From Allylic Alcohols to Aldols via a Novel, Efficient Reaction Catalyzed by Ni-Complexes

Article abstract of DOI:10.1055/s-2003-43373

Allylic alcohols react with aldehydes, in an atom economic aldol-type reaction, in the presence of a catalytic amount of (dppe)NiHCl/MgBr₂ mixture. This reaction occurs in good to excellent yields with total regiocontrol and very high chemo control under mild conditions. It is compatible with various types of aldehydes including very bulky ones.

Full text of DOI:10.1055/s-2003-43373

LETTER
93
From Allylic Alcohols to Aldols via a Novel, Efficient Reaction Catalyzed by
Ni-Complexes
From
Allylic
Alco
a
hols to
A
ldo
v
ls id Cuperly, Christophe Crévisy, René Grée*
ENSCR, Laboratoire de Synthèses et Activations de Biomolécules, UMR CNRS 6052, 35700 Rennes, France
Fax +33(2)23238108; E-mail: rene.gree@ensc-rennes.fr
Received 10 October 2003
the regio- and the chemoselectivities, however with a
modest diastereoselectivity.
Abstract: Allylic alcohols react with aldehydes, in an atom eco-
nomic aldol-type reaction, in the presence of a catalytic amount of
(dppe)NiHCl/MgBr2 mixture. This reaction occurs in good to excel-
lent yields with total regiocontrol and very high chemo control un-
der mild conditions. It is compatible with various types of aldehydes
including very bulky ones.
It has been demonstrated previously that using appropri-
ate Ni or Rh based catalysts, allylic alcoholates have been
isomerized to the corresponding enolates, followed by al-
dol reactions.7 However, this original stepwise procedure
is less efficient in term of atom economy. Nickel enolates
have been isolated and used in various reactions, includ-
ing aldol condensations.8 Recently, some nickel hydride
derivatives {(e.g. NiI2[(–)-Me-DUPHOS]/LiBHEt3} have
also been reported to be excellent catalysts for the asym-
metric isomerization of cyclic allyl acetals.9 So, it was an-
ticipated that such Ni-H complexes would be able to
catalyze also the isomerization of allylic alcohols. How-
ever, even if the isomerization step could be expected to
proceed, the possibility for the intermediate to be trapped
by a carbonyl derivative was more challenging. There-
fore, we have tested three different nickel hydride cata-
lysts 1–3 (Figure 1) to perform our tandem allylic alcohol
isomerization-aldol condensation reaction using octen-3-
ol 4 and isovaleraldehyde as models.
Key words: allylic alcohols, double bond isomerization, aldol reac-
tion, nickel hydrides
The aldol reaction is a fundamental process for the cre-
ation of C-C bonds.1 We have recently developed a novel
tandem isomerization–aldol condensation mediated by
transition metal catalysts.2 Following a known process,3
an allylic alcohol is converted by the catalyst to a com-
plexed enol and/or a transition metal enolate that is
trapped in situ by an aldehyde to give an aldol product
(Scheme 1).4
R'CHO
R
Cat.
R
R'
R
Cat
OH
OH
Scheme 1
OH
O
Fe
Ph
Ph
Ph
Ph
P
P
Ni
Ph
Ph
Ph
Ph
P
H
P
Cl
H
Ni
The reaction has first been performed using iron catalysts
such as Fe(CO)5, (bda)Fe(CO)3 or (COT)Fe(CO)3.2a–c
Then, it has been extended to Ru and Rh catalysts2d,5 [e.g.
(PPh3)3RhH or (PPh3)3RuHCl]. It must be underlined that,
in all cases, these reactions occur under mild conditions
and furthermore are atom economic processes.6 The limi-
tation of the use of an iron catalyst is the competitive
formation of regioisomeric aldols and ketones. Moreover,
the diastereoselectivity of the reaction is low. In the
presence of Ru or Rh catalysts, the reaction is regioselec-
tive but has a much narrow scope.
Cl
(dppe)NiClH
(dppf)NiClH
1
2
Ph
Ph
Ph
P
P
H
Ph
Ni
Cl
(dppp)NiClH
3
Figure 1
In order to improve the selectivity and the efficiency of
this reaction, we have screened various nickel complexes
and we report here our preliminary results demonstrating
that some nickel hydride catalysts are able to perform the
reaction very efficiently and with a complete control of
Catalyst 1 was first generated from (dppe)NiCl2 and
LiBHEt3 (Scheme 2, method A) and tested at 5 mol%
9
(Scheme 3 and entry 1, Table 1). The mixture of two dia-
stereoisomeric aldols was isolated in only 26% yield, the
main compound formed in the reaction being ketone 6.
Such Ni-H complexes can also be obtained by reaction of
the nickel chloride complex and a Grignard reagent fol-
lowed by addition of TMSCl to the resulting intermedi-
ate.10 (Table 1 and Scheme 2, method B).
SYNLETT 2004, No. 1, pp 0093–0096
0
5.
0
1.
2
0
0
4
Advanced online publication: 4.12.2003
DOI: 10.1055/s-2003-43373; Art ID: G27603ST
© Georg Thieme Verlag Stuttgart · New York

94
D. Cuperly et al.
LETTER
Table 1 Catalyst Optimization
Entry
R¢
Cat.a amount
(mol%)
Co-reagent
(mol%)
Time
5
5
6
Yield (%)
Syn/anti12
Yield (%)
1
2
3
4
5
6
7
8
9
Me2CHCH2
Me2CHCH2
Me2CHCH2
Me2CHCH2
Me2CHCH2
Ph
A, 5
B, 5
A, 5
A, 3
B, 5
B, 5
B, 5
A, 3
C, 3
–
2 h
26
94
95
92
89
85b
99
99
97
76/24
60/40
70/30
70/30
60/40
60/40
60/40
60/40
62/38
51
1
–
1 h 25 min
40 min
50 min
45 min
3 h 45 min
2 h
MgBr2, 5
MgBr2, 3
BEt3, 5
–
–
2
5
–
Ph
BEt3, 5
MgBr2, 3
MgBr2, 3
–
Ph
50 min
1 h
–
Ph
–
a Method A: (dppe)NiCl2 (1) + LiHBEt3; method B: (dppe)NiCl2 (1) + 2 Me2CHMgBr then TMSCl; method C: (dppf)NiCl2 (2) + LiHBEt3.
b +12% Recovered starting material.
Method A
underlined that the reaction can be performed with only 3
mol% of the Ni-H hydride catalyst and 3 mol% of MgBr2
(entry 4).
(dppe)NiCl2 + LiHBEt3
(dppe)NiHCl + BEt3 + LiCl
Method B
Moreover, formation of the catalyst from LiHBEt3 gener-
ates BEt3 as a by-product in the reaction mixture. There-
fore, we tried to evaluate the putative influence of this
compound on the reaction. We compared the reactivity of
the catalyst formed following the ‘Grignard’ procedure
(method B) in the presence and in the absence of added
BEt3 (5%). We could observe a slight increase of the
reactivity when benzaldehyde was used as the carbonyl
compound. Such improvement could not be noticed
when isovaleraldehyde was used (entries 5–7). So, it is
clear that BEt3 plays a less important role than MgBr2 in
this reaction.
(dppe)NiCl2 + 2 Me2CHMgBr
Me2CHCHMe2
MgBr2
+
(dppe)Ni
+
+
+ MgCl2
Si
TMSCl
(dppe)NiHCl
Scheme 2
Thus, Ni-H complex 1 was generated in this way and
tested. We observed a significant increase of the efficien-
cy of the reaction as the aldol products were isolated in
94% yield (entry 2). Such a result could be explained by a
possible key role of MgBr2, which is a by-product in the
formation of the active catalyst in method B. So, the reac-
tion has been performed with a solution of the catalyst 1
obtained from LiHBEt3 and in the presence of 5% MgBr2.
In this case, the reaction was completed in 40 minutes and
the mixture of the two aldols was isolated in 95% yield.
No trace of ketone or regioisomeric aldol was isolated
(entry 3). This result demonstrates the critical importance
of the presence of MgBr2. Although the exact role of
MgBr2 has not yet been established, it is known to act as a
coordinating agent in aldol reactions.11 Finally, it must be
Then, we tried catalysts 2 and 3, which bear different
phosphine ligands. In the presence of 3 mol% of catalyst
2 and 3 mol% of MgBr2 (Table 1, method C), the reaction
of octen-3-ol and benzaldehyde was completed after 1
hour and the mixture of the two diastereoisomeric aldols
was isolated in 97% yield (entry 9). So, catalyst 2 exhibits
a close reactivity compared to catalyst 1 (entry 8). On the
contrary, catalyst 3 is completely inefficient not only in
the aldolization reaction but also in the isomerization of
octen-3-ol to octanone 6, starting material being fully re-
covered. This establishes that the nature of the phosphine
ligand has a critical effect on the reaction.
n-Pent
3-5 mol% cat 1 or 2, R'CHO
n-Pent
R'
n-Pent
+
THF
–50 °C to r.t.
OH
O
OH
O
Co-reagent
5
6
4
Scheme 3
Synlett 2004, No. 1, 93–96 © Thieme Stuttgart · New York

LETTER
From Allylic Alcohols to Aldols
95
n-Pent
3 mol% 1,
3 mol% MgBr2
n-Pent
R'
n-Pent
+
OH
O
R'CHO
OH
O
4
6
5
THF
–50 °C to r.t.
Scheme 4
Table 2 Extension to Various Aldehydes
Entry
R¢
Time
5
5
6
Yield (%)
Syn/anti
Yield (%)
1
2
3
4
5
6
7
Ph
50 min
1 h
99
93
84
91
68
94
94a
60/40
66/34
69/31
67/33
76/24
57/43
–
–
Me2CH
2
Et2CH
2 h 30 min
1 h
4
AcNHPh
tBu
–
55 min
45 min
1 h
15
–
AcOCH2furyl
EtCH=CH(CH2)2
–
a Partial isomerization of the double bond of the aldehyde part was observed.
Having a very efficient catalyst in hands we screened var- perature and in the presence of a rather low catalyst load-
ious aldehydes (Scheme 4, Table 2). The reaction is com- ing (3 mol%). The reaction is far more efficient than the
patible with many different aldehydes even with those Ru-H or Rh-H catalyzed reaction2d as the latter is only
bearing a very bulky or a functionalized group. The results possible with a narrow panel of substrates. Furthermore,
obtained with pivalaldehyde (entry 5) and 2-ethylbutyral- the latter reactions were occurring at reflux of THF.
dehyde (entry 3) are particulary noteworthy as very few Similarly, the RuCl2(PPh3)3 catalyzed reaction needs high
aldol reactions (0–20% yield) were observed previously temperatures (90–110 °C).5 A final advantage of these
with Fe, Rh or Ru catalysts.
Ni-H catalysts is that the use of chiral non racemic ligands
can now be envisaged for this reaction. Therefore, ex-
tensions to asymmetric catalysis and total synthesis are
currently under way.
Phenyl allyl alcohol 7 proved to be also a very good sub-
strate for the reaction. Under the same reaction conditions
it gave excellent yields both with benzaldehyde and iso-
butyraldehyde (Scheme 5).
Procedure for Aldolization of an Allylic Alcohol with an Alde-
hyde via a Nickel Complex as Catalyst: Method A.
Ph
3 mol% 1, R'CHO
R'
Ph
To a solution of (dppe)NiCl2 (30 mg, 0.0567 mmol) and anhyd
MgBr2 (10 mg, 0.0567 mmol) in anhyd THF (3 mL) under a nitro-
gen atmosphere was added a 1 M LiBHEt3 solution in THF (57 mL,
0.0567 mmol). The reaction mixture was stirred at r.t. for 5 min and
cooled to –50 °C. Aldehyde (2.08 mmol) and allylic alcohol (1.89
mmol) were successively added. The temperature was raised to r.t.
and the reaction mixture was followed by TLC until disappearance
of the allylic alcohol. The reaction mixture was quenched with a sat.
solution of NH4Cl (15 mL) and the aq phase was extracted with
Et2O (3 × 50 mL). The organic layer was dried (MgSO4) and con-
centrated under reduced pressure. The crude product was purified
by column chromatography on silica gel.12
3 mol% MgBr2
OH
OH
O
THF –50 °C to r.t.
8
7
R' = Me2CH : 86% yield; syn/anti :50/50
R' = Ph : 97% yield ; syn/anti : 60/40
Scheme 5
To conclude, we have shown that NiHCl(dppe)/MgBr2 is
a very active catalytic system for the isomerization–al-
dolization reaction of allylic alcohols and aldehydes. The
reaction is compatible with a wide range of aldehydes in-
cluding very bulky ones. It gives good to excellent yields
and is completely regioselective, however with a modest
diastereoselectivity. It proceeds at, or below, room tem-
Acknowledgment
We thank P. Guenot (CRMPO Rennes) for performing mass
experiments.
Synlett 2004, No. 1, 93–96 © Thieme Stuttgart · New York

96
D. Cuperly et al.
LETTER
(7) (a) Motherwell, W. B.; Sandham, D. A. Tetrahedron Lett.
1992, 33, 6187. (b) Gazzard, L. J.; Motherwell, W. B.;
Sandham, D. A. J. Chem. Soc., Perkin Trans. 1 1999, 979.
(8) (a) Burkhardt, E. R.; Bergman, R. G.; Heathcock, C. H.
Organometallics 1990, 9, 30. (b) Campora, J.; Maya, C. M.;
Palma, P.; Carmona, E.; Gutierrez-Puebla, E.; Ruiz, C. J.
Am. Chem. Soc. 2003, 125, 1482.
(9) (a) Frauenrath, H.; Brethauer, D.; Reim, S.; Maurer, M.;
Raabe, G. Angew. Chem. Int. Ed. 2001, 40, 177.
(b) Frauenrath, H.; Reim, S.; Wiesner, A. Tetrahedron:
Asymmetry 1998, 9, 1103.
References
(1) Comprehensive Organic Chemistry, Additions to C-X p-
Bonds, Part 2, Vol. 2; Heathcock, C. H., Ed.; Pergamon
Press: Oxford, 1991, 99–319.
(2) (a) Crévisy, C.; Wietrich, M.; Le Boulaire, V.; Uma, R.;
Grée, R. Tetrahedron Lett. 2001, 42, 395. (b) Uma, R.;
Gouault, N.; Crévisy, C.; Grée, R. Tetrahedron Lett. 2003,
44, 6187. (c) Cuperly, D.; Crévisy, C.; Grée, R. J. Org.
Chem. 2003, 68, 6392. (d) Uma, R.; Davies, M.; Crévisy,
C.; Grée, R. Tetrahedron Lett. 2001, 42, 3069.
(3) (a) Uma, R.; Crévisy, C.; Grée, R. Chem. Rev. 2003, 103,
27. (b) van der Drift, R. C.; Bouwman, E.; Drent, E. J.
Organomet. Chem. 2002, 650, 1.
(4) In the case of propargylic alcohols a similar approach, using
vanadium catalysts, has been reported recently: Trost, B. M.;
Jonasson, C.; Wuchrer, M. J. Am. Chem. Soc. 2001, 123,
12736.
(5) (a) Yang, X.-F.; Wang, M.; Varma, R. S.; Li, C.-J. Org. Lett.
2003, 5, 657. (b) Wang, M.; Li, C.-J. Tetrahedron Lett.
2002, 43, 3589. (c) Wang, M.; Yang, X.-F.; Li, C.-J. Eur. J.
Org. Chem. 2003, 998.
(6) (a) Trost, B. M. Science 1991, 254, 1471. (b) Trost, B. M.
Angew. Chem., Int. Ed. Engl. 1995, 34, 259.
(10) Malanga, C.; Urso, A.; Lardicci, L. Tetrahedron Lett. 1995,
36, 1133.
(11) Swiss, K. A.; Choi, W.-B.; Liotta, D. C.; Abdel-Magid, A.
F.; Maryanoff, C. A. J. Org. Chem. 1991, 56, 5978.
(12) The stereoisomeric ratios were established by 1H NMR on
the crude reaction mixture. The relative stereochemistries
were attributed by comparison with literature data or
according to the following empirical rule: The C of the
CHOH group is deshielded in the anti-adduct compared to
the syn-adduct and the H of the CHOH group is shielded in
the anti-adduct compared to the syn-adduct.
Synlett 2004, No. 1, 93–96 © Thieme Stuttgart · New York