Dramatic ligand effect in catalytic asymmetric reductive aldol reaction of allenic esters to ketones

Article abstract of DOI:10.1021/ja0652565

A general catalytic asymmetric reductive aldol reaction of allenic esters to ketones is described. Two distinct constitutional isomers were selectively produced depending on the reaction conditions. A combination of CuOAc/(R)-DTBM-SEGPHOS/PCy₃

Full text of DOI:10.1021/ja0652565

Published on Web 10/25/2006
Dramatic Ligand Effect in Catalytic Asymmetric Reductive Aldol Reaction of
Allenic Esters to Ketones
Dongbo Zhao, Kounosuke Oisaki, Motomu Kanai,* and Masakatsu Shibasaki*
Graduate School of Pharmaceutical Sciences, The UniVersity of Tokyo, Tokyo 113-0033, Japan
Received July 22, 2006; E-mail: mshibasa@mol.f.u-tokyo.ac.jp
Chiral tertiary alcohols are important building blocks in many
biologically active naturally occurring compounds and artificial
1
pharmaceuticals. Catalytic enantioselective addition reactions of
carbon nucleophiles targeting ketones produce enantiomerically
enriched tertiary alcohols through carbon-carbon bond formation.
Among recent intensive studies to develop such reactions, the
Figure 1. Chiral phosphine ligands used in our system.
catalytic enantioselective aldol reaction to simple ketones has just
Table 1. Copper Salt and Additive Effects on γ-Selective
Reductive Aldol Reaction of 2b to 1a
2
begun to be explored. After the pioneering report by the Denmark
3a,b
group using trichlorosilyl enolate as a nucleophile, Campagne’s
3c
vinylogous aldol reaction of trimethylsilyl dienolate and our aldol
reaction using trimethylsilyl enolates3 were developed. Preacti-
vation of nucleophiles is essential in those three reactions. On the
other hand, a catalytic enantioselective reductive aldol reaction to
ketones via in situ generation of metal enolates is an attractive
alternative. The first example of this reaction type was demonstrated
d-f
Yield/%b
additive
4
a
by Lam’s group, but it was restricted to an intramolecular version.
entry
copper salt
(x mol %)
3a (ee/%)c
4
5a + 6
Our group4b and the Riant group independently developed catalytic
enantioselective intermolecular reductive aldol reactions to simple
ketones using either an allenic ester or an acrylate ester as the
prenucleophile. These reactions should be improved, however,
especially in terms of substrate generality. In this communication,
we report a general catalytic asymmetric reductive aldol reaction
of allenic esters to ketones. Specifically, two different constitutional
isomers (R- and γ-aldol products) were selectively synthesized from
the same starting materials, depending on the conditions.
4c
1
2
CuF‚3PPh3‚2EtOH
CuF2‚2H2O
CuOAc
CuOAc
CuOAc
46 (99)
trace
<4
trace
<3
22
31
18
6
9
4
d
3
4
5
6
7
32 (99)
60 (99)
90 (99)
96 (99)
78 (99)
PPh (7.5)
PCy3 (7.5)
PCy3 (5.0)
PCy3 (2.5)
3
CuOAc
CuOAc
8
a
Slow addition of pinacolborane for 4 h. b Yield was determined by H
1
NMR of the crude mixture. c Enantiomeric excess was determined by HPLC
after isolation of pure 3a. d Catalyst was prepared in refluxing methanol
for 2 h.
Successful catalytic asymmetric reductive aldol reactions of
allenic esters to ketones depend on the control of various types of
selectivities, namely, chemo- (ketone vs allenic ester in the initial
reduction step), regio- (R- vs γ-addition in the vinylogous aldol
reaction step), stereo- (cis vs trans of γ-adducts), diastereo- (syn
was used instead of PPh (entry 5). Finally, the best result was
3
obtained when CuOAc/(R)-DTBM-SEGPHOS/PCy (1:2:2) was
3
used as the catalyst (entry 6); the γ-cis-adduct (3a) was produced
predominantly in 96% yield with excellent γ/R regioselectivity
(25:1) and enantioselectivity (99% ee).
5
vs anti of R-adducts), and enantioselectivity. In our previous report,
there was only moderate γ/R regioselectivity (3a+4/5a+6) of the
reaction between acetophenone (1a) and allenic ester (2b) using
Substrate generality was investigated under the optimized condi-
tions of the γ-cis-selective reductive aldol reaction (Table 2). Both
aromatic (entries 1-4) and aliphatic (entries 5-9) ketones were
converted to γ-cis-products 3, generally in high yield and excellent
enantioselectivity. It is noteworthy that our system is applicable to
unsaturated ketones such as 1e and 1g via chemoselective reductive
enolate formation from the allenic ester. The products 3 are versatile
chiral building blocks in organic synthesis. For example, 3a was
converted to the corresponding R,â-unsaturated δ-lactone in greater
than 99% yield through heating in AcOH under reflux without any
loss of enantiopurity.
CuF‚3PPh ‚2EtOH-(R)-DTBM-SEGPHOS as the catalyst, al-
3
though the cis/trans selectivity (3a/4) and enantioselectivity of the
major γ-cis-product (3a) were promising (Table 1, entry 1).4b
To improve regioselectivity, effects of copper salt and an additive
were investigated. The representative results are summarized in
Table 1. When a chiral Cu complex was reductively generated in
6
situ from CuF
2
‚2H
2
O and excess (R)-DTBM-SEGPHOS, a very
small amount of the desired γ-product was produced (entry 2).
These contrasting results directed our attention to the effect of the
achiral phosphine (PPh
higher γ-selectivity when CuOAc was used as the Cu salt (entries
and 4); PPh -free CuOAc produced moderate γ/R regioselec-
tivity (entry 3), whereas the combination of CuOAc/(R)-DTBM-
SEGPHOS/PPh (1:2:3) significantly improved the γ/R regiose-
lectivity to 10:1 without any loss of the enantioselectivity of 3a
entry 4). Moreover, generation of the γ-trans-product (4) was
completely eliminated in the presence of the achiral phosphine
additive. The chemical yield of 3a was further improved when PCy
3
). A similar tendency was observed with
During the course of our optimization, regioselectivity was
switched to be exclusively R-selective when Taniaphos-type ligands
7
3
3
were used (Table 3). The initial trials were performed with 1a and
2b using commercially available L1 as the chiral ligand. As a result,
R-products were produced exclusively without the concomitant
generation of trace amounts of γ-products (entry 1). The diastereo-
selectivity of the R-products was high (5a/6 ) 6/1), and the major
isomer (5a) was obtained in 66% ee. Optimization of the allenic
3
(
3
14440
9
J. AM. CHEM. SOC. 2006, 128, 14440-14441
10.1021/ja0652565 CCC: $33.50 © 2006 American Chemical Society

C O M M U N I C A T I O N S
Scheme 1. Typical Conversion of R-Vinyl Aldol Product
Table 2. γ-cis-Selective Reductive Aldol Reaction to Ketones
Scheme 2. Preliminary Extension to Catalytic Asymmetric
Alkylative Aldol Reaction to Ketones
ratiob
yield/%c
ee/%d
entry
ketone (R)
γ/
R
product
1
2
3
4
5
6
7
8
9
1a (phenyl)
25/1
13/1
9/1
9/1
30/1
3/1
>8/1
>6/1
>8/1
3a
3b
3c
3d
3e
3f
3g
3h
3i
96
93
90
90
97
70
86
86
80
99e
98
97
99
84
96
89
88
98
1b (p-Cl phenyl)
1c (p-Me phenyl)
1d (m-Cl phenyl)
1e (cinnamyl)
1f (phenylethyl)
1g (homoallyl)
1h (n-butyl)
f
f
g
g
g
1). Previously, it was difficult to synthesize such a chiral building
block in enantiomerically enriched form.
h
1i (isopropyl)
Finally, the basic reaction pattern (i.e., conjugate addition
followed by aldol reaction) can be extended to another catalytic
asymmetric multicomponent reactionsan alkylative aldol reaction
to ketones initiated by a Cu-catalyzed conjugate addition of
dialkylzinc to an allenic ester (Scheme 2). Thus, using CuOAc-
DTBM-SEGPHOS complex as a catalyst (5 mol %), â-alkylated
δ-lactones (9 and 10) were produced in high enantioselectivity from
a ketone, an allenic ester, and dialkylzinc reagents. No R-aldol
products were isolated in these cases.
a
To a mixture of catalyst, 1, and 2b, pinacolborane was added slowly
b
1
over 4 h. The γ/R ratio was determined by H NMR of the crude mixture.
Yield was determined by H NMR of the crude mixture using an internal
standard. Deviation from the isolated yield was generally <10%. Enan-
tiomeric excess was determined by HPLC. The absolute configuration was
determined to be (R). Reaction was conducted at -20 °C. NMR peaks
were overlapping. 5 mol % of catalyst was used.
c
1
d
e
f
g
h
Table 3. R-Selective Reductive Aldol Reaction to Ketones
In conclusion, we developed an asymmetric reductive aldol
reaction between ketones and allenic esters catalyzed by chiral
Cu(I) complexes. The product constitution (R- or γ-aldol) can be
switched depending on the structure of chiral diphosphine ligands.
No preactivation of the substrates is necessary. Clarification of the
reaction mechanism and how the two reaction pathways are
ketone
(R)
allene
(R
5 or
entry ligand
′
)
product
7/6b
yield/%c ee/%d
10
differentiated depending on the conditions and further improve-
ment of the alkylative aldol reaction are ongoing.
1
2
3
4
5
6
7
8
9
L1 1a (phenyl)
L1 1a (phenyl)
L2 1a (phenyl)
L3 1a (phenyl)
L3 1b (p-Cl phenyl) 2a (Me)
L3 1d (m-Cl phenyl) 2a (Me)
L3 1j (p-I phenyl)
L3 1k (2-naphthyl)
L3 1e (cinnamyl)
2b (Et)
5a
7a
7a
7a
7b
7d
7j
6/1
86
90
90
90
89
89
86
91
87
66
67
72
84
83
82
84
84
67
e
e
e
2a (Me)
2a (Me)
2a (Me)
10/1
11/1
10/1
8/1
8/1
7/1
Acknowledgment. Financial support was provided by a Grand-
in-Aid for Specially Promoted Research of MEXT. K.O. thanks
JSPS for the research fellowship.
2a (Me)
2a (Me)
2a (Me)
Supporting Information Available: Results of reaction condition
optimizations, experimental procedures, and characterization of the
products (PDF). This material is available free of charge via the Internet
at http://pubs.acs.org.
7k
7e
10/1
9/1
a
To a mixture of catalyst, 1, and 2, pinacolborane was added slowly
over 4 h. Ratio was determined by H NMR of the crude mixture. Yield
of 5 or 7 was determined by H NMR of the crude mixture using an internal
standard. Deviation from the isolated yield was generally <10%. Enan-
tiomeric excess of 5 or 7 was determined by HPLC after hydrogenation of
the terminal olefin. The absolute configuration was determined to be (R,R).
b
1
c
1
References
d
(
(
(
1) (a) Denissova, I.; Barriault, L. Tetrahedron 2003, 59, 10105. (b) Corey,
E. J.; Guzman-Perez, A. Angew. Chem., Int. Ed. 1998, 37, 388.
2) For recent examples of catalytic enantioselective additions to ketones,
see references cited in ref 4c.
e
3) (a) Denmark, S. E.; Fan, Y. J. Am. Chem. Soc. 2002, 124, 4233. (b)
Denmark, S. E.; Fan, Y.; Eastgate, M. D. J. Org. Chem. 2005, 70, 5235.
(c) Moreau, X.; Tejeda, B.; Campagne, J.-M. J. Am. Chem. Soc. 2005,
127, 7288. (d) Oisaki, K.; Suto, Y.; Kanai, M.; Shibasaki, M. J. Am. Chem.
Soc. 2003, 125, 5644. (e) Oisaki, K.; Zhao, D.; Suto, Y.; Kanai, M.;
Shibasaki, M. Tetrahedron Lett. 2005, 46, 4325. (f) Oisaki, K.; Zhao, D.;
Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2006, 128, 7164.
4) (a) Lam, H. W.; Joensuu, P. M. Org. Lett. 2005, 7, 4225. (b) Zhao, D.;
Oisaki, K.; Kanai, M.; Shibasaki, M. Tetrahedron Lett. 2006, 47, 1403.
ester structure further improved the diastereoselectivity to 10/1
without affecting the enantioselectivity (67% ee), using methyl ester
a (entry 2). Encouraged by this finding, a series of Taniaphos
8
2
analogues were synthesized and screened (entries 2-4). Taniaphos
L3 containing bulkier di(3,5-xylyl)phosphines and a morpholine
unit gave the best results, producing the R-adducts with a 10:1
diastereoselectivity and 84% ee for the major isomer (7a; entry 4).
Next, the substrate scope of the R-selective reductive aldol
reaction was investigated (entries 4-9). Various aromatic ketones
were converted to the corresponding products with high diastereo-
and enantioselectivity. The reaction of an enone (1e) proceeded
chemoselectively with high diastereoselectivity, although the enan-
tioselectivity remained to be improved. The existence of a terminal
olefin in 7 allows for further conversion to various R-substituted
(
(
c) Deschamp, J.; Chuzel, O.; Hannedouche, J.; Riant, O. Angew. Chem.,
Int. Ed. 2006, 45, 1292.
(
5) For a general review of catalytic asymmetric multicomponent reactions,
see: Ramon, D. J.; Yus, M. Angew. Chem., Int. Ed. 2005, 44, 1602.
6) Wada, R.; Oisaki, K.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2004,
(
1
26, 8910.
(7) Ireland, T.; Tappe, K.; Grossheimann, G.; Knochel, P. Chem.sEur. J.
002, 8, 843 and references cited therein.
2
(
8) In contrast to the γ-cis-selective reductive aldol reaction, there was no
significant copper salt and additive effects in the R-selective reaction.
9) Morgan, J. P.; Grubbs, R. H. Org. Lett. 2000, 2, 3153.
(
(
10) For a review addressing the regioselectivity of vinylogous aldol reactions,
see: Denmark, S. E.; Heemstra, J. R., Jr.; Beutner, G. L. Angew. Chem.,
Int. Ed. 2005, 44, 4682.
9
aldol products using the cross-metathesis reaction. For example,
a was converted to 8 containing a longer alkyl chain at the
7
R-position without any epimerization and racemization (Scheme
JA0652565
J. AM. CHEM. SOC.
9
VOL. 128, NO. 45, 2006 14441