Zhang et al. Sci China Chem
3
verts to an imidate which undergoes Mumm rearrangement
to form an α-acylamino amide. Such a transformation is the
prototypical four-component Ugi reaction (Ugi-4CR) [24].
Alternatively, when trapped by water, the nitrilium ion
coverts to an α-amino amide derivative. Such a process is
commonly named as three-component Ugi reaction (Ugi-
proceeds with acid catalysis by lowering the lowest un-
occupied molecular orbital (LUMO) energies of the imines
[25,41,42]. Therefore, we envision that a chiral phosphoric
acid (CPA) [43,44] might enable aliphatic aldehydes to
function as feasible substrates for enantioselective Ugi-3CR.
Rapid imine formation, as well as its preferential activation
over the carbonyl group, are viable with a CPA catalyst to
suppress competing reactions such as the Passerini-type re-
action [29]. Suitable interactions between a bulky CPA and
the small substrates may lead to the improvement of en-
antioselectivity for the less sterically hindered aldehydes
[45].
3
CR) which was firstly developed by List et al. [25].
In light of the importance of α-amino amides and the ad-
vantages of MCRs, the pursuit of a mild, simple, highly
enantioselective Ugi-3CR with commercially available
starting materials is highly desirable (Figure 1(f), left). The
structures of the α-amino amides formed via Ugi-3CR are
more tunable compared to those made by the conventional
amidation strategy, as the former is based on three achiral
building blocks while the latter relies on the ready-made
chiral α-amino acids. In addition, enantioselective Ugi-3CR
offers facile delivery of distinct stereoisomers by judicious
choice of the catalyst configuration. The diversity of struc-
tures and the controllability of stereochemistry are necessary
for biological activity studies [26], drug development [27]
and catalyst modifications [28].
2 Results and discussion
online), we set out to optimize the model catalytic asym-
metric Ugi-3CR of 2-benzyloxy acetaldehyde (1a), 4-ni-
troaniline (2a), and cyclohexyl isocyanide (3a) in CH Cl at
2
2
−30 °C (Table 1). The BINOL-derived CPA1 and CPA2
gave moderate yields and low enantioselectivities (Table 1,
entries 1, 2). By tuning the substituents on 3,3′ positions of
the BINOL skeleton, we found that more steric hindrance led
to the improvement of enantioselectivity. CPA3 with 2,4,6-
triisopropylphenyl substituents provided a 40% enantiomeric
excess (ee) (entry 3), and bulkier 2,4,6-tricyclohexylphenyl
substituted CPA4 boosted the enantioselectivity to 81% ee
(entry 4). The CPA with 9-anthryl as substituents (CPA5)
catalyzed the reaction in 24% yield with 7% ee (entry 5).
Then we turned to the SPINOL-derived CPAs. The negli-
gible difference in enantioselectivity was detected for CPA6
and CPA1 despite the different chiral skeletons (entries 1, 6).
A slight increase in enantioselectivity was found when cat-
alyst changes from CPA2 to CPA7 (entries 2, 7). Notably, a
dramatic improvement was observed when CPA8 (88% ee)
was employed (entry 8). The enantiocontrol was further
enhanced (97% ee) when the reaction was facilitated by
CPA10; meanwhile, the best yield (72%) was achieved
among all the catalysts (entries 1–10). Attempts to further
improve the chemical yield upon evaluation of various sol-
vents (entries 11–14) and temperatures (entries 15, 16) met
with failure. Finally, the yield increased to 91% (entry 17)
when excess aldehyde was added and the reaction time was
prolonged.
Despite some advances have been made in en-
antioselective α-addition of isocyanides [29–31], the devel-
opment of catalytic asymmetric Ugi reactions is greatly
lagged due to the complex reaction system, the competing
reactions, and the difficulty in the enantiocontrol of the α-
addition of isocyanides. A catalytic enantioselective Ugi-
type reaction was firstly introduced in 2009 by Zhu and co-
workers [32,33]. After that, only limited examples of en-
antioselective Ugi reactions have been reported. Maruoka
and co-workers [34] developed a chiral dicarboxylic acid-
catalyzed Ugi-type reaction with acyclic azomethine imines.
Wang, Zhu, and co-workers [35] reported a catalytic asym-
metric four-center three-component Ugi reaction by dynamic
kinetic resolution of the primary multicomponent adduct. In
all cases mentioned above, transient nitrilium intermediates
were trapped by internal oxygens. Particularly worth to
mention is that Wulff and co-workers [36] successfully de-
veloped an asymmetric Ugi-3CR enabled by an elegantly
designed BOROX catalyst that was assembled in situ from a
chiral biaryl ligand, an amine, water, BH ·SMe , and an al-
cohol (or phenol) to afford α-amino amides. Nonetheless,
this enantioselective Ugi-3CR protocol is restricted to aro-
matic aldehydes, and the extension to aliphatic aldehydes
remains challenging. Recently, our group [37] reported a
catalytic asymmetric Ugi-4CR in which direct enantiocontrol
of the α-addition of isocyanides with activated imines as
electrophiles and external carboxylic acids as nucleophiles
was fulfilled. As our continuous research interest in asym-
metric MCRs [37–40], we aimed to develop an asymmetric
Ugi-3CR with complementary substrate scope, as aliphatic
aldehydes derived α-amino amides are more common in
naturally occurring compounds and drugs. Usually, Ugi-3CR
3
2
Having established the optimum conditions, the scope and
limitation of this reaction were evaluated by exploring dif-
ferent reaction components (Table 2). The reaction was ap-
plicable well to an array of linear aliphatic aldehydes which
gave moderate enantioselectivity in previous Ugi-type re-
action [32]. For acetaldehyde (5), a very small substrate, an
84% yield with 87% ee was obtained. Besides, the length of
the substitutions on the aldehydes was well-tolerated, as