From allylic alcohols to aldols through a new nickel-mediated tandem reaction: Synthetic and mechanistic studies

Article abstract of DOI:10.1002/chem.200501555

Nickel hydride type complexes have been successfully developed as catalysts for the tandem isomerization-aldolization reaction of allylic alcohols with aldehydes. Optimization of the reaction conditions has shown that a cocatalyst, such as MgBr₂, has a very positive effect on the kinetics of the reaction and in the yields of aldols. Under such optimized conditions {[NiHCl(dppe)] + MgBr₂ at 3-5 mol %)}, this reaction affords the aldols in good to excellent yields. It is a full-atom-economy-type reaction that occurs under mild conditions. Furthermore, it has a broad scope for the allylic alcohols and it is compatible with a wide range of aldehydes, including very bulky derivatives. The reaction is completely regioselective, but it exhibits a low stereoselectivity, except for allylic alcohols with a bulky substituent at the carbinol center. The use of chiral non-racemic catalysts was not successful, affording only racemic compounds. However, it was possible to use asymmetric synthesis for the preparation of optically active aldols. Various mechanistic studies have been performed using, for instance, a deuterated alcohol or a deuterated catalyst. They gave strong support to a mechanism involving first a transition-metal-mediated isomerization of the allylic alcohol into the free enol, followed by the addition of the latter intermediate onto the aldehyde in an "hydroxyl-carbonyl-ene" type reaction. These results confirm that allylic alcohols can be considered as new and useful partners in the development of the aldol reaction.

Full text of DOI:10.1002/chem.200501555

FULL PAPER
DOI: 10.1002/chem.200501555
From Allylic Alcohols to Aldols through a New Nickel-Mediated Tandem
Reaction: Synthetic and Mechanistic Studies
David Cuperly,[b] Julien Petrignet,[a] Christophe CrØvisy,[b] and RenØ GrØe*[a]
Abstract: Nickel hydride type com-
plexes have been successfully devel-
oped as catalysts for the tandem isom-
erization–aldolization reaction of allylic
alcohols with aldehydes. Optimization
of the reaction conditions has shown
that a cocatalyst, such as MgBr2, has
a very positive effect on the kinetics
of the reaction and in the yields of
aldols. Under such optimized condi-
tions {[NiHCl(dppe)] + MgBr2 at 3–
5 mol%)}, this reaction affords the
aldols in good to excellent yields. It is a
full-atom-economy-type reaction that
occurs under mild conditions. Further-
more, it has a broad scope for the allyl-
ic alcohols and it is compatible with a
wide range of aldehydes, including very
bulky derivatives. The reaction is com-
pletely regioselective, but it exhibits a
low stereoselectivity, except for allylic
alcohols with a bulky substituent at the
carbinol center. The use of chiral non-
racemic catalysts was not successful, af-
fording only racemic compounds. How-
ever, it was possible to use asymmetric
synthesis for the preparation of optical-
ly active aldols. Various mechanistic
studies have been performed using, for
instance, a deuterated alcohol or a deu-
terated catalyst. They gave strong sup-
port to a mechanism involving first a
transition-metal-mediated
isomeriza-
tion of the allylic alcohol into the free
enol, followed by the addition of the
latter intermediate onto the aldehyde
in an “hydroxyl–carbonyl–ene” type re-
action. These results confirm that allyl-
ic alcohols can be considered as new
and useful partners in the development
of the aldol reaction.
Keywords: aldol reaction · allylic
compounds · ene reaction · enols ·
nickel
Introduction
cessful developments include the uses of: 1) catalytic anti-
bodies,[3] 2) proline and analogues derived catalysts (“organ-
ocatalysis”),[4] and 3) transition-metal complexes.[5] Our goal
was to develop a new, full-atom-economy approach to the
aldol reaction starting from easily available allylic alcohols,
as indicated in Scheme 1.
The transposition of allylic alcohols 1 into saturated car-
bonyls 2, mediated by transition-metal catalysts (TM), is a
known process.[6] Mechanistic studies have strongly suggest-
The aldol reaction is one of the most commonly used reac-
À
tions to create new C C bonds. The original direct aldol
condensation between two carbonyl components was per-
formed under acid or base catalysis.[1] It is an early example
of a full-atom-economy reaction,[2] but the lackof regio- and
stereocontrol, as well as the presence of evolution products
(such as alkenes by crotonization), were some of the limita-
tions of this process. However, in the last 10 years many ef-
forts have been devoted to overcome these limitations and
to obtain fully controlled direct aldol reactions. Highly suc-
[a] J. Petrignet, Dr. R. GrØe
UniversitØ de Rennes1
Laboratoire de Synth›se et Electrosynth›se Organiques
CNRS UMR 6510, Avenue du GØnØral Leclerc
35042 Rennes Cedex (France)
Fax : (+33)223-236-978
E-mail: rene.gree@univ-rennes1.fr
[b] Dr. D. Cuperly, Dr. C. CrØvisy
ENSCR, Laboratoire de Synth›ses et Activations de BiomolØcules
CNRS UMR 6052, Avenue du GØnØral Leclerc
35700 Rennes Beaulieu (France)
Scheme 1. Working hypothesis for the tandem isomerization–aldolization
process.
Chem. Eur. J. 2006, 12, 3261 – 3274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3261

ed that it involves, as late intermediates, type A enols com-
plexed to the transition-metal catalyst as well as the free
enols B. Therefore, as a working hypothesis we considered
the possibility to trap such intermediates by aldehydes in a
new aldol-type process. Support to this was given by an
a first step, four commercially available complexes 3–6 have
been selected.
early workindicating that RuCl could perform an isomeri-
3
zation–aldolization followed by crotonization.[7] Further-
more, previous results from Bosnichꢁs group have demon-
strated that it is possible to trap free enols by strong electro-
philes, such as iminium salts or TCNE.[8] Finally, the results
from Motherwellꢁs group demonstrating the possible use of
rhodium or nickel catalysts to isomerize allylic alcoholates
into the corresponding enolates and trap the latter deriva-
tives in aldol reactions are also to be noticed.[9] Our first ex-
periments were performed by using iron–carbonyl-derived
catalysts, which proved to be very efficient for this new
tandem isomerization–aldolization process.[10] The corre-
sponding reactions occurred under very mild conditions and
gave, in most cases, good yields in aldol products. However,
they were not fully regiocontrolled and the stereocontrol
was also poor. Therefore, other transition-metal complexes
were screened for this tandem reaction. Among them, rhodi-
um and ruthenium derivatives were demonstrated as possi-
ble catalysts: they perform this reaction with a complete re-
giocontrol, but with a low stereocontrol and a more limited
scope.[11,12] Nickel catalysts appeared of much interest to us,
since such derivatives have been used in the isomerization
of allylic alcohols into saturated carbonyls.[13] Furthermore,
they have also been successfully developed recently in the
asymmetric isomerization of cyclic allylic ethers into vinyl
ethers in very good yields and excellent enantiomeric ex-
cesses (eeꢁs).[14] Therefore the purpose of this paper is: 1) to
describe our efforts in the development of new nickel hy-
dride catalysts for the tandem isomerization–aldolization of
allylic alcohols, 2) to define the scope and limitations of
these new catalysts, 3) to report our results regarding their
extension to asymmetric synthesis and catalysis, and 4) to
present the results of experiments using labeled compounds.
Based on these data, a tentative mechanism will be pro-
posed for these reactions: the transition-metal catalyst indu-
ces first the isomerization of the allylic alcohols into the free
enols. In a second step, these derivatives react with the alde-
hydes in an “hydroxyl–carbonyl–ene” type reaction to
afford the aldol products.[15] Such a mechanism not only ac-
counts for the experimental results, but it is also in full
agreement with the conclusions obtained from recent high-
level computational studies, using iron–carbonyl species as
catalysts.[16]
Essentially two methods have been developed to prepare
the nickel hydrides: they involve the reaction of nickel di-
chloride complexes either with LiBHEt3 (Method A),[17] or
with a combination of isopropyl Grignard and Me3SiCl
(Method B, Scheme 2).[18]
Scheme 2. Methods used for the synthesis of the nickel hydride catalysts.
Therefore, the first step of our study was to select the ap-
propriate catalysts and to optimize the reaction conditions.
The reaction of octen-3-ol (1a) with benzaldehyde was
chosen as a model reaction to screen the activity of the cata-
lysts derived from the four nickel complexes 3–6
(Scheme 3).
Scheme 3. Model reaction used for the selection of the catalyst.
In a first series of experiments, these catalysts were gener-
ated by addition of one molar equivalent of the Super-Hy-
drideTM solution to the nickel complexes in THF (Meth-
od A), followed by addition of octen-3-ol and benzaldehyde
at room temperature. The results are given in Table 1: at
5 mol%, the catalysts derived from complexes 3 and 4 were
completely inactive and only the starting products were re-
covered (entries 1 and 2). In contrast, in the case of com-
plexes 5 and 6, a total conversion was observed and the
aldol products 7 could be obtained in 44 and 32% yield, re-
spectively, together with the isomerization product 2a. Only
a low stereoselectivity was obtained in favor of the syn
isomer (entries 3 and 5). However, it was of much interest
to notice that this reaction was highly regiocontrolled, since
no trace of the regioisomeric aldols could be detected by
Results and Discussion
Development of new nickel hydride catalysts for the tandem
isomerization–aldolization: The nickel hydrides appeared as
very attractive candidates as catalysts, since they can be pre-
pared from nickel dichloride precursors bearing a large vari-
ety of ligands L2, including chiral nonracemic derivatives. In
3262
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3261 – 3274

Nickel Catalysts
FULL PAPER
Table 1. Selection of the nickel hydride catalysts.
this new catalytic system, it
became possible to use only
3 mol% of the catalysts
(entry 4) and furthermore, sim-
ilar results were obtained with
benzaldehyde (entries 5, 6).
Therefore, the magnesium salt
plays a key role as a cocatalyst
for this reaction. Such a cata-
lytic effect of MgBr2 has al-
ready been observed in aldol
reactions and this could be ex-
plained by the activation of the
carbonyl component by the
Precatalyst R4
T [8C]
t [min] Conversion [%] Aldol yield [%] syn/anti 2a [%]
1
2
3
4
5
6
7
8
3
4
5
5
6
6
5
5
5
5
5
Ph
Ph
Ph
Ph
Ph
Ph
H
RT
RT
RT
120
240
100
0
0
–
–
–
–
–
–
100
100
100
100
100
100
100
100
100
7 (44)
7 (81)
7 (32)
7 (37)
8 (30)
9 (26)
10 (15)
11 (26)
12 (18)
(60:40)
(60:40)
(74:26)
(67:33)
–
(68:32)
(72:28)
(76:24)
(70:30)
32
15
43
33
53
39
60
51
64
À50 to RT 120
RT
30
90
À50 to RT
À50 to RT 300
À50 to RT 240
CH3
9
10
11
nC5H11 À50 to RT 170
iBu
iPr
À50 to RT 120
À50 to RT 165
NMR analysis of the crude reaction mixtures. Lowering the
temperature afforded a significant improvement in yield in
the case of the complex 5 (entry 4), but no significant
change was observed in the case of 6 (entry 6). The same re-
action conditions were applied to formaldehyde and several
aliphatic aldehydes (entries 7 to 11), but the yields remained
low (<30%).
Lewis acid character of this salt.[19] The same results were
obtained starting from the diferrocenyl complex 6 (entry 7).
Using the same model reaction and under these optimized
reaction conditions, a large variety of other salts have been
evaluated as possible cocatalysts (Table 3).
From these results, it is clear that many of them can be
used in this reaction, except the very strong Lewis Acids
such as BiCl3 or TiCl4. It is worth noting that for these
nickel hydride catalysts, In(OAc)3 is not very efficient
So, clearly, a further optimization of the nature of the cat-
alyst and reaction conditions had to be performed and this
was done using octen-3-ol and
Table 2. Optimization of the reaction conditions.
isovaleraldehyde as models
(Table 2). We first checked the
method used for the prepara-
tion of the nickel hydride cata-
lyst: to our surprise, by using
the second method with the
Grignard reagent and Me3SiCl,
a quantitative transformation
was observed and the aldols
were obtained in excellent
overall yield (94%) with only
traces of the octan-3-one
(entry 2).
R4
Catalyst
amount [mol%]
Method Co-reagent
t [min] Aldol (yield [%]) syn/anti 2a [%]
(amount [mol%])
1
2
3
4
5
6
7
iBu
iBu
iBu
iBu
Ph
5
5
5
3
A
B
–
–
120
85
40
50
225
50
11 (26)
11 (94)
11 (95)
11 (92)
7 (85)
76:24
60:40
70:30
70:30
60:40
60:40
62:38
51
1
–
2
–
A
A
B
MgBr2 (5)
MgBr2 (3)
–
5
Ph
Ph
3
A
A
MgBr2 (3)
MgBr2 (3)
7 (99)
7 (97)
–
–
3[a]
60
[a] In this reaction [NiCl2(dppf)] (6) was used as the precatalyst.
Table 3. Isomerization–aldolization in the presence of various Lewis acid type salts.
R4
Co-reagent Catalyst t [min] Conversion [%] Aldol
(amount [mol%]) amount [mol%] (amount [%])
7 (90)
syn/anti 2a [%]
This was a strong indication
that, besides the nickel hy-
dride, some other important
co-reagent was present when
the second method was used.
Based on the proposed mecha-
nism, it appeared that magne-
sium salts could be generated
1
2
3
4
5
6
7
8
9
Ph MgCl2 (5)
Ph YCl3 (5)
Ph YbCl3 (5)
Ph InCl3 (7)
Ph CaCl2 (5)
Ph BiCl3 (5)
Ph TiCl4 (5)
Ph MgBr2 (3)
Ph LiBr (5)
5
5
5
7
5
5
5
3
5
5
7
5
7
5
3
5
5
7
7
5
5
5
45
75
105
130
50
1260
480
50
30
90
60
120
375
120
50
50
80
100
100
100
100
100
0
63:37
50:50
55:45
73:27
68:32
–
7
traces
8
7 (89)
7 (87)
7 (67)
7 (66)
–
10
11
–
0
–
–
–
100
100
100
100
100
25
100
100
100
100
100
100
100
77
7 (99)
7 (71)
7 (63)
7 (48)
7 (63)
7 (15)
7 (81)
11 (92)
11 (87)
11 (87)
11 (40)
11 (38)
11 (47)
11 (40)
11 (26)
60:40
60:40
73:27
74:26
66:34
78:22
60:40
70:30
58:42
63:37
74:26
78:22
70:30
70:30
76:24
–
conditions.[18]
20
30
46
20
5
15
2
9
13
34
54
41
20
51
under
those
10 Ph InBr3 (5)
Therefore, in a complementary
experiment, one equivalent of
MgBr2 was added to the cata-
lyst generated by the first
11 Ph ZnBr2 (7)
12 Ph Ti(OiPr)4 (5)
13 Ph In(OAc)3 (7)
14 Ph
–
15 iBu MgBr2 (3)
16 iBu YCl3 (5)
17 iBu YbCl3 (5)
18 iBu InCl3 (7)
19 iBu ZnBr2 (7)
20 iBu LiBr (5)
21 iBu Ti(OiPr)4 (5)
method
(using
LiBHEt3)
before starting the reaction
with the allylic alcohol and the
aldehyde: under those condi-
tions, the aldols were obtained
in excellent yield (95%) and
in only 40 min (entry 3)! With
100
75
40
960
120
22 iBu
–
100
Chem. Eur. J. 2006, 12, 3261 – 3274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3263

R. GrØe et al.
Table 4. Reactions of octen-3-ol with various types of aldehydes.
(entry 13), contrary to the results obtained with ruthenium
catalysts in water and protic solvents.[11b] In our case, MgBr2
proved to be the most active, since it is the only one allow-
ing a total conversion of the starting material at doses below
5 mol%. Similar results have been obtained by using isova-
leraldehyde as the carbonyl component and MgBr2 proved
to be also the best cocatalyst. However with this aliphatic al-
dehyde, the reactions performed in the presence of the
Lewis acids gave always improved yields as compared to the
reactions with the nickel hydride alone. It is worth noting
that the nature of the cocatalyst does not have any critical
influence on the diastereoselectivity of the reaction. Further-
more, using the benzaldehyde adducts as models, we have
established that this tandem isomerization–aldolization reac-
tion is occurring under kinetic control: in the presence of
the catalytic system (nickel hydride either with, and without,
MgBr2) both pure diastereoisomers 7 syn and 7 anti are
found to be stable. There is neither epimerization nor retro-
aldolization observed under these reaction conditions
(Scheme 4). This result is also in sharp contrast to the data
obtained with ruthenium catalysts used in dynamic kinetic
resolution and racemization processes.[20]
R4
t [min] Aldol Yield [%] syn/anti 2a [%]
1
2
3
4
5
6
Ph
50
50
60
150
55
60
7
11
12
13
14
15
99
92
93
84
68
91
60:40
70:30
66:34
69:31
76:24
67:33
–
2
2
4
15
–
Me2CHCH2
Me2CH
Et2CH
tBu
p-acetamidophenyl
7
45
16
94
57:43
–
The reaction was then extended to various types of allylic
alcohols with substituents on the different positions
(Scheme 5 and Table 5)
Scheme 5. Tandem isomerization–aldolization of various allylic alcohols
with two model aldehydes.
Starting from the 2-methylocten-3-ol (1b), neither the
isomerization nor the aldol reactions were observed and the
starting materials were fully recovered (entry 1). Therefore,
a substituent in the R2 position clearly inhibits the action of
the transition-metal catalyst. In contrast, it was possible to
introduce a methyl group in position R3: although the reac-
tion proceeded more slowly, 1c afforded in a good yield
(80%) the desired aldol products (entry 2). Finally, a large
variety of substituents can be introduced in position R1, as
in the alcohols 1d to 1h. In each case, excellent yields were
obtained, including for very bulky groups such as the tBu or
the gemdimethyl ester system (entries 3 to 10). Furthermore,
it is worthy of note that with the last substituents, the dia-
stereoselectivity became excellent in favor of the syn isomer
(over 90 and 95% respectively, entries 8 to 10). The relative
configuration of these aldols have been attributed as fol-
lows: for known compounds (7,[21] 19,[22] 20,[22a,23] 21,[22a]
24[22a]) the syn/anti configurations were established by com-
parison with literature data. For the other aldols, the follow-
Scheme 4. a) 7 syn (or
LiBHEt3 (5 mol%), MgBr2 (5 mol%), THF, À508C to RT, 1.5 h.
7 anti) (1 equiv), [NiCl2(dppe)] (5 mol%),
With this optimized catalytic system in hand, it became
possible to extend this reaction to a large variety of alde-
hydes (Table 4).
The reaction gave excellent yields with aromatic (en-
tries 1, 6), heteroaromatic (entry 7) as well as with aliphatic
aldehydes (entries 2 to 5). It is particularly noteworthy that
highly sterically hindered aldehydes (including pivalalde-
hyde) gave very high yields in aldol products. In all cases,
the diastereoselectivity remained low (2 or 3 to 1 in favor of
the syn diastereoisomer).
In the case of the allylic
Table 5. Tandem isomerization–aldolization of allylic alcohols with model aldehydes.
ether isomerization by nickel
hydrides, the reaction was sen-
sitive to the nature of the halo-
gens on the nickel and higher
yields were obtained with the
iodo complexes.[14] However, in
our isomerization–aldol reac-
tion, the iodo complex [NiI2-
(dppe)] gave the same results
as the corresponding dichlo-
ride.
1
R1
R2
R3
R4
Cat. [%]
t [min]
Aldol (yield [%])
syn/anti
1
2
3
4
5
6
7
8
9
1b
1c
1d
1d
1e
1 f
1 f
1g
1h
1h
nC5H11
nC5H11
Ph
Me
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
Ph
Ph
Ph
Me2CH
Ph
Ph
Me2CHCH2
Ph
Ph
7
7
3
3
5
5
5
5
5
3
240
1260
75
70
40
105
45
30
120
70
17 (0)
–
18 (80)
19 (97)
20 (86)
21 (68)
22 (90)
23 (88)
24 (89)
25 (93)
26 (88)
55:45
60:40
50:50
75:25
70:30
69:31
90:10
>95:5
>95:5
Ph
Me2CH
Et2CH
Et2CH
tBu
C(Me)2CO2Et
C(Me)2CO2Et
10
Me2CHCH2
3264
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3261 – 3274

Nickel Catalysts
FULL PAPER
ing empirical rule was followed: The C atom of the CHOH
group is deshielded in the anti adduct relative to that of the
syn adduct, while the H atom of the CHOH group is shield-
ed in the anti adduct. Moreover, except in one case (14), the
3J(H,H) coupling constant between the CHOH group and
the CHC=O group is larger in the anti aldol.
tion of the alcohol 1d with benzaldehyde. This indicates that
the chiral substituent vicinal to the carbonyl group has no
À
effect on the diastereoselectivity of this C2 C3 bond forma-
À
tion. On the other hand, if one considers the C3 C4 rela-
tionship, there is a good stereoselectivity in favor of the anti
diastereoisomer: in each case the facial stereoselectivity in-
duced by the chiral substituent is around 80:20.
Development of the tandem isomerization–aldolization in
asymmetric synthesis and asymmetric catalysis: The next
step in the development of this reaction was the extension
to asymmetric synthesis. For the corresponding studies we
selected two model allylic alcohols 1d and 1h and the chiral
lactaldehydes 27 and 29 bearing two different types of alco-
hol protective groups.
To obtain more information on the effect of the substitu-
ents, this reaction was extended to the alcohol 1h having a
very bulky substituent at the carbinol center and to the alde-
hydes 27 and 29.[26] In each case, the reaction afforded a
mixture of two aldol products in excellent yields
(Scheme 7). Their stereochemistry was established by NMR
spectroscopy, using the previously described criterion.
Under the previously optimized reaction conditions, the
addition of the allylic alcohol 1d to the 2-(S)-benzyloxypro-
panal 27[24] afforded a mixture of four diastereoisomers
(28a–d) in 79% overall yield (Scheme 6).
Scheme 7. a) 1h (1 equiv), 27 (or 29) (1.1 equiv), [NiCl2(dppe)]
(5 mol%), LiBHEt3 (5 mol%), MgBr2 (5 mol%), THF, À508C to RT,
1.5 h, overall yield 95%.
In agreement with the results obtained previously with
achiral aldehydes, this allylic alcohol afforded only the syn
diastereoisomers at the C2/C3 level. The facial diastereose-
lectivity was again excellent with the following syn/anti
ratios at the C3/C4 level: 78:22 for 30a and 30b, and 83:17
in the case of 31a and 31b. These stereoselectivity results
will be discussed in the last part of this publication.
Scheme 6. a) 1d (1 equiv), 27 (1.1 equiv), [NiCl2(dppe)] (5 mol%),
LiBHEt3 (5 mol%), MgBr2 (5 mol%), THF, À508C to RT, 1 h10 min,
overall yield 79%.
Their stereochemistry has been established by comparison
1
of their H and 13C NMR data with literature.[25] A careful
study of the latter data (Table 6) showed that the diaste-
After these encouraging results in asymmetric synthesis,
we were interested in the possibility to develop into asym-
metric catalysis. Taking into account the excellent results ob-
tained in the asymmetric isomerization of cyclic allyl ethers,
we selected three model nickel dichlorides 32, 33, and 34 as
precatalysts and studied the reactions with two different al-
lylic alcohols, the derivatives 1a and 1h. NMR measure-
Table 6. 13C Chemical shift comparison for the methyl groups in aldol
products 28a to 28d.
28a
28b
28c
28d
d(CH3) [ppm]
d(CH3) [ppm]
Dd [ppm]
11.8
15.5
3.7
14.2
15.8
1.6
15.6
15.7
0.1
15.2
15.7
0.5
reoisomer 29a had a strongly shielded methyl group (d=
11.8 ppm), together with a very large chemical shift differ-
ence between the two methyl groups (Dd=3.7 ppm versus
0.1 to 1.6 ppm). This characteristic appears general and will
be of much use with the other models. The ratio between
1
the diastereoisomers has been established by H NMR spec-
troscopy. The ratio between the isomers having the syn
versus the anti configuration at the C2/C3 level (28a and
28b versus 28c and 28d, respectively) was 60:40; it is worth
noting that exactly the same ratio was observed for the reac-
ments in the presence of a chiral shift reagent were selected
as the analytical technique to measure the eeꢁs, since we
have established that the corresponding racemic aldol ad-
ducts, previously prepared by using the dppe catalyst, exhib-
Chem. Eur. J. 2006, 12, 3261 – 3274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3265

R. GrØe et al.
ited excellent separations of the signals of the carbinol pro-
tons in the presence of [Eu(tfc)3].
The results of these reactions are given in Table 8. Here
again, the yields and selectivities were similar to those ob-
tained previously with the dppe complex. Analysis by NMR
spectroscopy with the same chiral shift reagent gave again a
The complexes 32 and 33 are not commercially available
and they were prepared following a literature method.[27]
The Me-DuPHOS complex 34 was obtained by the method
of Frauenrath.[14] The corresponding nickel hydride catalysts
were prepared by reaction of the nickel dichlorides with one
equivalent of LiBHEt3 and used under the previously opti-
mized reaction conditions, in the presence of one molar
equivalent of MgBr2 as cocatalyst (Scheme 8). The results of
the reactions performed starting from alcohol 1a are given
in Table 7.
Table 8. Isomerization–aldolization of alcohol 1h with two chiral nickel
hydride catalysts.
Catalyst
Yield [%]
syn/anti
syn (ee)
32
33
96
96
>95:5
94:6
0
0
disappointing 0% ee for these reactions. We had demon-
strated earlier that these reactions are under kinetic control,
excluding both epimerizations and reversibility. Therefore, a
racemization process on both stereocenters of these aldol
adducts, seems also very unlikely. As a conclusion of these
experiments, it appears that the chiral ligands on the transi-
tion-metal catalysts have no effect on the enantioselectivity
of this aldol reaction. As will be discussed later, a likely ex-
planation for this result is that the transition-metal complex
Scheme 8. Isomerization–aldolization reaction from alcohol 1a with three
chiral nickel hydride catalysts.
À
is not involved in the key C C bond-forming step in the al-
Table 7. Isomerization–aldolization of alcohol 1a with three chiral nickel
hydride catalysts.
dolization reaction: therefore, this step has to be envisaged
at later stage and the free enol appears as a very plausible
candidate as the key intermediate.
Catalyst
Yield [%]
syn/anti
syn (ee)
anti (ee)
32
33
34
95
91
23
60:40
62:38
60:40
0
0
0
0
0
0
Mechanistic studies using labeled compounds: Two sets of
experiments with deuterium labeled compounds were per-
formed in order to obtain more information about the
mechanism of this tandem isomerization–aldolization reac-
tions. The first reaction involved the use of a deuterated cat-
alyst, prepared from LiBDEt3 and MgBr2. Furthermore,
0.5 molar equivalent of the catalytic system was used in
order to minimize the risks of losing some of the label in
the experiment. The model reaction between octen-3-ol and
benzaldehyde was performed with this labeled catalyst and
the results are reported in Scheme 10. The reaction afforded
Starting with the PROPHOS and the CHIRAPHOS cata-
lysts, excellent chemical yields were obtained and the syn/
anti selectivities were also very similar to the results ob-
tained with the dppe complex. The diastereoisomeric aldols
were separated by chromatography and their optical purity
1
established by H NMR spectroscopy in the presence of the
chiral shift reagent [Eu(tfc)3]. With both the chiral catalysts,
only racemic mixtures were obtained. The same reaction
was also performed with the Me-DuPHOS catalyst: in that
case, a lower yield in aldols (23%) was obtained and only
the syn diastereoisomer could be isolated in pure form.
Analysis by 1H NMR spectroscopy in the presence of the
chiral shift reagent indicated again that a racemic mixture
was obtained for this compound. In conclusion, the three
model chiral catalysts afforded only racemic mixtures of
aldols from this allylic alcohol. Therefore, it was of much in-
terest to study another alcohol, such as 1h, which was
known to give a higher syn selectivity in the aldol reactions
(Scheme 9).
Scheme 10. Conditions: 1a (1 equiv), PhCHO (1.1 equiv), [NiCl2(dppe)]
(0.5 equiv), LiBDEt3 (0.5 equiv), MgBr2 (0.5 equiv), THF, À508C to RT,
40 min, overall yield 92%.
excellent yields in the aldol products in the usual 2:1 syn/
anti selectivity and analysis by high-field NMR spectroscopy
indicated that deuterium was not incorporated in the prod-
ucts. This result clearly excludes a mechanism in which the
catalyst reacts by an addition–elimination sequence: in this
last case deuterium would have been introduced either on
the methyl group, or on the next carbon atom or even on
both carbon atoms.[6a]
Scheme 9. Conditions: 1h (1 equiv), PhCHO (1.1 equiv), 32 (or 33)
(5 mol%), LiBHEt3 (5 mol%), MgBr2 (5 mol%), THF, À508C to RT,
1.5 h, overall yield 96%.
A second series of experiments was performed, starting
from a deuterated allylic alcohol. The compound 36, select-
3266
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3261 – 3274

Nickel Catalysts
FULL PAPER
ed as a model for this study, was prepared by reduction of
the enone 35 with NaBD4 (Scheme 11) and H NMR analy-
sis gave a 95% deuterium content for 36.
Tentative mechanistic proposal for the nickel hydride medi-
ated tandem isomerization–aldolization: Based on the re-
sults described in the previous parts, we propose the mecha-
nism described in Scheme 13. The reaction of the allylic al-
cohol with the nickel hydride generates first an alcoholate
complexed to the nickel catalyst (a). A b-hydride elimina-
tion process leads to the enone coordinated to the nickel hy-
dride (b). Then, nonregioselective and reversible additions
on the double bond, afford the type d and c intermediates.
The latter compounds are known to exist in the s-bound
form c, in equilibrium with the nickel enolate form c’ as well
as with the p-oxo allyl complex c’’.[28] It is important to note
that this type of mechanism has been already proposed in
the isomerization of allylic alcohols, mediated by ruthenium
and rhodium complexes.[29,11d]
This type c intermediate reacts with another molecule of
allylic alcohol to regenerate the corresponding alcoholate a
and to liberate the free enol e. The latter key intermediate
can either tautomerize to the saturated carbonyl compound
h or react with the aldehyde to afford the aldol product g.
The different lines of evidence in favor of such a mechanis-
tic proposal are the following:
1
Scheme 11. a) 1d (1 equiv), IBX (1.5 equiv), DMSO, RT, 80%; b) NaBD4
(1 equiv), CeCl3 (1 equiv), MeOH, RT, 77%.
Under the same reaction conditions as before (with
3 mol% catalyst) this alcohol afforded a complex mixture of
four aldol products 19, 37, 38, and 39 in excellent yield
(Scheme 12). A careful analysis by mass spectrometry indi-
*
The nickel enolates, which could be considered as alter-
native intermediates in this reaction, have been clearly
excluded. The nickel enolate derived from ketone 40 was
prepared following literature procedure for similar com-
pounds.[28] In the presence of benzaldehyde (with or with-
out MgBr2), no reaction was obtained and only the start-
ing materials were recovered (Scheme 14).[30] This result
is in agreement with the literature data indicating that
the nickel enolates derived from ketones have a very low
Scheme 12. a) 36 (1 equiv), [NiCl2(dppe)] (3 mol%), LiBHEt3 (3 mol%),
MgBr2 (3 mol%), PhCHO (1.1 equiv), THF, À508C to RT, 50 min, over-
all yield 98%.
cated a mixture of a nondeuterated aldol 19 (38%), a mono-
deuterated species (44%), and
a bisdeuterated aldol (18%).
1
A further analysis by H and
13C NMR spectroscopy con-
firmed these data and gave in-
formation on the position of
the deuterium label: within the
monodeuterated compounds
the derivative with the label
on the methyl group 37 repre-
sented 20–30%, while the de-
rivative with the deuterium in
central position 38 was in the
range 14–24%. For the bisdeu-
terated compound 39, the deu-
terium atoms were both on the
methyl group and also on the
vicinal carbon atom.
These results clearly ex-
cludes a mechanism through
p-allyl intermediates:[6a] in
such
a case, the deuterium
label should be exclusively on
the methyl group (affording
type 37 aldols).
Scheme 13. Mechanistic proposal for the tandem isomerization–aldolization, mediated by nickel hydride
catalysts.
Chem. Eur. J. 2006, 12, 3261 – 3274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3267

R. GrØe et al.
*
If we use a Zimmermann–Traxler transition state with
the hydrogen atom coordinating the two oxygen atoms, it
is also possible to rationalize the trends in the syn/anti
stereoselectivity of this aldol reaction (Scheme 15). The
Scheme 14. a) LDA (1 equiv), THF, À788C, 1 h, then [NiCl2(dppe)]
(1 equiv), À308C, 3 h, then MgBr2 (1 equiv) and PhCHO (1.1 equiv),
À508C to RT, 12 h; b) LDA (1 equiv), THF, À788C, 1 h, then MgBr2
(1 equiv), À788C, 1.5 h, then PhCHO (1.1 equiv), À508C to RT, 12 h.
reactivity, affording very small amount of the expected
aldol products.[28]
*
The magnesium enolates, which could be also considered
Scheme 15. Aldol reactions starting from (Z)- and (E)-enols.
as possible intermediates by metal exchange from the
nickel enolates, were unambiguously excluded as well.
The magnesium enolate derived from ketone 40 was pre-
pared independently from lithium enolate. Under our
standard reaction conditions, it afforded only the reduc-
tion product (benzyl alcohol) and no aldols (Scheme 14).
Furthermore it is worth mentioning that the magnesium
enolates are known to afford mainly the anti adducts,[31a]
eventually by means of epimerization processes,[31b] while
our reaction affords the syn adducts as major compounds
(>90% in the case of 1h, the allylic alcohol correspond-
ing to 40). Finally, a transformation of the free enol into
the magnesium enolate appears also very unlikely, since
it would generate HBr, which is a very efficient catalyst
for the tautomerization of enols into saturated carbon-
yls.[8] Therefore, the role of MgBr2 as a cocatalyst appears
to be a Lewis acid type activation of the carbonyl com-
pound.
isomerization of the allylic alcohol can afford either the
(Z)-enol (leading to the syn aldols) or the (E)-enol (af-
fording the anti aldols). If we assume that the steric inter-
action between the groups R1 and R’ controls the enol
geometry, then it should be directly connected to the syn/
anti ratios. This tendency has been indeed observed in
this tandem reaction (Table 9): for groups with a limited
steric bulkthe syn/anti ratios remained low (2:1 to 3:1).
In contrast, the bulky R1 groups such as the tBu or the
gemdimethyl ester afforded very high syn selectivities
(over 90–95%).
Table 9. Syn/anti stereoselectivity dependence as a function of the nature
of the R1 group.[a]
Allylic alcohol
R1
syn
60
75
70
90
>95
anti
*
As demonstrated in the previous part, various chiral li-
1a
1e
1 f
1g
1h
nC5H11
Me2CH
Et2CH
tBu
CMe2CO2Et
40
25
30
10
<5
gands on the nickel catalysts have no effect on the enan-
tioselectivity of the reaction. This is a strong indication
that the transition-metal complex is no longer involved
À
during the formation of the key C C bond in the aldol.
[a] Data for: R4 =Ph; R’=Me.
*
This mechanism is in full agreement with previous com-
putational studies, although these have been performed
with a different catalyst (iron–carbonyl compounds).
These calculations have shown that the addition of the
free enol is the strongly favored pathway and this process
can be considered as an “hydroxyl–carbonyl–ene” type
reaction.[16a]
*
The facial selectivity can also be rationalized by using the
free enols as the key intermediates (Scheme 16).
Recent literature data have established that, for the aldol
reactions, the Felkin–Anh–Houk model is not suitable and
the Cornforth model has to be used preferentially.[32] In this
model, the polar group has to be situated as far as possible
of the carbonyl group to minimize the dipole moment. In
the case of the (Z)-enol, the transition state TS2 is strongly
disfavored by a strong syn-pentane interaction between the
two methyl groups. Such an interaction is not anymore pres-
ent in TS1 and explains the favored formation of the anti
*
It is fully consistent with our labeling studies: no deuteri-
um was obtained in the aldol product when the reaction
was performed with a deuterated catalyst. The allylic al-
cohol deuterated on the carbinol center afforded a mix-
ture of aldols: this is fully consistent with reversible addi-
tions, on both directions, of the nickel hydride on the
enone, together with ligand exchanges between the deu-
terated and nondeuterated b type intermediates.
3268
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3261 – 3274

Nickel Catalysts
FULL PAPER
alcohol into the free enol, fol-
lowed by the addition of the
latter intermediate onto the al-
dehyde in a “hydroxyl–carbon-
yl–ene” type reaction. These
results confirm that allylic al-
cohols can be considered as
new and useful partners in the
development of the aldol reac-
tion.
Experimental Section
General methods: Tetrahydrofuran
was distilled from sodium/benzophe-
none ketyl. All aldehydes were fresh-
ly purified before reactions. The
NMR data were obtained at 400 MHz
for 1H and 100 MHz for 13C. IR spec-
tra were recorded on a FT-IR instru-
ment by using NaCl plates. Elemental
analyses and mass spectral analyses
were performed at the Centre RØ-
gional de Mesures Physiques de
lꢁOuest
(CRMPO)
in
Rennes
Scheme 16. Transition-state models for the aldolization reactions starting from chiral aldehydes.
(France).
Allylic alcohols 1a and 1d are com-
mercially available. The others have
been prepared by standard proce-
C3C4 diastereoisomer 28a or 30a. This is in full agreement
dure: addition of Grignard (1e,[33] 1 f, 1g[33]) or organolithium derivatives
(1b,[34] 1c,[34] 1h[35]) on the corresponding aldehyde.
with the experimental results: the allylic alcohol 1h, with
the bulky C(Me)2CO2Et substituent on the carbinol center,
afforded a 78:22 mixture of 30a and 30b aldols. Starting
now from the (E)-enol two transition states are again possi-
ble; TS3 is disfavored with respect to TS4 by the syn inter-
action between the methyl group and the axial hydrogen.
Therefore, for these compounds, the anti C3/C4 isomer
should be favored: this is indeed observed in the addition of
allylic alcohol 1d with the protected lactaldehyde 27 (the
C3C4 anti/syn ratio was 81:19 for 28c/28d).
Procedure for aldolization of an allylic alcohol with an aldehyde by
means of a nickel complex as catalyst—method A (with MgBr2): A 1m
solution of LiBHEt3 in THF (57 mL, 0.0567 mmol) was added to a solu-
tion of [NiCl2(dppe)] (30 mg, 0.0567 mmol) in anhydrous THF (3 mL) at
room temperature under nitrogen. The reaction mixture was stirred at
room temperature for 5 min before to be transferred into a flaskcontain-
ing anhydrous MgBr2 (10 mg, 0.0567 mmol). This reaction mixture was
stirred at room temperature for a further 5 min and was cooled to
À508C. Then, the aldehyde (2.08 mmol) and allylic alcohol (1.89 mmol)
were added successively. The temperature was raised to room tempera-
ture and the reaction mixture was monitored by TLC until the disappear-
ance of the allylic alcohol. The reaction mixture was quenched with a sa-
turated solution of NH4Cl (15 mL) and the aqueous phase was extracted
with Et2O (3”50 mL). The organic phase was dried (MgSO4) and con-
centrated under reduced pressure. Purification by column chromatogra-
phy on silica gel afforded the desired aldol products.
Conclusion
In conclusion, we have demonstrated that the [NiHCl-
(dppe)]/MgBr2 combination is a very active catalytic system
for the tandem isomerization–aldolization reaction of allylic
alcohols with aldehydes. The reaction has a large scope in
term of allylic alcohols and it is compatible with a wide
range of aldehydes, including very bulky ones. This reaction
proceeds at, or below, room temperature and in the pres-
ence of a low catalyst loading (3 mol%). It affords aldols in
good to excellent yields and it is completely regioselective.
Although asymmetric catalysis could not be developed at
this stage, asymmetric synthesis can be used for the prepara-
tion of aldols in optically active form. Detailed mechanistic
studies afforded strong support to a mechanism involving
first a transition-metal-mediated isomerization of the allylic
The same procedure was used with [NiCl2(dppf)] as precatalyst.
Procedure for aldolization of an allylic alcohol with an aldehyde with a
nickel complex as catalyst—method B: A 1m solution of isopropylmagne-
sium bromide in THF (379 mL, 0.379 mmol) was added to a solution of
[NiCl2(dppe)] (100.0 mg, 0.189 mmol) in anhydrous THF (10 mL) at 08C
under nitrogen. The temperature was raised to room temperature, then
TMSCl (24 mL, 0.189 mmol) was added. The reaction mixture was stirred
at room temperature for 5 min and cooled to À508C. Then the aldehyde
(4.17 mmol) and allylic alcohol (3.79 mmol) were added successively. The
temperature was raised to room temperature and the reaction mixture
was monitored by TLC until the disappearance of the allylic alcohol. The
reaction mixture was quenched with a saturated solution of NH4Cl
(15 mL) and the aqueous phase was extracted with Et2O (3”50 mL). The
organic phase was dried (MgSO4) and concentrated under reduced pres-
sure. Purification by column chromatography on silica gel afforded the
desired aldol products.
Chem. Eur. J. 2006, 12, 3261 – 3274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3269

R. GrØe et al.
1-Hydroxy-2-methyl-1-phenyloctan-3-one (7): These aldols were obtained
by using the general procedure (method A) with [NiCl2(dppe)] (30 mg,
0.0567 mmol), a 1m solution of LiBHEt3 (57 mL, 0.0567 mmol), MgBr2
(10 mg, 0.0567 mmol), benzaldehyde (211 mL, 2.08 mmol), and 1a
(292 mL, 1.89 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 then 2:1 v/v) afforded a separable mixture of diastereoisomeric
17.1, 7.3 Hz, 1H; CH2CO), 2.53 (dt, J=17.1, 7.5 Hz, 1H; CH2CO), 2.73
(dq, J=2.9, 7.2 Hz, 1H; CHCO), 2.88 (d, J=3.2 Hz, 1H; OH), 3.51 ppm
(ddd, J=2.9, 3.2, 8.5 Hz, 1H; CHOH); 13C NMR (100 MHz, CDCl3): d=
9.3, 13.8, 18.9, 19.0, 22.4, 23.2, 30.5, 31.3, 41.6, 47.3, 76.2, 216.5 ppm.
Compound 12 anti: 1H NMR (400 MHz, CDCl3): d=0.89 (t, J=7.1 Hz,
3H; CH3CH2), 0.91 (d, J=6.7 Hz, 3H; (CH3)2CH), 0.96 (d, J=6.8 Hz,
3H; (CH3)2CH), 1.11 (d, J=7.2 Hz, 3H; CH3CH), 1.21–1.35 (m, 4H;
CH2CH2CH3), 1.52–1.61 (m, 2H; CH2CH2CO), 1.66–1.79 (m, 1H; CH-
(CH3)2), 2.46 (dt, J=17.4, 7.3 Hz, 1H), CH2CO, 2.55 (dt, J=17.4, 7.5 Hz,
1H; CH2CO), 2.58 (d, J=6.9 Hz, 1H; OH), 2.76 (dq, J=6.9, 7.2 Hz, 1H;
CHCO), 3.44 ppm (ddd, J=4.9, 6.9, 6.9 Hz, 1H; CHOH); 13C NMR
(100 MHz, CDCl3): d=13.8, 14.3, 15.9, 19.9, 22.4, 22.9, 30.4, 31.3, 42.9,
48.2, 78.3, 216.8 ppm. IR (neat): n˜ =3487 (w), 1705 cmÀ1; HRMS (EI;
70 eV): m/z calcd for C11H21O2: 185.1542 [MÀMe]+; found: 185.1570
(15 ppm); elemental analysis calcd (%) for (syn/anti mixture) C12H24O2:
C 71.95, H 12.08; found: C 71.74, H 12.24.
aldols as
a
colorless oil (syn/anti 62:38 by 1H NMR spectroscopy,
438.3 mg, 99% yield). The two diastereoisomers were separated by
column chromatography (pentane/Et2O, 10:1 v/v).
Compound 7 syn: 1H NMR (400 MHz, CDCl3): d=0.87 (t, J=7.2 Hz,
3H; CH3CH2), 1.07 (d, J=7.2 Hz, 3H; CH3CH), 1.18–1.35 (m, 4H;
CH2CH2CH3), 1.50–1.60 (m, 2H; CH2CH2CO), 2.33 (dt, J=17.2, 7.3 Hz,
1H; CH2CO), 2.45 (dt, J=17.2, 7.4 Hz, 1H; CH2CO), 2.83 (dq, J=4.0,
7.2 Hz, 1H; CHCH3), 3.20 (d, J=2.5 Hz, 1H; OH), 5.05 (dd, J=2.5,
4.0 Hz, 1H; CHOH), 7.23–7.38 ppm (m, 5H; HAr); 13C NMR (100 MHz,
CDCl3): d=10.5, 13.8, 22.4, 23.0, 31.2, 42.2, 52.4, 73.2, 125.9, 127.3, 128.2,
141.8, 216.0 ppm.
Compound 7 anti: 1H NMR (400 MHz, CDCl3): d=0.88 (t, J=7.1 Hz,
3H; CH3CH2), 0.94 (d, J=7.2 Hz, 3H; CH3CH), 1.15–1.35 (m, 4H;
CH2CH2CH3), 1.47–1.59 (m, 2H; CH2CH2CO), 2.41 (dt, J=17.2, 7.2 Hz,
1H; CH2CO), 2.51 (dt, J=17.2, 7.5 Hz, 1H; CH2CO), 2.92 (dq, J=8.1,
7.2 Hz, 1H; CHCH3), 2.96 (d, J=4.6 Hz, 1H; OH), 4.75 (dd, J=4.6,
8.1 Hz, 1H; CHOH), 7.23–7.38 ppm (m, 5H; HAr); 13C NMR (100 MHz,
CDCl3): d=13.9, 14.3, 22.4, 22.9, 31.2, 43.2, 52.7, 76.5, 126.5, 127.8, 128.4,
142.1, 215.8 ppm; elemental analysis calcd (%) for (syn/anti mixture)
C15H22O2: C 76.88, H 9.46; found: C 76.67, H 9.56.
3-Ethyl-4-hydroxy-5-methylundecan-6-one (13): These aldols were ob-
tained by using the general procedure (method A) with [NiCl2(dppe)]
(30 mg, 0.0567 mmol), a 1m solution of LiBHEt3 (57 mL, 0.0567 mmol),
MgBr2 (10 mg, 0.0567 mmol), 2-ethylbutanal (256 mL, 2.08 mmol), and 1a
(292 mL, 1.89 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 then 2:1 v/v) afforded a separable mixture of diastereoisomeric
aldols as
a
colorless oil (syn/anti: 69:31 by 1H NMR spectroscopy,
363.0 mg, 84% yield). The two diastereoisomers were separated by
column chromatography (pentane/Et2O, 6:1 v/v).
Compound 13 syn: 1H NMR (400 MHz, CDCl3): d=0.83–0.92 (m, 9H;
CH3CH2), 1.12 (d, J=7.2 Hz, 3H; CH3CH), 1.19–1.46 (m, 8H; CH2CH,
CH2CH2CH3), 1.53–1.70 (m, 3H; CHCHOH, CH2CH2CO), 2.45 (dt, J=
17.1, 7.3 Hz, 1H; CH2CO), 2.52 (dt, J=17.1, 7.5 Hz, 1H; CH2CO), 2.67
(d, J=3.4 Hz, 1H; OH), 2.73 (dq, J=3.2, 7.2 Hz, 1H; CHCO), 3.78 ppm
(ddd, J=3.2, 3.4, 6.7 Hz, 1H; CHOH); 13C NMR (100 MHz, CDCl3): d=
9.8, 10.2, 10.4, 13.9, 20.2, 20.8, 22.5, 23.3, 31.4, 41.8, 42.0, 47.1, 71.8,
216.7 ppm.
2,5-Dimethyl-4-hydroxyundecan-6-one (11): These aldols were obtained
by using the general procedure (method A) using [NiCl2(dppe)] (30 mg,
0.0567 mmol), a 1m solution of LiBHEt3 (57 mL, 0.0567 mmol), MgBr2
(10 mg, 0.0567 mmol), isovaleraldehyde (227 mL, 2.08 mmol), and 1a
(292 mL, 1.89 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 then 2:1 v/v) afforded a separable mixture of diastereoisomeric
aldols as
a
colorless oil (syn/anti: 70:30 by 1H NMR spectroscopy,
Compound 13 anti: 1H NMR (400 MHz, CDCl3): d=0.87–0.94 (m, 9H;
CH3CH2), 1.09 (d, J=7.2 Hz, 3H; CH3CH), 1.19–1.62 (m, 11H; CH2CH,
CH2CH2CH3, CH2CH2CO, CHCHOH), 2.46 (dt, J=17.3, 7.3 Hz, 1H;
CH2CO), 2.51 (d, J=6.6 Hz, 1H; OH), 2.55 (dt, J=17.3, 7.5 Hz, 1H;
CH2CO), 2.81 (dq, J=7.2, 7.2 Hz, 1H; CHCO), 3.71 ppm (ddd, J=3.9,
6.6, 7.2 Hz, 1H; CHOH); 13C NMR (100 MHz, CDCl3): d=11.4, 11.8,
13.9, 14.4, 20.4, 22.4, 23.0, 31.3, 42.9, 43.2, 48.2, 74.6, 216.8 ppm. IR
370.4 mg, 92% yield). The two diastereoisomers were separated by
column chromatography (pentane/Et2O, 8:1 v/v).
Compound 11 syn: 1H NMR (400 MHz, CDCl3): d=0.86–0.94 (m, 9H;
(CH3)2CH, CH3CH2), 1.01–1.09 (m, 1H; CH2CHOH), 1.12 (d, J=7.1 Hz,
3H; CH3CH), 1.21–1.35 (m, 4H; CH2CH2CH3), 1.46 (ddd, J=5.6, 9.4,
14.4 Hz, 1H; CH2CHOH), 1.57 (ddt, J=7.4, 7.4, 7.4 Hz, 2H;
CH2CH2CO), 1.70–1.82 (m, 1H; CH(CH3)2), 2.44 (dt, J=17.0, 7.4 Hz,
1H; CH2CO), 2.51 (dt, J=17.0, 7.6 Hz, 1H; CH2CO), 2.48–2.55 (m, 1H;
CHCO), 2.70 (brd, J=2.8 Hz, 1H; OH), 3.96–4.03 ppm (m, 1H;
CHOH); 13C NMR (100 MHz, CDCl3): d=9.9, 13.9, 22.0, 22.4, 23.2, 23.4,
24.5, 31.4, 41.9, 43.0, 50.2, 68.8, 216.6 ppm.
(neat): n˜ =3493 (w), 1703 cmÀ1
; HRMS (EI; 70 eV): m/z calcd for
C13H25O2: 213.1855 [MÀMe]+; found: 213.1853 (1 ppm); elemental analy-
sis calcd (%) for (syn/anti mixture) C14H28O2: C 73.63, H 12.36; found: C
73.37, H 12.54.
Compound 11 anti: 1H NMR (400 MHz, CDCl3): d=0.86–0.95 (m, 9H;
(CH3)2CH, CH3CH2), 1.12 (d, J=7.1 Hz, 3H; CH3CH), 1.15–1.43 (m,
6H; CH2CHOH, CH2CH2CH3), 1.52–1.62 (m, 2H; CH2CH2CO), 1.79–
1.90 (m, 1H; CH(CH3)2), 2.42 (dt, J=17.0, 7.4 Hz, 1H; CH2CO), 2.51
(dt, J=17.0, 7.6 Hz, 1H; CH2CO), 2.55–2.63 (m, 1H; CHCO), 2.57 (d,
J=6.6 Hz, 1H; OH), 3.70–3.79 ppm (m, 1H; CHOH); 13C NMR
(100 MHz, CDCl3): d=13.9, 14.2, 21.5, 22.5, 23.1, 23.8, 24.4, 31.4, 43.0,
44.0, 51.6, 71.8, 216.7; IR (neat): n˜ =3470 (w), 1703 cmÀ1; HRMS (EI;
70 eV): m/z calcd for C13H24O: 196.1827 [MÀH2O]+; found: 196.1822
(2 ppm); elemental analysis calcd (%) for (syn/anti mixture) C13H26O2: C
72.84, H 12.22; found: C 72.59, H 12.35.
3-Hydroxy-2,2,4-trimethyldecan-5-one (14): These aldols were obtained
by using the general procedure (method A) with [NiCl2(dppe)] (30 mg,
0.0567 mmol), a 1m solution of LiBHEt3 (57 mL, 0.0567 mmol), MgBr2
(10 mg, 0.0567 mmol), pivalaldehyde (226 mL, 2.08 mmol), and 1a
(292 mL, 1.89 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 then 2:1 v/v) afforded a separable mixture of diastereoisomeric
aldols as
a
colorless oil (syn/anti: 76:24 by 1H NMR spectroscopy,
275.5 mg, 68% yield). The two diastereoisomers were separated by
column chromatography (pentane/Et2O, 6:1 v/v).
Compound 14 syn: 1H NMR (400 MHz, CDCl3): d=0.90 (t, J=7.1 Hz,
3H; CH3CH2), 0.94 (s, 9H; tBu), 1.14 (d, J=7.1 Hz, 3H; CH3CH), 1.21–
1.38 (m, 4H; CH2CH2CH3), 1.52–1.61 (m, 2H; CH2CH2CO), 2.45 (dt, J=
17.2, 7.3 Hz, 1H; CH2CO), 2.47 (d, J=3.7 Hz, 1H; OH), 2.54 (dt, J=
17.2, 7.5 Hz, 1H; CH2CO), 2.83 (dq, J=2.9, 7.1 Hz, 1H; CHCO),
3.58 ppm (dd, J=2.9, 3.7 Hz, 1H; CHOH); 13C NMR (100 MHz, CDCl3):
d=11.5, 13.9, 22.5, 23.3, 26.7, 31.3, 35.4, 41.5, 46.5, 76.9, 216.1 ppm.
Compound 14 anti: 1H NMR (400 MHz, CDCl3): d=0.88 (s, 9H; tBu),
0.90 (t, J=6.9 Hz, 3H; CH3CH2), 1.22–1.38 (m, 4H; CH2CH2CH3), 1.29
(d, J=7.2 Hz, 3H; CH3CH), 1.47–1.64 (m, 2H; CH2CH2CO), 2.46–2.64
(m, 2H; CH2CO), 2.87 (dq, J=2.2, 7.2 Hz, 1H; CHCO), 3.22 (dd, J=2.2,
8.9 Hz, 1H; CHOH), 4.24 ppm (d, J=8.9 Hz, 1H; OH); 13C NMR
(100 MHz, CDCl3): d=13.9, 18.1, 22.4, 22.7, 26.7, 31.1, 36.1, 43.4, 43.8,
84.3, 219.1 ppm; IR (neat): n˜ =3503 (w), 1701 cmÀ1; HRMS (EI; 70 eV):
m/z calcd for C12H23O2: 199.1698 [MÀMe]+; found: 199.1691 (3 ppm); el-
2,4-Dimethyl-3-hydroxydecan-5-one (12): These aldols were obtained by
using the general procedure (method A) with [NiCl2(dppe)] (30 mg,
0.0567 mmol), a 1m solution of LiBHEt3 (57 mL, 0.0567 mmol), MgBr2
(10 mg, 0.0567 mmol), isobutyraldehyde (190 mL, 2.08 mmol), and 1a
(292 mL, 1.89 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 then 2:1 v/v) yielded a separable mixture of diastereoisomeric
aldols as
a
colorless oil (syn/anti: 66:34 by 1H NMR spectroscopy,
352.3 mg, 93% yield). The two diastereoisomers were separated by
column chromatography (pentane/Et2O, 8:1 v/v).
Compound 12 syn: 1H NMR (400 MHz, CDCl3): d=0.86 (d, J=6.8 Hz,
3H; (CH3)2CH), 0.90 (t, J=7.1 Hz, 3H; CH3CH2), 1.02 (d, J=6.5 Hz,
3H; (CH3)2CH), 1.11 (d, J=7.2 Hz, 3H; CH3CH), 1.22–1.38 (m, 4H;
CH2CH2CH3), 1.53–1.72 (m, 3H; CH2CH2CO, CH(CH3)2), 2.47 (dt, J=
3270
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3261 – 3274

Nickel Catalysts
FULL PAPER
emental analysis calcd (%) for (syn/anti mixture) C13H26O2: C 72.85, H
12.23; found: C 72.61, H 12.31.
(22 mg, 0.119 mmol), benzaldehyde (190 mL, 1.87 mmol), and 1c (295 mL,
1.70 mmol). Column chromatography on silica gel (pentane/Et2O, 8 :1
then 2:1 v/v) afforded a separable mixture of diastereoisomeric aldols as
a colorless oil (syn/anti: 55:45 by 1H NMR spectroscopy, 336.1 mg, 80%
yield). The two diastereoisomers were separated by column chromatogra-
phy (pentane/Et2O, 8:1 v/v).
Compound 18 syn: 1H NMR (400 MHz, CDCl3): d=0.83 (t, J=7.5 Hz,
3H; CH3), 0.84 (t, J=7.2 Hz, 3H; CH3), 1.07–1.16 (m, 2H; CH2), 1.17–
1.28 (m, 2H; CH2), 1.32–1.50 (m, 2H; CH2), 1.62–1.84 (m, 2H; CH2),
2.16 (ddd, J=6.5, 8.0, 17.6 Hz, 1H; CH2CO), 2.29 (ddd, J=6.5, 8.2,
17.6 Hz, 1H; CH2CO), 2.79 (d, J=2.4 Hz, 1H; OH), 2.81 (ddd, J=4.0,
6.0, 9.8 Hz, 1H; CHCO), 4.85 (dd, J=2.4, 6.0 Hz, 1H; CHOH), 7.23–
7.36 ppm (m, 5H; HAr); 13C NMR (100 MHz, CDCl3): d=12.1, 13.8, 20.5,
22.3, 22.5, 31.1, 45.1, 60.3, 74.0, 126.2, 127.6, 128.3, 142.1, 215.3 ppm.
Compound 18 anti: 1H NMR (400 MHz, CDCl3): d=0.84 (t, J=7.6 Hz,
3H; CH3), 0.86 (t, J=7.1 Hz, 3H; CH3), 1.13–1.32 (m, 4H;
CH2CH2CH3), 1.34–1.45 (m, 1H; CH2CH), 1.44–1.53 (m, 2H;
CH2CH2CO), 1.54–1.67 (m, 1H; CH2CH), 2.31 (dt, J=17.7, 7.3 Hz, 1H;
CH2CO), 2,39 (dt, J=17.7, 7.4 Hz, 1H; CH2CO), 2,84 (ddd, J=5.2, 7.1,
8.8 Hz, 1H; CHCO), 2.94 (d, J=5.7 Hz, 1H; OH), 4.80 (dd, J=5.7,
7.1 Hz, 1H; CHOH), 7.25–7.38 ppm (m, 5H; HAr); 13C NMR (100 MHz,
CDCl3): d=11.7, 13.9, 22.4, 22.5, 22.8, 31.2, 45.4, 59.8, 75.3, 126.2, 127.8,
128.4, 142.7, 216.2 ppm; HRMS (EI; 70 eV): m/z calcd for (syn/anti mix-
ture) C16H22O: 230.1671 [MÀH2O]+; found: 230.1676 (2 ppm).
1-(4-N-acetamidophenyl)-1-hydroxy-2-methyloctan-3-one (15): These
aldols were obtained by using the general procedure (method A) with
[NiCl2(dppe)] (30 mg, 0.0567 mmol), a 1m solution of LiBHEt3 (57 mL,
0.0567 mmol), and MgBr2 (10 mg, 0.0567 mmol). In this case, solid 4-N-
acetamidobenzaldehyde (340 mg, 2.08 mmol) and 1a (292 mL, 1.89 mmol)
were dissolved in anhydrous THF (4 mL) before addition to the reaction
mixture. Column chromatography on silica gel (CH2Cl2/MeOH; 98:2 v/v)
afforded an inseparable mixture of diastereoisomeric aldols as a white
1
solid (syn/anti: 67:33 by H NMR spectroscopy, 500.3 mg, 91% yield).
Compound 15 syn, anti: 1H NMR (400 MHz, CDCl3): d=0.86 (syn) (t,
J=7.4 Hz, 3H; CH3CH2), 0.87 (anti) (d, J=7.4 Hz, 3H; CH3CH), 0.88
(anti) (t, J=7.2 Hz, 3H; CH3CH2), 1.07 (syn) (d, J=7.1 Hz, 3H;
CH3CH), 1.13–1.35 (syn, anti) (m, 4H; CH2CH2CH3), 1.43–1.53 (syn) (m,
2H; CH2CH2CO), 1.51–1.60 (anti) (m, 2H; CH2CH2CO), 2.11 (syn, anti)
(s, 3H; CH3CO), 2.25–2.58 (syn, anti) (m, 2H; CH2CO), 2.81 (syn) (dq,
J=4.7, 7.1 Hz, 1H; CHCO), 2.88 (anti) (dq, J=8.5, 7.2 Hz, 1H; CHCO),
3.44 (anti) (d, J=4.2 Hz, 1H; OH), 3.62 (syn) (d, J=2.7 Hz, 1H; OH),
4.66 (anti) (dd, J=4.2, 8.5 Hz, 1H; CHOH), 4.93 (syn) (dd, J=2.7,
4.7 Hz, 1H; CHOH), 7.14–7.23 (syn, anti) (m, 2H; HAr), 7.37–7.46 (syn,
anti) (m, 2H; HAr), 8.17 (syn) (s, 1H; NH), 8.22 ppm (anti) (s, 1H; NH);
13C NMR (100 MHz, CDCl3): d=10.8 (syn), 13.8 (syn), 13.8 (anti), 14.2
(anti), 22.3 (syn), 22.4 (anti), 22.9 (syn), 22.9 (anti), 24.2 (syn), 24.2 (anti),
31.1 (syn), 31.2 (anti), 42.2 (syn), 43.0 (anti), 52.4 (syn), 52.8 (anti), 73.1
(syn), 76.1 (anti), 119.8 (syn), 119.9 (anti), 126.5 (syn), 127.0 (anti), 137.1
(syn), 137.6 (anti), 137.7 (syn), 137.9 (anti), 168.9 (anti), 169.0 (syn), 216.0
(anti), 216.1 ppm (syn); IR (neat): n˜ =3461 (w), 3250, 3190, 3121, 3067,
1699, 1661 cmÀ1; HRMS (EI; 70 eV): m/z calcd for C17H25NO3 : 291.1834
[M]+; found: 291.1829 (1 ppm); elemental analysis calcd (%) for (syn/
anti mixture) C17H25NO3: C 70.07, H 8.65, N 4.81; found: C 70.11, H 8.72,
N 4.75.
1,3-Diphenyl-3-hydroxy-2-methylpropan-1-one (19): These aldols were
obtained by using the general procedure (method A) with [NiCl2(dppe)]
(30 mg, 0.0567 mmol), a 1m solution of LiBHEt3 (57 mL, 0.0567 mmol),
MgBr2 (10 mg, 0.0567 mmol), benzaldehyde (211 mL, 2.08 mmol), and 1d
(248 mL, 1.89 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 then 2:1 v/v) afforded a separable mixture of diastereoisomeric
aldols as
a
colorless oil (syn/anti: 60:40 by 1H NMR spectroscopy,
441.2 mg, 97% yield). The two diastereoisomers were separated by
column chromatography (pentane/Et2O, 8:1 v/v).
1-(5-Acetoxymethyl-2-furyl)-1-hydroxy-2-methyloctan-3-one (16): These
aldols were obtained by using the general procedure (method A) with
[NiCl2(dppe)] (30 mg, 0.0567 mmol), a 1m solution of LiBHEt3 (57 mL,
0.0567 mmol), and MgBr2 (10 mg, 0.0567 mmol). In this case, solid 5-ace-
toxymethyl-2-furaldehyde (349 mg, 2.08 mmol) and 1a (292 mL,
1.89 mmol) were dissolved in anhydrous THF (2 mL) before addition to
the reaction mixture. Column chromatography on silica gel (toluene/
AcOEt, 20:1 then 13:1 then 5:1 v/v) afforded a separable mixture of dia-
stereoisomeric aldols as a colorless oil (syn/anti: 57:43 by 1H NMR spec-
troscopy, 525.3 mg, 94% yield). The two diastereoisomers were separated
by column chromatography (pentane/Et2O, 4:1 v/v).
Compound 16 syn: 1H NMR (400 MHz, CDCl3): d=0.89 (t, J=7.1 Hz,
3H; CH3CH2), 1.18 (d, J=7.2 Hz, 3H; CH3CH), 1.21–1.36 (m, 4H;
CH2CH2CH3), 1.55 (ddt, J=7.3, 7.4, 7.4 Hz, 2H; CH2CH2CO), 2.08 (s,
3H; CH3CO), 2.42 (dt, J=17.3, 7.3 Hz, 1H; CH2CO), 2.51 (dt, J=17.3,
7.4 Hz, 1H; CH2CO), 3.04 (dq, J=4.4, 7.2 Hz, 1H; CHCO), 3.09 (brd,
J=3.9 Hz, 1H; OH), 5.01 (s, 2H; CH2O), 5.03 (brdd, J=3.9, 4.4 Hz, 1H;
CHOH), 6.25 (d, J=3.2 Hz, 1H; Hfuryl), 6.35 ppm (d, J=3.2 Hz, 1H;
Hfuryl); 13C NMR (100 MHz, CDCl3): d=11.2, 13.9, 20.9, 22.4, 23.0, 31.3,
41.8, 49.6, 58.1, 68.3, 107.7, 111.5, 148.6, 155.3, 170.6, 215.1 ppm.
Compound 16 anti: 1H NMR (400 MHz, CDCl3): d=0.89 (t, J=7.1 Hz,
3H; CH3CH2), 1.05 (d, J=7.2 Hz, 3H; CH3CH), 1.19–1.36 (m, 4H;
CH2CH2CH3), 1.57 (ddt, J=7.2, 7.5, 7.3 Hz, 2H; CH2CH2CO), 2.07 (s,
3H; CH3CO), 2.46 (dt, J=17.3, 7.2 Hz, 1H; CH2CO), 2.55 (dt, J=17.3,
7.5 Hz, 1H; CH2CO), 3.17 (dq, J=7.3, 7.2 Hz, 1H; CHCO), 3.30 (d, J=
6.2 Hz, 1H; OH), 4.77 (dd, J=6.2, 7.3 Hz, 1H; CHOH), 5.01 (s, 2H;
CH2O), 6.25 (d, J=3.2 Hz, 1H; Hfuryl), 6.35 ppm (d, J=3.2 Hz, 1H;
Hfuryl); 13C NMR (100 MHz, CDCl3): d=13.8, 14.0, 20.8, 22.4, 22.9, 31.2,
42.8, 49.5, 58.0, 70.0, 108.3, 111.3, 149.0, 155.6, 170.6, 215.3 ppm; IR
(neat): n˜ =3465 (w), 1742, 1712 cmÀ1; HRMS (EI; 70 eV): m/z calcd for
C16H24O5: 296.1624 [M]+; found: 296.1627 (1 ppm); elemental analysis
calcd (%) for (syn/anti mixture) C16H24O5: C 64.84, H 8.16; found: C
64.89, H 8.34.
Compound 19 syn: 1H NMR (400 MHz, CDCl3): d=1.19 (d, J=7.2 Hz,
3H; CH3), 3.70 (d, J=2.1 Hz, 1H; OH), 3.70 (dq, J=3.1, 7.2 Hz, 1H;
CHCO), 5.24 (dd, J=2.1, 3.1 Hz, 1H; CHOH), 7.23–7.50 (m, 7H; HAr),
7.54–7.61 (m, 1H;
H
Ar), 7.91–8.01 ppm (m, 2H; HAr); 13C NMR
(100 MHz, CDCl3): d=11.1, 47.0, 73.0, 126.0, 127.3, 128.2, 128.4, 128.7,
133.6, 135.5, 141.7, 205.8 ppm.
Compound 19 anti: 1H NMR (400 MHz, CDCl3): d=1.06 (d, J=7.2 Hz,
3H; CH3), 3.03 (d, J=4.5 Hz, 1H; OH), 3.83 (dq, J=8.1, 7.2 Hz, 1H;
CHCO), 4.99 (dd, J=4.5, 8.1 Hz, 1H; CHOH), 7.23–7.50 (m, 7H; HAr),
7.54–7.61 (m, 1H;
H
Ar), 7.91–8.01 ppm (m, 2H; HAr); 13C NMR
(100 MHz, CDCl3): d=15.7, 47.9, 76.7, 126.7, 127.9, 128.4, 128.5, 128.6,
133.3, 136.6, 142.1, 204.9 ppm.
2,4-Dimethyl-3-hydroxy-1-phenylpentan-1-one (20): These aldols were
obtained by using the general procedure (method A) with [NiCl2(dppe)]
(30 mg, 0.0567 mmol), a 1m solution of LiBHEt3 (57 mL, 0.0567 mmol),
MgBr2 (10 mg, 0.0567 mmol), isobutyraldehyde (190 mL, 2.08 mmol), and
1d (248 mL, 1.89 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 v/v) afforded a separable mixture of diastereoisomeric aldols as
a colorless oil (50:50 by 1H NMR spectroscopy, 335.6 mg, 86% yield).
The two diastereoisomers were separated by column chromatography
(pentane/Et2O, 6:1 v/v).
Compound 20 syn: 1H NMR (400 MHz, CDCl3): d=0.96 (d, J=6.8 Hz,
3H; (CH3)2CH), 1.04 (d, J=6.6 Hz, 3H; (CH3)2CH), 1.25 (d, J=7.1 Hz,
3H; CH3CH), 1.72–1.85 (m, 1H; CH(CH3)2), 3.16 (d, J=2.8 Hz, 1H;
OH), 3.64 (ddd, J=2.8, 2.9, 8.2 Hz, 1H; CHOH), 3.68 (dq, J=2.9,
7.1 Hz, 1H; CHCO), 7.46–7.53 (m, 2H; HAr), 7.57–7.63 (m, 1H; HAr),
7.93–7.98 ppm (m, 2H; HAr); 13C NMR (100 MHz, CDCl3): d=10.7, 19.0,
19.1, 30.7, 41.7, 76.6, 128.4, 128.8, 133.4, 135.8, 205.9 ppm.
1
Compound 20 anti: Colorless solid; H NMR (400 MHz, CDCl3): d=0.94
(d, J=6.8 Hz, 3H; (CH3)2CH), 1.00 (d, J=6.7 Hz, 3H; (CH3)2CH), 1.28
(d, J=7.2 Hz, 3H; CH3CH), 1.74–1.87 (m, 1H; CH(CH3)2), 2.98 (d, J=
7.6 Hz, 1H; OH), 3.58 (ddd, J=5.7, 7.6, 7.6 Hz, 1H; CHOH), 3.71 (dq,
J=5.7, 7.2 Hz, 1H; CHCO), 7.46–7.52 (m, 2H; HAr), 7.56–7.62 (m, 1H;
3-[Hydroxy(phenyl)methyl]nonan-4-one (18): These aldols were obtained
by using the general procedure (method A) with [NiCl2(dppe)] (63 mg,
0.119 mmol), a 1m solution of LiBHEt3 (119 mL, 0.119 mmol), MgBr2
H
Ar), 7.95–8.00 ppm (m, 2H; HAr); 13C NMR (100 MHz, CDCl3): d=16.0,
Chem. Eur. J. 2006, 12, 3261 – 3274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3271

R. GrØe et al.
17.0, 20.0, 31.2, 42.4, 79.2, 128.3, 128.7, 133.4, 136.6, 206.3 ppm; HRMS
(EI; 70 eV): m/z calcd for (syn/anti mixture) C10H11O2: 163.0759
[MÀC3H7]+; found: 163.0757 (1 ppm).
1-Hydroxy-2,4-dimethyl-1-phenylpentan-3-one (21): These aldols were
obtained by using the general procedure (method A) with [NiCl2(dppe)]
(35 mg, 0.066 mmol), a 1m solution of LiBHEt3 (66 mL, 0.066 mmol),
MgBr2 (12 mg, 0.066 mmol), benzaldehyde (147 mL, 1.44 mmol), and 1e
(157 mL, 1.31 mmol). Column chromatography on silica gel (pentane/
Et2O, 8:1 then 2:1 v/v) afforded an inseparable mixture of diastereoiso-
meric aldols as colorless oil (syn/anti: 75:25 by 1H NMR spectroscopy,
182.9 mg, 68% yield).
Compound 21 (syn, anti): 1H NMR (400 MHz, CDCl3): d=0.97 (syn) (d,
J=6.9 Hz, 3H; (CH3)2CH), 0.98 (anti) (d, J=7.1 Hz, 3H; (CH3)2CH),
1.00 (anti) (d, J=7.1 Hz, 3H; (CH3)2CH), 1.05 (syn) (d, J=6.9 Hz, 3H;
(CH3)2CH), 1.07 (anti) (d, J=7.1 Hz, 3H; CH3CH), 1.10 (syn) (d, J=
7.1 Hz, 3H; CH3CH), 2.58 (syn) (sept, J=6.9 Hz, 1H; CH(CH3)2), 2.63
(anti) (sept, J=7.1 Hz, 1H; CH(CH3)2), 3.01 (syn) (dq, J=4.5, 7.1 Hz,
1H; CHCO), 3.04 (anti) (d, J=5.1 Hz, 1H; OH), 3.09 (anti) (dq, J=7.3,
7.1 Hz, 1H; CHCO), 3.23 (syn) (d, J=2.2 Hz, 1H; OH), 4.76 (anti) (dd,
J=5.1, 7.3 Hz, 1H; CHOH), 4.99 (syn) (dd, J=2.2, 4.5 Hz, 1H; CHOH),
7.23–7.38 ppm (syn, anti) (m, 5H; HAr); 13C NMR (100 MHz, CDCl3): d=
11.2 (syn), 14.9 (anti), 17.6 (anti), 17.6 (anti), 17.7 (syn), 17.9 (syn), 40.5
(syn), 41.3 (anti), 50.8 (syn), 51.0 (anti), 73.6 (syn), 76.7 (anti), 126.0 (syn),
126.4 (anti), 127.3 (syn), 127.7 (anti), 128.2 (syn), 128.3 (anti), 141.9 (syn),
142.4 (anti), 219.4 (syn), 219.5 (anti).
CDCl3): d=9.2, 11.8, 12.0, 22.1, 23.2, 23.3, 24.5, 24.6, 43.0, 49.4, 54.3, 68.4,
220.3 ppm.
Compound 23 anti: 1H NMR (400 MHz, CDCl3): d=0.84 (t, J=7.4 Hz,
3H; CH3CH2), 0.90 (t, J=7.4 Hz, 3H; CH3CH2), 0.90 (d, J=6.5 Hz, 3H;
(CH3)2CH), 0.93 (d, J=6.7 Hz, 3H; (CH3)2CH), 1.12 (d, J=7.2 Hz, 3H;
CH3CH), 1.19 (ddd, J=2.9, 9.7, 12.8 Hz, 1H; CH2CHOH), 1.35–1.49 (m,
3H; CH2CHOH, CH2CH3), 1.60–1.74 (m, 2H; CH2CH3), 1.79–1.94 (m,
1H; CH(CH3)2), 2.46 (dddd, J=5.1, 6.3, 6.3, 8.2 Hz, 1H; CH(CH2)2), 2.67
(dq, J=6.4, 7.2 Hz, 1H; CHCH3), 2.74 (d, J=6.4 Hz, 1H; OH), 3.77 ppm
(dddd, J=2.9, 6.4, 6.4, 9.4 Hz, 1H; CHOH); 13C NMR (100 MHz,
CDCl3): d=11.6, 12.0, 14.0, 21.5, 22.5, 23.8, 23.9, 24.4, 43.9, 51.3, 54.5,
71.7, 219.5 ppm; HRMS (EI; 70 eV): m/z calcd for (syn/anti mixture)
C13H24O: 196.1827 [MÀH2O]+; found: 196.1839 (5 ppm).
1-Hydroxy-2,4,4-trimethyl-1-phenylpentan-3-one (24): These aldols were
obtained by using the general procedure (method A) with [NiCl2(dppe)]
(68 mg, 0.1285 mmol), a 1m solution of LiBHEt3 (128 mL, 0.1285 mmol),
MgBr2 (24 mg, 0.1285 mmol), benzaldehyde (287 mL, 2.83 mmol), and 1g
(293 mg, 2.57 mmol). Column chromatography on silica gel (pentane/
Et2O, 8:1 then 2:1 v/v) afforded an inseparable mixture of diastereoiso-
meric aldols as colorless oil (syn/anti: 90:10 by 1H NMR spectroscopy,
502.8 mg, 89% yield).
Compound 24 syn,anti: 1H NMR (400 MHz, CDCl3): d=1.03 (anti) (d,
J=7.0 Hz, 3H; CH3CH), 1.03 (anti) (s, 9H; tBu), 1.06 (syn) (d, J=
6.9 Hz, 3H; CH3CH), 1.08 (syn) (s, 9H; tBu), 3.19 (anti) (d, J=5.8 Hz,
1H; OH), 3.25 (syn) (dq, J=4.0, 6.9 Hz, 1H; CHCH3), 3.32 (anti) (dq,
J=7.0, 7.0 Hz, 1H; CHCH3), 3.52 (syn) (d, J=1.4 Hz, 1H; OH), 4.77
(anti) (dd, J=5.8, 7.0 Hz, 1H; CHOH), 4.89 (syn) (dd, J=1.4, 4.0 Hz,
4-Ethyl-1-hydroxy-2-methyl-1-phenylhexan-3-one (22): These aldols were
obtained by using the general procedure (method A) with [NiCl2(dppe)]
(60 mg, 0.114 mmol), a 1m solution of LiBHEt3 (114 mL, 0.114 mmol),
MgBr2 (21 mg, 0.114 mmol), benzaldehyde (255 mL, 2.50 mmol), and 1 f
(344 mL, 2.28 mmol). Column chromatography on silica gel (pentane/
Et2O, 8:1 then 2:1 v/v) afforded a separable mixture of diastereoisomeric
1H; CHOH), 7.22–7.36 ppm (syn, anti) (m, 5H;
H
Ar); 13C NMR
(100 MHz, CDCl3): d=12.0 (syn), 16.6 (anti), 25.8 (syn), 25.9 (anti), 45.0
(anti), 45.1 (syn), 46.2 (syn), 46.7 (anti), 73.9 (syn), 77.4 (anti), 126.1 (syn),
126.3 (anti), 127.4 (syn), 127.7 (anti), 128.2 (syn), 128.3 (anti), 141.8 (syn),
143.0 (anti), 221.1 (anti), 221.4 ppm (syn); HRMS (EI; 70 eV): m/z calcd
for (syn/anti mixture) C14H20O2: 220.1463 [M]+; found: 220.1477 (6 ppm).
aldols as
a
colorless oil (syn/anti 70:30 by 1H NMR spectroscopy,
477.8 mg, 90% yield). The two diastereoisomers were separated by
column chromatography (pentane/Et2O, 6:1 v/v).
Ethyl 4-[hydroxy(phenyl)methyl]-2,2-dimethyl-3-oxopentanoate (25):
These aldols were obtained by using the general procedure (method A)
with [NiCl2(dppe)] (43.6 mg, 0.0825 mmol), a 1m solution of LiBHEt3
(82 mL, 0.0825 mmol), MgBr2 (15.2 mg, 0.0825 mmol), benzaldehyde
(184 mL, 1.81 mmol), and 1h (293 mL, 1.65 mmol). Column chromatogra-
phy on silica gel (pentane/Et2O, 8:1 then 2:1 v/v) afforded the diastereo-
Compound 22 syn: 1H NMR (400 MHz, CDCl3): d=0.80 (t, J=7.4 Hz,
3H; CH3CH2), 0.84 (t, J=7.4 Hz, 3H; CH3CH2), 1.05 (d, J=7.2 Hz, 3H;
CH3CH), 1.34–1.49 (m, 2H; CH2), 1.50–1.62 (m, 1H; CH2), 1.61–1.73 (m,
1H; CH2), 2.39–2.47 (m, 1H; CHCH2), 2.90 (dq, J=3.6, 7.2 Hz, 1H;
CHCH3), 3.40 (d, J=2.0 Hz, 1H; OH), 5.05 (dd, J=2.0, 3.6 Hz, 1H;
CHOH), 7.23–7.37 ppm (m, 5H; HAr); 13C NMR (100 MHz, CDCl3): d=
9.8, 11.6, 12.0, 23.0, 24.2, 51.9, 54.5, 72.8, 126.0, 127.3, 128.2, 141.7,
219.6 ppm.
Compound 22 anti: 1H NMR (400 MHz, CDCl3): d=0,80 (t, J=7.4 Hz,
3H; CH3CH2), 0.83 (t, J=7.4 Hz, 3H; CH3CH2), 0.97 (d, J=7.3 Hz, 3H;
CH3CH), 1.28–1.48 (m, 2H; CH2), 1.52–1.74 (m, 2H; CH2), 2.37–2.46 (m,
1H; CHCH2), 3.00 (dq, J=7.8, 7.3 Hz, 1H; CHCH3), 3.07 (d, J=5.0 Hz,
1H; OH), 4.79 (dd, J=5.0, 7.8 Hz, 1H; CHOH), 7.25–7.37 ppm (m, 5H;
HAr); 13C NMR (100 MHz, CDCl3): d=11.3, 11.8, 14.4, 22.3, 23.3, 52.4,
54.9, 76.4, 126.5, 127.7, 128.3, 142.3, 218.6 ppm; HRMS (EI; 70 eV): m/z
calcd for (syn/anti mixture) C15H22O2: 234.1620 [M]+; found: 234.1624
(1 ppm).
1
isomeric aldols as colorless oil (syn/anti >95:5 by H NMR spectroscopy,
427.2 mg, 93% yield).
Compound 25 syn: 1H NMR (400 MHz, CDCl3): d=1.05 (d, J=6.9 Hz,
3H; CH3CH), 1,28 (t, J=7.1 Hz, 3H; CH3CH2), 1.31 (s, 3H; CH3), 1.35
(s, 3H; CH3), 3.07 (dq, J=3.2, 6.9 Hz, 1H; CHCH3), 3.42 (d, J=1.4 Hz,
1H; OH), 4.23 (dd, J=10.9, 7.1 Hz, 1H; CH2O), 4.26 (dd, J=10.9,
7.1 Hz, 1H; CH2O), 4.97–5.02 (m, 1H; CHOH), 7.23–7.38 ppm (m, 5H;
H
Ar); 13C NMR (100 MHz, CDCl3): d=11.5, 14.0, 21.6, 21.6, 48.4, 56.4,
61.6, 73.4, 125.9, 127.4, 128.2, 141.2, 173.1, 214.2; HRMS (EI; 70 eV): m/z
calcd for C9H16O3: 172.1100 [MÀC6H5CHO]+; found: 172.1100 (0 ppm).
Ethyl 5-hydroxy-2,2,4,7-tetramethyl-3-oxooctanoate (26): These aldols
were obtained by using the general procedure (method A) with [NiCl2-
3-Ethyl-6-hydroxy-5,8-dimethylnonan-4-one (23): These aldols were ob-
tained by using the general procedure (method A) with [NiCl2(dppe)]
(66.5 mg, 0,126 mmol), a 1m solution of LiBHEt3 (126 mL, 0.126 mmol),
MgBr2 (23.3 mg, 0.126 mmol), isovaleraldehyde (297 mL, 2.77 mmol), and
1 f (380 mL, 2.52 mmol). Column chromatography on silica gel (pentane/
Et2O, 6:1 v/v) afforded a separable mixture of diastereoisomeric aldols as
a colorless oil (syn/anti: 69:31 by 1H NMR spectroscopy, 473.7 mg, 88%
yield). The two diastereoisomers were separated by column chromatogra-
phy (pentane/Et2O, 6:1 v/v).
Compound 23 syn: 1H NMR (400 MHz, CDCl3): d=0.85 (t, J=7.4 Hz,
3H; CH3CH2), 0.89 (t, J=7.5 Hz, 3H; CH3CH2), 0.92 (d, J=6.6 Hz, 3H;
(CH3)2CH), 0.93 (d, J=6.7 Hz, 3H; (CH3)2CH), 1.08 (dddd, J=1.1, 4.4,
8.4, 13.9 Hz, 1H; CH2CHOH), 1.11 (d, J=7.2 Hz, 3H; CHCH3), 1.37–
1.53 (m, 3H; CH2CHOH, CH2CH3), 1.57–1.72 (m, 2H; CH2CH3), 1.71–
1.82 (m, 1H; CH(CH3)2), 2.45–2.54 (m, CH(CH2)2), 2.59 (dq, J=2.5,
7.2 Hz, 1H; CHCH3), 2.99 (dd, J=1.1, 2.5 Hz, 1H; OH), 3.99 ppm
(dddd, J=2.5, 2.5, 4.4, 4.7 Hz, 1H; CHOH); 13C NMR (100 MHz,
(dppe)] (32.1 mg, 0.0608 mmol),
a 1m solution of LiBHEt3 (61 mL,
0.0608 mmol), MgBr2 (11.2 mg, 0.0608 mmol), isovaleraldehyde (239 mL,
2.23 mmol), and 1h (358 mL, 2.03 mmol). Column chromatography on
silica gel (pentane/Et2O, 6:1 then 2:1 v/v) afforded the aldols as colorless
1
oil (syn/anti>95:5 by H NMR spectroscopy, 460.9 mg, 88% yield).
Compound 26 syn: 1H NMR (400 MHz, CDCl3): d=0.90 (d, J=6.6 Hz,
3H; (CH3)2CH), 0.92 (d, J=6.7 Hz, 3H; (CH3)2CH), 1.03 (dddd, J=1.3,
4.1, 8.6, 13.8 Hz, 1H; CH2CHOH), 1.12 (d, J=7.0 Hz, 3H; CH3CH), 1.27
(t, J=7.2 Hz, 3H; CH3CH2), 1.38 (s, 3H; CH3), 1.41 (s, 3H; CH3), 1.49
(ddd, J=5.5, 9.1, 13.8 Hz, 1H; CH2CHOH), 1.69–1.81 (m, 1H; CH-
(CH3)2), 2.77 (dq, J=2.1, 7.0 Hz, 1H; CHCO), 2.97 (dd, Jꢀ1.5, 1.5 Hz,
1H; OH), 3.87 (dddd, J=1.5, 2.1, 4.1, 9.1 Hz, 1H; CHOH), 4.20 (dd, J=
10.9, 7.2 Hz, 1H; CH2O), 4.25 ppm (dd, J=10.9, 7.2 Hz, 1H; CH2O);
13C NMR (100 MHz, CDCl3): d=11.1, 14.0, 21.8, 21.8, 22.0, 23.3, 24.5,
43.0, 45.9, 56.3, 61.5, 69.2, 173.1, 214.7; HRMS (EI; 70 eV): m/z calcd for
C9H16O3: 172.1100 [MÀC4H9CHO]+; found: 172.1097 (1 ppm).
3272
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3261 – 3274

Nickel Catalysts
FULL PAPER
(4S)-4-(Benzyloxy)-3-hydroxy-2-methyl-1-phenylpentan-1-one (28a–d):
These aldols were obtained in 79% overall yield from 1d and aldehyde
27,[25] by using the general procedure (method A). Analysis of the crude
reaction mixture by 1H and 13C NMR spectroscopy, and by comparison
with literature data,[26] indicated the following ratios for the aldol prod-
ucts: 28a (48%), 28b (12%), 28c (32%), 28d (8%).
8.20–8.27 (m, 2H; HAr), 8.30–8.37 ppm (m, 2H; HAr); 31P{1H} NMR
(162 MHz, CD3COCD3): d=45.9 (d, J=78.5 Hz), 64.0 ppm (d, J=
78.5 Hz).
1
Compound 33: H NMR (400 MHz, CDCl3): d=0.91–0.97 (m, 6H; CH3),
2.11–2.16 (m, 2H; CHCH3), 7.46–7.68 (m, 12H; HAr), 7.72–7.78 (m, 4H;
H
Ar), 8.32–8.38 ppm (m, 4H; HAr); 31P {1H} NMR (162 MHz, CDCl3): d=
59.7 ppm (s).
(6S)-Ethyl 6-(benzyloxy)-5-hydroxy-2,2,4-trimethyl-3-oxoheptanoate (30):
These aldols were obtained by using the general procedure (method A)
with [NiCl2(dppe)] (46.5 mg, 0.088 mmol), a 1m solution of LiBHEt3
(88 mL, 0.088 mmol), MgBr2 (16.2 mg, 0.088 mmol), 27 (317 mL,
1.94 mmol, 1.1 equiv), and the allylic alcohol 1h (312 mL, 1.76 mmol).
Column chromatography on silica gel (hexane/Et2O, 7:1 then 4:1 then 2:1
v/v) afforded a separable mixture of diastereoisomeric aldols as a color-
less oil (syn/anti: 78:22 by 1H NMR spectroscopy, 561.8 mg, 95% yield).
The two diastereoisomers were separated by column chromatography
(hexane/Et2O 6:1 v/v).
Complex 34 was prepared following the method of Frauenrath.[27]
Starting from each of these three complexes, the experiments were per-
formed by using the previously described procedure (method A) with al-
lylic alcohol 1a and benzaldehyde. After separation by chromatography,
the aldols 7 syn and 7 anti were studied by 1H NMR spectroscopy with
[Eu(tfc)3]. In the presence of this chiral shift reagent, the racemic mix-
ture (Æ)-7 syn, prepared from the dppe complex, exhibited a nice split-
ting of the signal of the carbinol proton CHOH at d=5.49 and 5.56 ppm.
Similarly, the (Æ)-7 anti aldol exhibited a significant splitting of the
CHCH3 signals at d=1.06 and 1.07 ppm. For each of the experiments
performed with 32, 33, and 34, the NMR spectra obtained for corre-
sponding syn and anti aldols were identical to those obtained using the
racemic catalyst, with equal intensities for the two signals. The same ex-
periments (method A) were performed starting from allylic alcohol 1h
and benzaldehyde. The NMR spectra, in the presence of [Eu(tfc)3], were
again identical to those obtained with the racemic aldol (Æ)-25 syn, with
a significant splitting of the signals corresponding to the carbinol proton
CHOH at d=5.43 and 5.49 ppm.
1
Compound 30a: H NMR (400 MHz, CDCl3): d=0.96 (d, J=6.9 Hz, 3H;
CH3CH), 1.22 (t, J=7.2 Hz, 3H; CH3CH2), 1.27 (d, J=6.0 Hz, 3H;
CH3CH), 1.37 (s, 3H; CH3), 1.39 (s, 3H; CH3), 3.07 (d, J=1.6 Hz, 1H;
OH), 3.28 (dq, J=2.2, 6.9 Hz, 1H; CHCO), 3.38 (dq, J=8.0, 6.0 Hz, 1H;
CHO), 3.64 (ddd, J=1.6, 2.2, 8.0 Hz, 1H; CHOH), 4.08–4.18 (m, 2H;
CH2CH3), 4.38 (d, J=11.8 Hz, 1H; CH2Ph), 4.63 (d, J=11.8 Hz, 1H;
CH2Ph), 7.25–7.39 ppm (m, 5H; HAr); 13C NMR (100 MHz, CDCl3): d=
11.0, 13.9, 16.0, 21.8, 21.9, 41.7, 56.3, 61.5, 70.4, 73.6, 74.4, 127.6, 127.8,
128.3, 138.3, 173.1, 215.1 ppm.
1
Compound 30b: H NMR (400 MHz, CDCl3): d=1.17 (d, J=6.9 Hz, 3H;
1-Phenylprop-2-en-1-one (35): IBX (10.92 g, 39 mmol, 1.5 equiv) was
added portionwise to a solution of 1d (3.50 g, 26 mmol) in DMSO
(25 mL) at room temperature. The reaction mixture was stirred for 1.5 h,
then water (15 mL) and CH2Cl2 (15 mL) were added. After filtration on
Celite, the aqueous phase was extracted with CH2Cl2 (3”30 mL). The or-
ganic phases were washed with water (2”15 mL), dried (MgSO4), and
concentrated under vacuum. Purification by column chromatography on
silica gel (pentane/Et2O, 20:1 v/v) afforded the enone 35 as a pale yellow
oil (2.75 g, 80% yield). 1H NMR (400 MHz, CDCl3): d=5.92 (dd, J=1.7,
10.6 Hz, 1H; CH2 =CH), 6.44 (dd, J=1.7, 17.1 Hz, 1H; CH2 =CH), 7.16
(dd, J=10.6, 17.1 Hz, 1H; CH2 =CH), 7.43–7.51 (m, 2H; HAr), 7.54–7.60
(m, 1H; HAr), 7.91–7.98 ppm (m, 2H; HAr); 13C NMR (100 MHz, CDCl3):
d=128.5, 128.6, 130.1, 132.2, 132.9, 137.1, 190.9.
CH3CH), 1.24 (d, J=6.2 Hz, 3H; CH3CH), 1.25 (t, J=7.1 Hz, 3H;
CH3CH2), 1.37 (s, 3H; CH3), 1.38 (s, 3H; CH3), 2.59 (d, J=6.1 Hz, 1H;
OH), 3.12 (dq, J=5.5, 6.9 Hz, 1H; CHCO), 3.50 (dq, J=4.4, 6.2 Hz, 1H;
CHO), 3.68 (ddd, J=4.4, 5.5, 6.1 Hz, 1H; CHOH), 4.16 (q, J=7.1 Hz,
2H; CH2CH3), 4.45 (d, J=11.5 Hz, 1H; CH2Ph), 4.63 (d, J=11.5 Hz,
1H; CH2Ph), 7.26–7.38 ppm (m, 5H; HAr); 13C NMR (100 MHz, CDCl3):
d=13.6, 14.0, 16.2, 22.2, 22.2, 44.1, 56.2, 61.4, 70.8, 75.5, 75.6, 127.6, 128.4,
138.3, 173.3, 212.2; HRMS (EI; 70 eV): m/z calcd for C10H17O4: 201.1127
[MÀC9H11O]+; found: 201.1109 (8 ppm).
(6S)-Ethyl 6-(tert-butyldimethylsiloxy)-5-hydroxy-2,2,4-trimethyl-3-oxo-
heptanoate (31): These aldols were obtained by using the general proce-
dure (method A) with [NiCl2(dppe)] (55 mg, 0.104 mmol), a 1m solution
of LiBHEt3 (104 mL, 0.104 mmol), MgBr2 (19.2 mg, 0.104 mmol), 29
(495 mL, 2.29 mmol, 1.1 equiv), and allylic alcohol 1h (369 mL,
2.08 mmol). Column chromatography on silica gel (hexane/Et2O, 12:1
then 8:1 then 4:1 v/v) afforded a separable mixture of diastereoisomeric
1-Deuterio-1-phenylprop-2-en-1-ol (36): CeCl3·7H2O (7.45 g, 20 mmol,
1 equiv) was added to a solution of 35 (2.68 g, 20 mmol) in methanol
(30 mL) at room temperature and under stirring. The reaction mixture
was cooled to 08C and then NaBD4 (0.85 g, 20 mmol, 1 equiv) was added
portionwise. The reaction mixture was stirred at room temperature for
25 min before the addition of a saturated NH4Cl solution (15 mL). The
aqueous phase was extracted with ether (3”30 mL), and the organic
phases were dried (MgSO4) and concentrated under vacuum. Purification
by column chromatography on silica gel (pentane/Et2O, 4:1 v/v) afforded
the alcohol 36 as a colorless oil (2.12 g, 77% yield). 1H NMR (400 MHz,
CDCl3): d=2.26 (s, 1H; OH), 5.17 (dd, J=1.4, 10.3 Hz, 1H; CH2=CH),
5.32 (dd, J=1.4, 17.1 Hz, 1H; CH2=CH), 6.02 (dd, J=10.3, 17.1 Hz, 1H;
CH2=CH), 7.25–7.36 ppm (m, 5H; HAr); 13C NMR (100 MHz, CDCl3):
d=74.8 (t, J=43.9 Hz), 115.1, 126.3, 127.7, 128.5, 140.1, 142.4 ppm;
HRMS (EI; 70 eV): m/z calcd for C9H9DO: 135.0794 [M]+; found:
135.0791 (2 ppm).
aldols as
a
colorless oil (syn/anti: 83:17 by 1H NMR spectroscopy,
632.6 mg, 85% yield). The two diastereoisomers 31a and 31b were sepa-
rated by column chromatography (hexane/Et2O 12:1 v/v).
Compound 31a: 1H NMR (400 MHz, CDCl3): d=0.05 (s, 3H; CH3Si),
0.06 (s, 3H; CH3Si), 0.87 (s, 9H; tBu), 1.07 (d, J=7.0 Hz, 3H; CH3CH),
1.19 (d, J=5.9 Hz, 3H; CH3CH), 1.26 (t, J=7.1 Hz, 3H; CH3CH2), 1.37
(s, 3H; CH3), 1.38 (s, 3H; CH3), 3.09 (d, J=1.3 Hz, 1H; OH), 3.29 (dq,
J=2.0, 7.0 Hz, 1H; CHCO), 3.46 (ddd, J=1.3, 2.0, 8.0 Hz, 1H; CHOH),
3.62 (dq, J=8.0, 5.9 Hz, 1H; CHO), 4.16 ppm (q, J=7.1 Hz, 2H;
CH2CH3); 13C NMR (100 MHz, CDCl3): d=À5.0, À4.0, 10.8, 14.0, 17.8,
20.7, 21.8, 21.9, 25.7, 41.4, 56.3, 61.4, 67.9, 76.0, 172.9, 215.1 ppm.
Compound 31b: 1H NMR (400 MHz, CDCl3): d=0.09 (s, 3H; CH3Si),
0.10 (s, 3H; CH3Si), 0.90 (s, 9H; tBu), 1.14 (d, J=6.9 Hz, 3H; CH3CH),
1.17 (d, J=6.2 Hz, 3H; CH3CH), 1.27 (t, J=7.1 Hz, 3H; CH3CH2), 1.39
(s, 3H; CH3), 1.41 (s, 3H; CH3), 2.63 (d, J=5.3 Hz, 1H; OH), 3.02 (dq,
J=4.9, 6.9 Hz, 1H; CHCO), 3.52–3.58 (m, 1H; CHOH), 3.72 (dq, J=4.8,
6.2 Hz, 1H; CHO), 4.18 ppm (q, J=7.1 Hz, 2H; CH2CH3); 13C NMR
(100 MHz, CDCl3): d=À4.7, À4.0, 12.9, 14.0, 18.1, 20.7, 22.3, 22.4, 25.8,
44.0, 56.1, 61.4, 70.0, 75.4, 173.4, 211.8 ppm; HRMS (EI; 70 eV): m/z
calcd for C14H27O5Si: 303.1628 [MÀtBu]+; found: 303.1634 (2 ppm).
The reaction of 36 with benzaldehyde and the nickel catalyst was per-
formed as described earlier for the synthesis of aldols 19. By suing EI
HRMS, the isotopic pattern was analyzed for the [MÀC6H5CHO]+ ions
and gave the three species: [D0] (38%), [D1] (44%), and [D2] (18%).
The ratio between the two [D1] compounds was determined by compari-
son of the integration of the 1H NMR signals of the CHOH signals (d=
5.24 and 5.02 ppm) with both the signals of the CHCH3 and the (CH2D+
CH3) groups. From the integration of the CHCH3 signals (at d=3.73,
3.48, 2.88 ppm), values 29% for the [D1](CH2D) compound and 15% for
the [D1](CDCH3) compound were obtained. From the integration of the
CH3 and CH2D signals (at d=1.17 and 0.95 ppm) values of 20 and 24%
were obtained, respectively. So, the range was evaluated as 20–30% for
the [D1](CH2D) compound and 14–24% for the [D1](CDCH3) com-
pound. These results were confirmed by the analysis of the methyl
Catalytic studies using chiral nickel complexes: The complexes 32 and 33
were prepared following a literature method.[27]
1
Compound 32: H NMR (400 MHz, CDCl3): d=1.03 (dd, J=6.9, 11.9 Hz,
3H; CH3), 1.90–2.02 (m, 1H; CH2), 2.12–2.32 (m, 1H; CH2), 2.44–2.56
(m, 1H; CHCH3), 7.43–7.66 (m, 12H; HAr), 7.71–7.85 (m, 4H; HAr),
Chem. Eur. J. 2006, 12, 3261 – 3274
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3273

R. GrØe et al.
groups signals in the 13C{1H} NMR spectra of these aldols. The anti ad-
ducts exhibited the following signals: d=15.92 (s, CH3 for 19), 15.82 (s,
Maurer, G. Raabe, Angew. Chem. 2001, 113, 176–178; Angew.
Chem. Int. Ed. 2001, 40, 177–179.
CH3 for 38), 15.66 (t, JCD =20 Hz, CH3 for 37 or 39), 15.56 ppm (t, JCD
=
[15] For a preliminary communication on this research see: D. Cuperly,
C. CrØvisy, R. GrØe, Synlett 2004, 1, 93–96.
20 Hz, CH3 for 39 or 37). The syn adducts exhibit the following signals:
d=12.18 (s, CH3 for 19), 12.11 (s, CH3 for 38), 11.92 (t, JCD =20 Hz, CH3
for 37 or 39), 11.84 ppm (t, JCD =20 Hz, CH3 for 39 or 37).
[16] a) V. Branchadell, C. CrØvisy, R. GrØe, Chem. Eur. J. 2004, 10,
5795–5803; see also: b) T. J. Dickerson, T. Lovell, M. M. Meijler, L.
Noodleman, K. D. Janda, J. Org. Chem. 2004, 69, 6603–6609; c) X.
Zhang, K. N. Houk, J. Org. Chem. 2005, 70, 9712–9716.
[17] a) M. L. H. Green, T. Saito, J. Chem. Soc. D 1969, 208; b) M. L. H.
Green, T. Saito, P. J. Tanfield, J. Chem. Soc. A 1971, 152–154.
[18] C. Malanga, A. Urso, L. Lardicci, Tetrahedron Lett. 1995, 36, 1133–
1136.
[19] T. Fujisawara, Y. Sasaki, T. Mukaiyama, Chem. Lett. 2001, 190–191.
[20] B. Martin-Matute, M. Edin, K. Bogar, F. B. Kaynak, J.-E. Bꢂckvall,
J. Am. Chem. Soc. 2005, 127, 8817–8825, and references therein.
[21] I. Kuwajima, M. Kato, A. Mori, Tetrahedron Lett. 1980, 21, 4291–
4294.
Acknowledgements
We thankvery much Pr. Vicenc Branchadell (Universitat Autonoma de
Barcelona, Spain) for particularly fruitful discussions. D.C. and J.P. thank
the MENRT for PhD thesis fellowships. We thankvery much P. Guenot
(CRMPO Rennes) for his help during the mass spectral analysis.
À
[1] Comprehensive Organic Chemistry, Addition to C X p Bonds, Part 2
[22] a) C. H. Heatchcock, S. K. Davidsen, K. T. Hug, L. A. Filippin, J.
Org. Chem. 1986, 51, 3027–3037; b) S. E. Denmark, R. A. Staveng-
er, K.-T. Wong, X. Su, J. Am. Chem. Soc. 1999, 121, 4982–4991.
[23] I. Shibata, T. Suwa, H. Sakakibara, A. Baba, Org. Lett. 2002, 4, 301–
303.
(Ed.: C. H. Heathcock), Pergamon Press, 1991, 99–319.
[2] a) B. M. Trost, Science 1991, 254, 1471–1477; b) B. M. Trost, Angew.
Chem. 1995, 107, 287–307; Angew. Chem. Int. Ed. Engl. 1995, 34,
259–281.
[3] T. Hoffmann, G. Zhong, B. List, D. Shabat, J. Anderson, S. Gramati-
kova, R. A. Lerner, C. F. Barbas III, J. Am. Chem. Soc. 1998, 120,
2768–2779.
[4] a) B. List, R. A. Lerner, C. F. Barbas III, J. Am. Chem. Soc. 2000,
122, 2395–2396; b) W. Notz, B. List, J. Am. Chem. Soc. 2000, 122,
7386–7387; c) B. List, P. Pojarliev, C. Castello, Org. Lett. 2001, 3,
573–575; d) B. List, Acc. Chem. Res. 2004, 37, 548–557; e) Z. Tang,
Z.-H. Yang, L.-F. Cun, L.-Z. Gong, A.-Q. Mi, Y.-Z. Jiang, Org. Lett.
2004, 6, 2285–2287; f) W. Notz, F. Tanaka, C. F. Barbas III, Acc.
Chem. Res. 2004, 37, 580–591, and references therein.
[24] A. SolladiØ-Cavallo, F. Bonne, Tetrahedron: Asymmetry 1996, 7,
171–180.
[25] a) M. T. Reetz, K. Kesseler, A. Jung, Tetrahedron 1984, 40, 4327–
4336; b) C. H. Heathcock, S. K. Davidsen, K. T. Hug, L. A. Flippin,
J. Org. Chem. 1986, 51, 3027–3037.
[26] M. Hirama, T. Shigemoto, S. Ito, J. Org. Chem. 1987, 52, 3342–3346.
[27] F. Morandini, G. Consiglio, O. Piccolo, Inorg. Chim. Acta 1982, 57,
15–19.
[28] E. R. Burkhardt, R. G. Bergman, C. H. Heathcock, Organometallics
1990, 9, 30–44.
[5] a) N. Yoshikawa, Y. M. A. Yamada, J. Das, H. Sasai, M. Shibasaki, J.
Am. Chem. Soc. 1999, 121, 4168–4178; b) B. M. Trost, H. Ito, J. Am.
Chem. Soc. 2000, 122, 12003–12004.
[6] a) R. Uma, C. CrØvisy, R. GrØe, Chem. Rev. 2003, 103, 27–51, and
references therein; b) L. A. Yanovskaya, Kh. Shakhidayatov, Russ.
Chem. Rev. 1970, 39, 859–874; c) R. C. van der Drift, E. Bouwman,
E. Drent, J. Organomet. Chem. 2002, 650, 1–24.
[29] a) B. M. Trost, R. J. Kulawiec, J. Am. Chem. Soc. 1993, 115, 2027–
2036; b) I. E. Marko, A. Gautier, M. Tsukazaki, A. Llobet, E. Plan-
talech-Mir, C. J. Urch, S. M. Brown, Angew. Chem. 1999, 111, 2126–
2128; Angew. Chem. Int. Ed. 1999, 38, 1960–1962; for heterogene-
ous catalysts, see also: c) K. Yamaguchi, T. Koike, M. Kotani, M.
Matsushita, S. Shinachi, N. Mizuno, Chem. Eur. J. 2005, 11, 6574–
6582.
[7] M. Dedieu, Y.-L. Pascal, Bull. Soc. Chim. Fr. 1978, 11/12, 593–594.
[8] S. H. Bergens, B. Bosnich, J. Am. Chem. Soc. 1991, 113, 958–967.
[9] a) G. L. Edwards, W. B. Motherwell, D. M. Powell, D. A. Sandham,
J. Chem. Soc. Chem. Commun. 1991, 1399–1401; b) W. B. Mother-
well, D. A. Sandham, Tetrahedron Lett. 1992, 33, 6187–6190; c) L. J.
Gazzard, W. B. Motherwell, D. A. Sandham, J. Chem. Soc. Perkin
Trans. 1 1999, 979–994.
[10] a) C. CrØvisy, M. Wietrich, V. Le Boulaire, R. Uma, R. GrØe, Tetra-
hedron Lett. 2001, 42, 395–398; b) R. Uma, N. Gouault, C. CrØvisy,
R. GrØe, Tetrahedron Lett. 2003, 44, 6187–6190.
[11] a) R. Uma, M. Davies, C. CrØvisy, R. GrØe, Tetrahedron Lett. 2001,
42, 3069–3072; see also b) M. Wang, C.-J. Li, Eur. J. Org. Chem.
2003, 998–1003; c) M. Wang, C.-J. Li, Org. Lett. 2003, 5, 657–660;
d) B. Martin-Matute, K. Bogar, M. Edin, F. B. Kaynak, J.-E. Bꢂck-
vall, Chem. Eur. J. 2005, 11, 5832–5842.
[12] A similar reaction has been demonstrated also, starting from propar-
gylic alcohols, see: B. M. Trost, S. Oi, J. Am. Chem. Soc. 2001, 123,
1230–1231.
[30] The same result (recovery of the starting materials) was also ob-
tained when this reaction was performed in the presence of the allyl-
ic alcohol. This strongly suggests that, at the enol level, the competi-
tion between the two reaction pathways (isomerization or aldoliza-
tion) is very sensitive to the reaction conditions. The transition-
metal-mediated reaction involves a slow generation of the reactive
intermediates, in the presence of a very large excess of the allylic al-
cohol and the aldehydes, and this is probably an important issue for
the success of this tandem reaction.
[31] a) C. H. Heathcock, S. Arseniyadis, Tetrahedron Lett. 1985, 26,
6009–6012; b) K. A. Swiss, W.-B. Choi, D. C. Liotta, A. F. Abdel-
Magid, C. A. Maryanoff, J. Org. Chem. 1991, 56, 5978–5980.
[32] a) B. W. Gung, X. Xue, Tetrahedron: Asymmetry 2001, 12, 2955–
2959; b) D. A. Evans, S. J. Siska, V. J. Cee, Angew. Chem. 2003, 115,
1803–1807; Angew. Chem. Int. Ed. 2003, 42, 1761–1765.
[33] M. M. Midland, R. W. Koops, J. Org. Chem. 1990, 55, 5058–5065.
[34] J-L. Luche, J.-C. Damiano, J. Am. Chem. Soc. 1980, 102, 7926–7927.
[35] A. B. Smith III, P. A. Levenberg, Synthesis 1981, 567–570.
[13] a) C. F. Lochow, R. G. Miller, J. Org. Chem. 1976, 41, 3020–3022;
b) H. Bricout, E. Monflier, J.-F. Carpentier, A. Mortreux, Eur. J.
Inorg. Chem. 1998, 1739–1744.
[14] a) H. Frauenrath, S. Reim, A. Wiesner, Tetrahedron: Asymmetry
1998, 9, 1103–1106; b) H. Frauenrath, D. Brethauer, S. Reim, M.
Received: December 13, 2005
Published online: February 28, 2006
3274
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3261 – 3274