Organic Letters
Letter
benzotriazoles, and polycyclic systems.13 Two methods of
limited scope have been described for assembling triazenyl-
1,2,3-triazoles, the reaction between alkynyllithiums and aryl
azides14 and the azo coupling of 5-diazo-1,2,3-triazole-4-
carboxamide with amines (Scheme S1).15 Very recently, Cui
and co-workers reported the synthesis of triazenyl function-
alized 1,2,3-triazoles 7 through the Ir(I)-catalyzed cyclo-
addition of azides 5 and internal alkynes 1-alkynyltriazenes 6
(Scheme 1, eq 2).16
Table 1. Optimization of the Cu-Catalyzed Azide−Alkyne−
Azide Reaction
We describe herein an unprecedented chelation-assisted
CuAAC domino reaction whereby aryl azides 8 bearing a polar
X+−O− group (X+−O− = P(O)(NR2)2, P(O)Ph2, SO3H) at
the ortho position undergo a copper-catalyzed cycloaddition to
terminal alkynes 9 followed by in situ electrophilic trapping
with a second equivalent of azide to give fully substituted 1,2,3-
triazoles 10 containing a triazenyl moiety chemo- and
regioselectively (Scheme 1, eq 3).
We initially investigated the experimental conditions of the
copper-catalyzed reaction of N,N,N′,N’-tetraethyl-2-azidophen-
yl-phosphonodiamide (8a) with phenylacetylene (9a). The
parameters evaluated included the screening of copper(I) and
copper(II) salts, loading of reducing agent, solvent, temper-
ature, and reaction time (Table 1).
a
yield (%)
entry
[Cu]/mol %
Solvent
H2O
H2O
t (h) 10a 11a 12a
1
2
3
CuSO4·5H2O/2
CuSO4·5H2O/2
CuSO4·5H2O/5
CuSO4·5H2O/5
CuSO4·5H2O/5
CuSO4·5H2O/5
CuSO4·5H2O/5
CuSO4·5H2O/2
CuSO4·5H2O/5
Cu(OAc)2·1H2O/5
[Cu(ACN)4][PF6]/5
(EtO3P)CuI/5
5
24
1
1
1
1
1
1
0.7
1
12
12
1
1
1
46
46
48
25
75
63
68
37
76
69
62
9
9
9
7
7
6
6
9
7
10
5
9
8
11
0
0
3
2
H2O
46
69
15
2
22
2
10
18
23
86
78
76
86
b
4
H2O/tBuOH
DMF
5
6
7
c
DMF
DMF
DMF
DMF
The reaction in the presence of 2 mol % CuSO4·5H2O and
sodium ascorbate (1 equiv) as the reducing agent at room
temperature in water for 5 h afforded triazenyltriazole 10a as
the major product (NMR yield of 46%, entry 1), together with
minor amounts of the conventional click 1,4-disubstituted
1,2,3-triazole 11a (9%) and the aminotriazole 12a (7%)
derived from 10a (see below and Table S1). Increasing the
reaction time to 24 h had no apparent effect (entry 2).
Completion of the reaction was achieved in 1 h at a loading of
5 mol % CuSO4·5H2O (entry 3). However, the yield of 10a
remained unchanged, whereas that of 11a (46%) increased by
a factor of ca. 5. The use of a mixture H2O/tBuOH 2:1 as
solvent1b further increased the amount of 11a formed at the
expense of 10a (entry 4). Changing the solvent to DMF while
keeping all other variables constant was crucial. In this way,
compound 10a was obtained with 75% NMR yield (entry 5).
Decreasing the amount of acetylene (entry 6), NaAsc (entry
7), and catalyst loading (entry 8) was disadvantageous. A
further improvement in efficiency was achieved by decreasing
the reaction time to 40 min and the amount of solvent to 0.5
mL (entry 9). The reaction provided 10a in 76% NMR yield
(isolated yield of 71%). A decline in the yield of 10a was
observed for lower reaction times (15 and 30 min, data not
shown) and when Cu(OAc)2 (5 mol %, entry 10) was used as
a source of copper. Copper(I) salts such as [Cu(CH3CN)4]-
[PF6] in acetone can be used directly without adding the
reducing agent (entry 11). However, completion of the
reaction required 12 h and large amounts of byproducts 11a
(23%) and 12a (11%) were detected. On the other hand,
copper(I) halides favored the predominant ((EtO3P)CuI in
acetone, entry 12) or exclusive ((Ph3P)CuBr in DCM, entry
13) formation of the click product 11a with yields of ca. 80%.
Triazole 11a also became the major product for the CuSO4·
5H2O/NaAsc catalytic system when large loadings of copper-
(II) were used (0.3 to 1 equiv, entries 14 and 15). In summary,
5-triazenyl-1,2,3-triazole 10a was obtained in high yield (71%)
when the reaction was carried out in DMF with CuSO4 (5 mol
%) and sodium ascorbate (1 equiv) at room temperature under
open air for 40 min. The same performance was observed on a
gram-scale synthesis (yield of 70%).
d
8
9
e
10
11
12
13
DMF
f
Acetone
Acetone
CH2Cl2
DMF
f
f
(Ph3P)CuBr/5
CuSO4·5H2O/30
CuSO4·5H2O/100
0
15
10
14
15
DMF
a
NMR yield determined through integration of the 1H and 31P NMR
spectrum of the crude reaction mixture. The remaining material up to
100%, if any, was the starting reagent 8a. In some cases, traces of
b
other minor compounds <3% were detected. H2O/tBuOH 2:1 (v).
c
d
e
1 equiv of 9a was used. 60 mol % of NaAsc was used. The amount
of solvent was reduced to 0.5 mL, and the isolated yield of 10a was
f
71%. No ascorbate was used.
With the optimized reaction conditions in hand, the scope of
alkynes was explored. The reaction showed high tolerance of
the electronic nature of the substituents in the acetylene
component (Scheme 2). Triazenyltriazoles 10a−l were
obtained in moderate to good yield together with small
amounts of the byproducts 11 and 12 (Table S1). Compounds
10 were purified through column chromatography. An extra
step of precipitation in DCM/hex was required for some of
them. Acetylenes with R1 = 2-OMeC6H4, OEt, C4H9, TMS,
and CO2Me reacted more slowly requiring reaction times ≥16
h for achieving good yields. The apparent low yield of 10e is
due, to a significant extent, to its partial in situ transformation
into the aminotriazole 12e (33%, Table S1). Compound 10b
was characterized through single-crystal X-ray diffraction
analysis (Scheme 2 and Figure S8). Importantly, the reaction
of 4 equiv of 8a with 1,3-diethynylbenzene provided a mixture
of the bis(triazenyltriazole) 10m, the mixed triazenyltriazole-
triazole 13, and the mono adduct 10l in a ratio 47:16:37
(Scheme S2). Purification through column chromatography
afforded 10m in a yield of 28%.
To extend further the scope of the reaction, the ability of
other functional groups for promoting this domino trans-
formation was tested using phenylacetylene 9a as the click
B
Org. Lett. XXXX, XXX, XXX−XXX