Communication
obtained with high selectivity (up to 93%) by simply mixing an
aniline derivative, a cycloalkylamine, glyoxal and formaldehyde
in the presence of acetic acid at 408C for few minutes. There-
fore, we postulated that the replacement of the cycloalkyl-
amine component by an enantiopure amino alcohol or amino
acid would lead to a powerful and practical synthesis of bulky
chiral bifunctional unsaturated NHC ligand precursors with no
motif. To our delight, the strategy appeared also efficient in
presence of enantiopure amino acids to afford chiral NHC pre-
cursors 3 bearing a carboxylic acid function beyond reach oth-
erwise. In fact, this substitution pattern is inaccessible by qua-
ternarization of aryl-imidazole with chiral secondary alkyl hal-
[16]
ides. Moreover, the original multistep method to access satu-
rated NHC analogues included a reduction step with lithium
[
13]
[6]
equivalent in the literature. Nevertheless, the success of this
approach was uncertain owing to several key issues. First, the
selectivity for the unsymmetrical salt can be dramatically af-
fected by a change in the electronic and steric properties of
aluminum hydride (Scheme 1, Step 3), which is incompatible
with the use of amino acid. Interestingly, a workup under basic
conditions with NaHCO afforded the carboxylate imidazolium
3
[17]
zwitterion 3b, which could be isolated in satisfactory yield.
With this new family of imidazolium salts in hand, we natu-
rally became interested in evaluating their potential as chiral li-
gands for metal-catalyzed asymmetric transformations. The cre-
ation of enantioenriched all-carbon quaternary centers is of
[
12]
the aliphatic amine. Second, b-amino-alcohols are prone to
undergo condensation reactions with formaldehyde to form
[
14]
oxazolidines and bis(oxazolidine) adducts. Moreover similar
[15]
condensation reactions were reported with glyoxal.
Third
[
18]
the selectivity may also suffer from the difference of solubility
between the two different amines, especially when free a-
amino acids are involved.
high synthetic interest.
asymmetric conjugate addition (ACA) is an attractive strategy
Moreover, the copper-catalyzed
[19]
to construct such challenging motifs and some hydroxy-che-
lating diaminocarbene-based ligands have been previously em-
To our delight, the first attempt validated our hypothesis
and after optimization of the reaction conditions, the introduc-
tion of one equivalent of (S)-leucinol in the multicomponent
procedure afforded the desired chiral imidazolium salt 2a·PF6
with high 91% selectivity and good 59% isolated yield (2 g
[7,20]
ployed with success in this powerful transformation.
Never-
theless, it should be noted that, chiral hydroxyalkyl-ligands 1,
have not been previously described in the copper-catalyzed
ACA of dialkylzinc reagent to a b-substituted cyclic enone.
Therefore, with the objective to extend the scope of our
ligand family, the newly developed chiral bidentate unsymmet-
rical imidazolium salts were first evaluated in this challenging
transformation (Table 2).
scale) (Table 1). The NHC precursors 2b·PF derived from (S)-va-
6
linol was obtained with a slight decrease of the selectivity
(83%) and reasonable 48% isolated yield. On the other hand,
a large increase of the steric demand on the branched-chain
amino alcohol considerably disfavored the formation of the de-
sired unsymmetrical imidazolium salt, with accumulation of the
symmetrical 1,3-bis(2,4,6-trimethylphenyl)imidazolium salt. In
fact, poor 18% selectivity and 15% isolated yield were ob-
served with the highly sterically congested (S)-tert-leucinol
Table 2. Evaluation of chiral NHC ligands for copper-catalyzed ACA of di-
alkylzinc.
Table 1. Multicomponent synthesis of chiral bidentate imidazolium salts.
[
a]
[b]
[c]
Entry
Ligand
Conv. [%]
Yield [%]
e.r.
1
2
3
4
5
6
7
8
2a·PF
1a·PF
2b·PF
6
6
99
99
99
99
99
12
52
74
84
85
76
86
84
10
46
43
99:1
98:2
98:2
99:1
6
2c·PF
2a·Cl
3a·PF
6
99:1
6
66:34
79.5:20.5
81.5:18.5
3b·PF
3c·PF
6
6
1
[
[
a] Determined by H NMR spectroscopy. [b] NMR spectroscopic yield.
c] Determined by chiral GC analysis.
The copper–NHC catalyst prepared in situ by deprotonation
of 2a·PF (Table 1) with n-butyllithium in presence of copper(I)
6
triflate (1 mol%) promoted efficiently the alkyl addition to
form the desired quaternary carbon center with excellent
regio- (>98% of 1,4 addition) and enantioselectivity (99:1 e.r.).
1
By comparison, the saturated analogue 1a·PF (Scheme 1; R =
6
2
R =Me and R=iBu) behaved very similarly to furnish almost
identical results (Table 2, entry 1 vs. 2). Moreover, under the
1
[
a] Selectivity based on H NMR spectroscopic analysis of the crude reac-
same conditions, ligands 2b·PF , 2c·PF , and 2a·Cl afforded
6
6
tion mixture. [b] Yield of isolated product.
very similar efficiency (entries 3–5), while poor to modest enan-
&
&
Chem. Eur. J. 2014, 20, 1 – 6
2
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!