.
Angewandte
Communications
in pharmaceuticals and in organic synthesis, and have
attracted chemists for many years.[12] However, efficient and
general catalytic methods are still under intensive develop-
ment.[13–15]
The reaction between tert-butylisonitrile (1a) and N-
benzyl-1,3-diphenylprop-2-yl-1-amine (2a)[16] was initially
investigated by varying the metal salts, the solvents, and the
temperature (Scheme 2). Key observations are summarized
Scheme 2. Multicatalysis: Synthesis of 1,4,5-trisubstituted imidazoles.
as follows: a) Y(OTf)3, La(OTf)3, and Yb(OTf)3 were effec-
tive Lewis acids to catalyze the insertion of isocyanide to the
À
N H bond of the secondary amine, with Yb(OTf)3 being the
best, to afford the amidine 5a in a quantitative yield.
Interestingly, other metal salts such as CuCl, AgOAc,
AgOTf, NaAuCl3, which are known to catalyze/promote the
amine insertion to isonitriles,[17] failed to produce 5a,
probably because of the preferential coordination of these
p-acidic metal salts to the triple bond, thus inhibiting the
insertion reaction; b) Yb(OTf)3 is an ineffective catalyst for
the conversion of amidine 5a into 3a. Silver salts, especially
AgOTf, catalyzed the 5-exo-dig cyclization of 5a leading to 3a
in a quantitative yield;[18] c) Yb(OTf)3 and AgOTf are
compatible and the two catalytic cycles did not interfere
with each other. Therefore in practice, there was neither need
to isolate the amidine intermediate nor to introduce the two
catalysts sequentially;[19] d) Xylenes gave the best result
among solvents screened (THF, 1,4-dioxane, DMF, DCE,
toluene and xylene). Overall, the all-in-one protocol consisted
of heating a xylenes solution (1308C) of 1a and 2a in the
presence of Yb(OTf)3 (0.2 equiv) and AgOTf (0.1 equiv).
Under these reaction conditions, the imidazole 3a, whose
structure was confirmed by X-ray structural analysis, was
isolated in 99% yield.[20]
The scope of this imidazole synthesis turned out to be very
general (Scheme 3). Aryl groups with different electronic
properties including heterocycles (3e–h) and alkyl groups can
be introduced into the 1-, 4-, and 5-positions of imidazoles.
The reaction was not sensitive to steric hindrance as
compound 3q (R3 = 2,4,6-trimethylbenzyl) can be obtained
in 90% yield. Chiral propargylamines can be converted into
enantioenriched imidazoles without racemization as is evi-
denced by the synthesis of the two enantiomers 3v and 3w.
The 1-benzyl-5,6-dihydroimidazo[5,1-a]isoquinoline 3x, an
important bicyclic compound in medicinal chemistry,[21] is
readily synthesized from the corresponding cyclic secondary
amine. It is interesting to note that when N-aryl propargyl-
amines (R3 = Ar, 3r–u) were used as substrates, the compet-
itive direct cyclization of 3r–u leading to quinolines was not
observed.[22]
Scheme 3. Scope of reaction. [a] Reaction was performed at 1408C.
Scheme 4. Further structural elaboration of functionalized imidazole.
The so-obtained imidazole could be further elaborated to
more complex polycyclic structure (Scheme 4). For example,
heating a toluene solution of 3u in the presence of a catalytic
amount of Pd(OAc)2, PPh3, and Cs2CO3 afforded 1-
benzylimidazo[1,5-f]phenanthridine (6) in 88% yield.[20,23]
It is known that isonitrile can form a complex with both
YbIII[24] and AgI[25] salts. In addition, we found that the pre-
synthesized complex Ag(OTf)(tBuNC)n has the same cata-
lytic activity as AgOTf.[26] It is therefore reasonable to assume
that both Yb(OTf)3 and AgOTf, upon mixing with isocyanide,
will form the complexes 7 and 8, respectively (Scheme 5).
Ytterbium triflate, which is azophilic, would further coordi-
nate to the amine,[27] while silver triflate, which is more p-
acidic,[28] would prefer to coordinate to the triple bond to
produce 9. Migrative insertion from 9 would produce 10 and
TfOH. Salt metathesis between 10 and TfOH would produce
the amidine 11 with the concurrent regeneration of Yb(OTf)3.
Intramolecular nucleophilic addition of the amidine nitrogen
atom to the silver-coordinated triple bond through a 5-exo-dig
mode afforded the intermediate 12, which upon dealkylation
(R = tBu), would produce 13 and TfOH.[29] Protonation of 13
followed by double-bond isomerization would produce 3 with
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!