Y. Hasegawa et al.
Bull. Chem. Soc. Jpn. Vol. 85, No. 3 (2012)
317
Ts
Rn
chiral Ir cat 1:
C6H5
N
Ru
N
R
S
N
O
O
Ir
C6H5
H
1a: R = CH3 (Ms)
C6H5
C6H5
H
H
1b: R = 4-CH3-C6H4 (Ts)
1c: R = (1R)-camphor (Cs)
1d: R = (CH3)5C6 (PMs)
Ar
Ar
amine hydrido
complex
N
H
O
O
H
chiral Ru cat 2:
H
R
O
S
N
O
OH
OH
Ts
Rn
C6H5
C6H5
2a: R = CH3 (Ms)
C6H5
C6H5
N
2b: R = 4-CH3-C6H4 (Ts)
2c: R = (1R)-camphor (Cs)
2d: R = (CH3)5C6 (PMs)
Ru
Ru
N
H
N
H
amido complex
Scheme 1. Representative bifunctional catalysts and mechanism of asymmetric transfer hydrogenation with chiral catalyst.
The modern view of the most effective chiral molecular
catalysts has drifted from the initial ideas of structural
robustness and monofunctionality,1f and it is now widely
recognized that chiral bifunctional catalysts with acid-base
synergy are especially effective in creating chiral 3D pockets
suitable for arranging highly enantioselective transforma-
tions.2i,2q,18,19 In addition, it has been recently recognized that
metal and cooperating ligands can activate reacting substrates
in a bifunctional mode and facilitate the enantioselective bond
forming steps.20 Ultimately such catalysts with a variable
structure would be able to perform multiple enantioselective
transformations in response to an external signal.21,22
Recently, we have shown that chiral bifunctional Ir and Ru
catalysts, Cp*Ir[(S,S)-N-sulfonated dpen] 1 and Ru[(S,S)-N-
sulfonated dpen](©6-arene) 2 (DPEN: 1,2-diphenylethylene-
diamine) that were initially developed for asymmetric trans-
fer hydrogenation of ketones and imines with 2-propanol as
shown in Scheme 1,23 are applicable to other stereoselective
transformations24 triggered by the proton transfer from a
pronucleophile to the 16e amido Ir or Ru complex, thus effec-
tuating the catalytic C-C20d,25,26 or C-N20d,27 bond formation
on the resulting catalyst intermediate, chiral amine complex.
The objectives of this work are mechanistic consideration
of two new catalytic reactions promoted by chiral bifunc-
tional amido complexes 1 and 2: asymmetric C-N and C-C
bond formation in the reactions of ¡-cyanoacetates 3 as the
pronucleophile with dialkyl azodicarboxylates 427-29 or acety-
lenic esters 626,30,31 as the acceptor molecule, respectively.
Noticeably, both reactions gave the corresponding chiral
adducts 5 and 7 with excellent ees’, but with the opposite
configuration of the quaternary carbon atom centers. It means
that different forms of the catalyst or the catalyst-donor
complex become active in the reaction with different acceptor
molecules, in which the structures and properties of the
substrates determine the stereochemistry of the bond formation
steps. Some examples of such substrate-controlled inversion
in terms of enantioselectivity are known,32 however, they
remain relatively rare. Detailed mechanistic studies of such
asymmetric catalytic reactions can reveal the structural features
responsible for the switch of the reaction pathways, and hence
of the sense of enantioselection.1f,2p,33
In this paper we report the optimization of reaction
conditions for these two reactions, catalyst tuning, and struc-
tural modification of the substrates, as well as an experimental
mechanistic study combined with a computational investigation
of the catalytic cycle and the origin of the enantioselectivity
in these two reactions, leading to the chiral adducts with the
opposite configuration.
Results and Discussion
Enantioselective C-N Bond Formation in the Conjugate
Addition of ¡-Substituted ¡-Cyanoacetates to Azodicar-
boxylic Diesters. We have recently reported that a chiral
amido Ir complex, Cp*Ir[(S,S)-Msdpen] 1a (MsDPEN: (S,S)-
N-(methylsulfonyl)-1,2-diphenylethylenediamine)
catalyzed
asymmetric direct amination of tert-butyl ¡-phenyl-¡-cyano-
acetate (3a) using dimethyl azodicarboxylate (4a) (a substrate/
catalyst ratio, S/C = 100, 3a:4a = 1:1) in a 0.5 M toluene
solution at 0 °C for 2 h to produce the corresponding hydrazine
adduct 5a in 99% yield and with 85% ee as shown in Scheme 2
and Table 1 (Run 1).27 Since the noncatalyzed reaction of
3a and 4a gave quantitatively the racemic product, 5a (Run 2),
the reaction conditions were modified to minimize the non-
catalyzed reaction. In fact, a slow addition of azodicarboxylate,
4a, to a toluene solution of 3a containing the Ir catalyst 1a
with a syringe pump for 20 min at 0 °C followed by stirring of
the reaction mixture for 2 h afforded 5a in 95% ee. Decreasing
the reaction temperature to ¹40 °C resulted in an additional
increase in the ee value of 5a to 97% (Runs 4 and 5).
The stereochemical outcome of the reaction with the chiral
Ir catalyst was delicately influenced by the structure of the
Ir complexes. The CsDPEN (Cs: (R)-camphorsulfonyl) com-
plex 1c gave the best selectivity; the ee value of the product
reached 98% ee in this case (Run 7). In toluene, THF, or
C2H5(CH3)2COH, high ee values were observed (Runs 4, 8,
and 9), while acetone and CH2Cl2 gave unsatisfactory results
(Runs 10 and 11). Acetonitrile gave a racemic adduct probably
because of its strong binding to the center metal. An increase in