.
Angewandte
Communications
essential factor. We therefore decided to examine other
hypervalent iodine reagents for the reaction. Indeed, the
domino process was more efficient when using bistrifluoro-
methyl benziodoxole reagent 10,[11a] which gave 50% of 11a
together with 14% of 2-phenyl furan (6g), resulting probably
from protonation of the putative organogold intermediate
(Table 2, entry 7). The fact that no 2-alkynylation product was
obtained in this case is probably due to the lower efficiency of
reagent 10 in direct C–H alkynylation.[9d,14] In contrast, no
product was observed with the dimethyl-substituted reagent
13 or alkynyl iodonium salt 14, with a more or less basic
oxygen atom bound to the iodine, respectively (Table 2,
entries 8 and 9). On the other hand, no product was observed
when alkynyl iodide 15 was used (Table 2, entry 10). These
results further demonstrated that success in this challenging
domino process can be achieved only for a very narrow
window of electronic density at the iodine atom. In presence
of sodium carbonate, the formation of 2-phenyl furan (6g)
was suppressed, albeit at the cost of the yield of 11a (Table 2,
entry 11). Fortunately in this case, a larger excess of hyper-
valent iodine reagent 10 allowed increasing the yield sub-
stantially to 73% (Table 2, entry 12). Finally, reinvestigation
of the solvent led to the use of isopropanol, for which the 3-
alkynylated product 11a was obtained in 93% yield deter-
mined by using GC (Table 2, entries 13–16). On a 0.3 mmol
scale, furan 11a could finally be isolated in 68% yield
(Table 3, entry 1).
nearly quantitative yields (94–95%), independently of its
position (Table 3, entries 2–4). A para-methoxy or a phenyl
group were also well-tolerated in the reaction (Table 3,
entries 5 and 6), as well as a furyl substituent (entry 7). A
current limitation of the method is that electron-withdrawing
groups were not tolerated, because in this case Michael
addition of the solvent on the allene was observed in
isopropanol and decomposition was obtained in other sol-
vents. The reaction was not limited to aromatic substituents,
and excellent yields were obtained with both primary
(Table 3, entries 8–9) and a secondary (entry 10) aliphatic
substituent. A more sensitive benzyl group was also well
tolerated, giving the alkynylation product 11k in 77% yield
(Table 3, entry 11). Finally, an important preliminary result
was obtained for the synthesis of polysubstituted furans:
Starting from allene ketone 16, trisubstituted furan 17 was
obtained in 94% yield (Table 3, entry 12). The synthesis of
such a product would be very challenging through C–H
functionalization, owing to serious issues of reactivity and
regioselectivity.
The obtained silylated alkynyl furan 8a was easily
deprotected to give the corresponding free acetylene. By
using a methodology developed previously in our group,[9b]
one-pot dimerization and thiophene formation then resulted
in the formation of alternating heterocyclic oligomer 18
(Scheme 3, (1)). Heterocyclic oligomers are important in
Investigation of the scope of the reaction showed that
substitution of the benzene ring by a methyl group led to
Table 3: Scope of the domino cyclization/alkynylation.
Scheme 3. Functionalization of alkynyl furans 8a and 8b. Reaction
conditions: a) tetrabutylammonium fluoride, THF, 08C, 1 h, 78%;
b) Cu(OAc)2, CH3CN, 808C, 12 h; then Na2S·3H2O, 808C, 24 h, 65%;
c) 10 wt% Pd/C, H2, pentane/EtOH, 12 h, 74%, 5:1 d.r.
Entry
Substrate
Product
Yield [%][a]
organic materials, but are usually composed of a single class of
heterocycles. New properties can be expected to emerge with
an easier access to more-complex oligomers. Furthermore,
hydrogenation gave access to tetrahydrofurans, which are
important building blocks for the synthesis of bioactive
synthetic and natural products (Scheme 3, (2)).
The results obtained for the C2-alkynylation of furans are
in line with our previous work on the alkynylation of
heterocycles.[9a–d] In particular, we had shown that AuCl
reacts instantaneously with TIPS-EBX (7) to form bis(triiso-
propylsilyl)diyne as major product. This led us to propose
a catalytic cycle involving first oxidative addition of a AuI
species on the reagent, followed by C–H auration and
reductive elimination as one of the most probable mecha-
nisms for this transformation. In this context, the fact that the
new domino cyclization/alkynylation process is catalyzed by
a AuIII catalyst is mechanistically intriguing. In fact, the only
domino cyclization/alkynylation process reported to date has
been proposed to proceed through a AuI–AuIII catalytic
cycle.[8d] To explain our results, an unprecedented electro-
1
2
3
4
5
6
7
8
R1 =Phenyl 9a
R1 =4-MeC6H4 9b
R1 =3-MeC6H4 9c
R1 =2-MeC6H4 9d
R1 =4-MeOC6H4 9e
R1 =4-PhC6H4 9 f
R1 =2-furyl 9g
11a
11b
11c
11d
11e
11 f
11g
11h
11i
68
94
95
94
83
53
75
93
96
97
77
R1 =hexyl 9h
9
10
11
R1 =ethyl 9i
R1 =cyclohexyl 9j
R1 =benzyl 9k
11j
11k
12
94
[a] Reaction conditions: 9a (0.3 mmol), Au catalyst 12 (5 mol%),
isopropanol (15 mL), RT, 5–72 h, yields of isolated products after column
chromatography are given.
4
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1 – 6
These are not the final page numbers!