Chiral lithium(I) binaphtholate salts for the enantioselective direct Mannich-type reaction with a change of syn/anti and absolute stereochemistry

Article abstract of DOI:10.1021/ja909874b

(Chemical Equation Presented) A highly diastereo- and enantioselective direct Mannich-type reaction of aldimines with 1,3-dicarbonyl compounds using Li(I) BINOLate salts as effective Lewis acid-Bronsted base catalysts has been developed. Li(I) BINOLate sa

Full text of DOI:10.1021/ja909874b

Published on Web 12/14/2009
Chiral Lithium(I) Binaphtholate Salts for the Enantioselective Direct Mannich-Type
Reaction with a Change of Syn/Anti and Absolute Stereochemistry
Manabu Hatano,† Takahiro Horibe,† and Kazuaki Ishihara*,†,‡
Graduate School of Engineering, Nagoya UniVersity, and CREST, Japan Science and Technology Agency (JST),
Furo-cho, Chikusa, Nagoya 464-8603, Japan
Received November 21, 2009; E-mail: ishihara@cc.nagoya-u.ac.jp
Table 2. Syn- and Enantioselective Reaction of 1 with 5
Acid-base combination chemistry, particularly with the use of chiral
1,1′-bis(2-naphthol) (BINOL) derivatives, has been widely accepted
as an important concept in modern asymmetric catalysis.1,2 Chiral Li(I)
BINOLate salts3 are among the simplest acid-base bifunctional
catalysts, and Holmes and Kagan4 developed the trimethylsilylcya-
nation of aldehydes with the use of dry chiral Li(I) BINOLate salts.
Later, we reported that wet or alcoholic Li(I) BINOLate salts showed
dramatically improved catalytic activity as Lewis acid-Lewis base
catalysts for trimethylsilylcyanation by activating both an aldehyde
and Me3SiCN.5 By taking advantage of the strong basicity of
naphtholate oxygen, Li(I) BINOLate salts should be good Lewis
acid-Brønsted base catalysts, which activate both a substrate and an
acidic pronucleophile. In this paper, we describe a highly diastereo-
and enantioselective direct Mannich-type reaction of aldimines with
1,3-dicarbonyl compounds that involves a change of syn/anti and
absolute stereochemistry, with the use of Li(I) BINOLate salts as the
simplest chiral acid-base catalysts.
First we examined the enantioselective direct Mannich-type reaction
of aldimine 1a with acetylacetone (2) (Table 1).6,7 Mono- and dilithium
salts of (R)-BINOL (5 mol %) gave (R)-3 with poor enantioselectivities
(entries 1 and 2). The monolithium salt of (R)-3,3′-Ph2-BINOL
improved the enantioselectivity to 38% ee (entry 3). The addition of
t-BuOH significantly improved the catalytic activity, and (R)-3 was
obtained in 93% yield with 82% ee (entry 4). Moreover, the
monolithium salt of (R)-3,3′-(3,4,5-F3-C6H2)2-BINOL (4) with t-BuOH
was found to be more effective, and the catalyst loading could be
decreased to 2.5 mol % (entries 5 and 6).
entry
1
(R, Ar)
product yield (%) dr (syn/anti) ee (%)a
1
2
3
4
5
6
7
8
1a (Boc, Ph)
6a
6b
6c
6d
6e
6f
6g
6h
6a
6a
6a
6i
98
93
93
95
94
88:12
88:12
88:12
96:4
96:4
91:9
95:5
97:3
90:10
88:12
92:8
91
93
93
95
90
91
87
95
84
95
90
87
1b (Boc, 3-MeC6H4)
1c (Boc, 4-MeC6H4)
1d (Boc, 4-MeOC6H4)
1e (Boc, 4-ClC6H4)
1f (Boc, 2-Naph)
1g (Boc, 2-furyl)
1h (Boc, 3-thienyl)
96
>99
>99
97
71
96
9b 1a (Boc, Ph)
10c 1a (Boc, Ph)
11d 1a (Boc, Ph)
12
1i
(Cbz, Ph)
91
95:5
a Values are for syn-6. b 4 (1 mol %), n-BuLi (1 mol %), and t-BuOH
(2 mol %) were used. c 4 (5 mol %), n-BuLi (5 mol %), and t-BuOH
(10 mol %) were used. The reaction time was 3 min. d LiOH (2.5 mol
%) was used in place of n-BuLi and t-BuOH.
for syn-6 as the major product, with the formation of a chiral quaternary
carbon center (entries 1-8). A catalyst loading of 1 mol % was still
effective, and 6a was obtained in 97% yield with high diastereo- and
enantioselectivity (entry 9). Since the observed turnover frequency was
284 h-1 (71% yield at 3 min) (entry 10), the high reactivity of this
catalyst is quite unlike those of other previous catalysts.6,7 Moreover,
another advantage is that inexpensive LiOH can be used as the lithium
precursor in place of n-BuLi/t-BuOH (entry 11). When N-Cbz-aldimine
1i was used, syn-6i was obtained as the major product with 87% ee
(entry 12). Notably, we found that a dramatic changeover of the
absolute configuration at the amino carbon center took place in going
from 3 to 6 (see below).
Table 1. Screening of Li(I) BINOLate Salts
n-BuLi
(mol %)
t-BuOH
(mol %)
yield
(%)
ee
(%)
entry
BINOL [mol %]
1
2
3
4
5
6
(R)-BINOL [5]
(R)-BINOL [5]
(R)-3,3′-Ph2-BINOL [5]
(R)-3,3′-Ph2-BINOL [5]
4 [5]
5
10
5
5
5
0
0
0
10
10
5
72
18
40
93
>99
>99
8
0
38
82
80
82
To demonstrate the synthetic utility of this approach, adduct 6a
(90% ee) was transformed to the useful spiro ꢀ-lactam8 7 with three
consecutive chiral carbon centers in a highly diastereo- and
enantioselective manner (eq 1):
4 [2.5]
2.5
Under the optimized conditions using 4, we next examined the
reaction of a cyclic ketoester (5) with a variety of aryl aldimines with
an electron-withdrawing or electron-donating group (1a-f and 1i) and
with heteroaryl aldimines (1g and 1h) (Table 2). The reactions
proceeded smoothly at -78 °C within 2 h, and the desired products 6
were obtained in high yields with high diastereoselectivities (syn/anti
) 88:12-97:3). High enantioselectivities (87-95% ee) were observed
Furthermore, we explored the scope of the diastereo- and enanti-
oselective direct Mannich-type reaction with a class of 1,3-dicarbonyl
compounds (Table 3). Consequently, the Li(I) BINOLate salt was
shown to be one of the most efficient catalysts and, unlike previous
catalysts, showed high catalytic activity (at -78 °C within 2 h) toward
cyclic and acyclic 1,3-dicarbonyl compounds. Cyclic ketoesters, acyclic
ketoesters, a ketothioester, and a ketoamide gave the corresponding
† Nagoya University.
‡ Japan Science and Technology Agency.
9
56 J. AM. CHEM. SOC. 2010, 132, 56–57
10.1021/ja909874b  2010 American Chemical Society

C O M M U N I C A T I O N S
Table 3. Diastereo- and Enantioselective Reaction
pathway (eq 2, syn). Therefore, we rationalized that syn isomers would
be obtained via path b while anti isomers would be obtained via path
a. In path a, a pronucleophile would be activated by coordination
with a Brønsted basic naphtholate oxygen, after which attack of the
Li-coordinated aldimine on the re face would occur. In path b, the
aldimine would be activated by a resulting Brønsted acidic naphthol
proton, and the Li enolate would attack the aldimine on the si face.
Figure 1. Proposed precursor and transition states.
a Using 10 mol % catalyst. b The temperature was -40 °C.
In summary, we have developed a highly enantioselective direct
Mannich-type reaction of aldimines with 1,3-dicarbonyl compounds.
A simple Li(I) BINOLate salt was effective for selectively synthesizing
both syn adducts from cyclic reagents and unprecedented anti adducts
from acyclic reagents with an interesting change in absolute
stereochemistry.
products (9a-h) with high enantioselectivities (85-97% ee).9 Notably,
anti products (9d-h) were selectively obtained from acyclic reagents
without epimerization at the R-3°-carbon center, and these are valuable
since previous catalysts often gave syn/anti mixtures or the stereo-
chemistry has not yet been determined.6,7 A ketolactone and a cyclic
diketone also gave the adducts (9i and 9j) in high yields with up to
99% ee. From an S-heterocyclic ketoester, the desired adduct (9k) was
obtained in 90% yield with a syn/anti ratio of 88:12 in 93% ee.
Subsequent reduction and desulfurization of 9k with Raney Ni gave
the valuable acyclic ꢀ-amino dicarbonyl compound 9l having three
consecutive chiral carbons and involving a methyl-substituted quater-
nary center.
Interestingly, syn-9a-c,k and anti-9d-h from cyclic and acyclic
pronucleophiles, respectively, offered opposite absolute configurations
at the amino carbon center (Table 3). However, during our investigation
of these unexpected changes in stereochemistry (see above), we found
that cyclic ketoester 10 gave a mixture of syn-11 and anti-11 at a
diastereomeric ratio of 52:48 with high enantioselectivities (90 and
97% ee, respectively) (eq 2): Moreover, when 3,5-xylenol was used
Acknowledgment. Financial support for this project was provided
by JSPS KAKENHI (20245022), MEXT KAKENHI (21750094,
21200033), and the Global COE Program of MEXT. We greatly thank
the referees for helpful suggestions regarding the mechanistic aspects.
Supporting Information Available: Experimental procedures. This
material is available free of charge via the Internet at http://pubs.acs.org.
References
(1) Reviews of acid-base chemistry: (a) Kanai, M.; Kato, N.; Ichikawa, E.;
Shibasaki, M. Synlett 2005, 1491. (b) Ishihara, K.; Sakakura, A.; Hatano,
M. Synlett 2007, 686.
(2) Reviews of BINOL chemistry: (a) Chen, Y.; Yekta, S.; Yudin, A. K. Chem.
ReV. 2003, 103, 3155. (b) Shibasaki, M.; Matsunaga, S. Chem. Soc. ReV.
2006, 35, 269. (c) Brunel, J. M. Chem. ReV. 2007, 107, PR1.
(3) Chiral Li(I) BINOLate catalyses: (a) Schiffers, R.; Kagan, H. B. Synlett
1997, 1175. (b) Loog, O.; Ma¨eorg, U. Tetrahedron: Asymmetry 1999, 10,
2411. (c) Nakajima, M.; Orito, Y.; Ishizuka, T.; Hashimoto, S. Org. Lett.
2004, 6, 3763. (d) Ichibakase, T.; Orito, Y.; Nakajima, M. Tetrahedron
Lett. 2008, 49, 4427.
(4) Holmes, I. P.; Kagan, H. B. Tetrahedron Lett. 2000, 41, 7453.
(5) (a) Hatano, M.; Ikeno, T.; Miyamoto, T.; Ishihara, K. J. Am. Chem. Soc.
2005, 127, 10776. (b) Hatano, M.; Ikeno, T.; Matsumura, T.; Torii, S.;
Ishihara, K. AdV. Synth. Catal. 2008, 350, 1776.
(6) Reviews of direct Mannich-type reactions: (a) Co´rdova, A. Acc. Chem.
Res. 2004, 37, 102. (b) Ting, A.; Schaus, S. E. Eur. J. Org. Chem. 2007,
5797. (c) Verkade, J. M. M.; van Hemert, L. J. C.; Quaedflieg, P. J. L. M.;
Rutjes, F. P. J. T. Chem. Soc. ReV. 2008, 37, 29.
in place of t-BuOH, 11 was obtained with similar diastereoselectivity,
but the enantioselectivity of syn-11 was significantly decreased to 43%
ee.10 Overall, these results strongly suggest that the syn and anti isomers
were obtained thorough different reaction pathways.
(7) Asymmetric direct Mannich-type reactions of aldimines with 1,3-dicarbonyl
compounds: (a) Uraguchi, D.; Terada, M. J. Am. Chem. Soc. 2004, 126,
5356. (b) Hamashima, Y.; Sasamoto, N.; Hotta, D.; Somei, H.; Umebayashi,
N.; Sodeoka, M. Angew. Chem., Int. Ed. 2005, 44, 1525. (c) Lou, S.; Taoka,
B. M.; Ting, A.; Schaus, S. E. J. Am. Chem. Soc. 2005, 127, 11256. (d)
Tillman, A. L.; Ye, J.; Dixon, D. J. Chem. Commun. 2006, 1191. (e) Song,
J.; Wang, Y.; Deng, L. J. Am. Chem. Soc. 2006, 128, 6048. (f) Fini, F.;
Bernardi, L.; Herrera, R. P.; Pettersen, D.; Ricci, A.; Sgarzani, V. AdV.
Synth. Catal. 2006, 348, 2043. (g) Yamaoka, Y.; Miyabe, H.; Yasui, Y.;
Takemoto, Y. Synthesis 2007, 2571. (h) Lou, S.; Dai, P.; Schaus, S. E. J.
Org. Chem. 2007, 72, 9998. (i) Chen, Z.; Morimoto, H.; Matsunaga, S.;
Shibasaki, M. J. Am. Chem. Soc. 2008, 130, 2170. (j) Hatano, M.; Maki,
T.; Moriyama, K.; Arinobe, M.; Ishihara, K. J. Am. Chem. Soc. 2008, 130,
16858. (k) Kang, Y. K.; Kim, D. Y. J. Org. Chem. 2009, 74, 5734. (l)
Han, X.; Kwiatkowski, J.; Xue, F.; Huang, K.-W.; Lu, Y. Angew. Chem.,
Int. Ed. 2009, 48, 7604.
Although the role of the alcoholic Li(I) salt is not completely clear,
two major mechanisms can be considered (Figure 1).11 On one hand,
the aldimine would be activated by the Lewis acidic Li(I) center (path
a). On the other hand, the pronucleophile would be activated as a Li
enolate (path b). In both paths, the addition of alcohols such as t-BuOH
would promote the dissociation to a monomeric precursor, Li(I)
BINOLate ·(ROH)n (Figure 1, left).5 Since an acyclic 1,3-dicarbonyl
compound would have a rather flexible conformation, activation of
the aldimine on the Li(I) center would be likely (path a). In path a,
alcohols (t-BuOH or 3,5-xylenol) would still coordinate to the Li(I)
center, which would not strongly affect the enantioselectivity (eq 2,
anti). In sharp contrast, a cyclic 1,3-dicarbonyl compound would be
preferentially activated as a Li enolate because the conformationally
rigid structure could promote chelation to the Li(I) center (path b).
Ligand 4 protonated via Li enolization would activate the aldimine
through hydrogen bonding. However, 3,5-xylenol, which is more acidic
than t-BuOH, might compete with 4 and could trigger a nonselective
(8) Bhalla, A.; Venugopalan, P.; Bari, S. S. Eur. J. Org. Chem. 2006, 4943.
(9) The reaction of PMPNdCCO2Et with 10 gave the corresponding products
in 76% yield (dr ) 71:29) with 91/26% ee.
(10) We can exclude a pathway with achiral 3,5-Me2C6H3OLi in eq 2 because
the dr ratios in eq 2 were nearly the same even though the reaction of 1a
and 10 with 3,5-xylenol/n-BuLi (5 mol%) gave syn-11 in 98% yield.
(11) A more detailed mechanism is discussed in the Supporting Information.
JA909874B
9
J. AM. CHEM. SOC. VOL. 132, NO. 1, 2010 57