Angewandte
Chemie
Table 2: Substrate scope with propargylic derivatives.[a]
limiting their coordination with the pyridine by-product.
Biarylphosphonite ligands appeared to be promising candi-
dates as they would fulfill the above-mentioned electronic
and steric requirements. The new biarylphosphonite gold(I)
complexes [L2Au]NTf2 and [L3Au]NTf2 were therefore
synthesized by a convenient and efficient four-step procedure
(Scheme 2),[7,12,13] and their catalytic activity evaluated
(Table 1, entries 14 and 15). We were delighted to observe
that, as expected, these complexes exhibit a superior reac-
tivity with increased initial reaction rates. [L3Au]NTf2, from
which L3 was calculated to have a %VBur value of 54.5 (similar
to that of tBuXPhos),[14] was shown to be the most active
catalyst (entry 15).[15] In this case, 6a was completely con-
sumed after only 4 hours and 7a could be isolated in an
enhanced 77% yield. The experimental conditions could be
further improved: working with only 1.2 equivalents of 8e[16]
at 608C furnished 7a with the same 77% yield (entries 16 and
17). It is noteworthy that the combination of
[(Me4tBuXPhos)Au]NTf2 with 8i, which was shown by the
group of Zhang to be optimal for the related oxidative
cyclization of propargyl aryl ethers, was mostly inefficient in
our case (entry 18).[1b]
The reaction scope was next examined using the optimal
catalytic conditions noted in entry 17, of Table 1: 4 mol% of
[L3Au]NTf2 and 1.2 equivalents of 8e in CHCl3 at 608C. We
first focused on the use of aromatic and heteroaromatic
substrates without a subsituent on the propargylic position
(Table 2). A series of these derivatives (6a–m) were smoothly
converted in less than 2 hours into the corresponding cyclized
prodcuts 7a–m, which were isolated in moderate to good
yields (53–77%). This transformation exhibits an interesting
functional group compatibility since various alkyl (6b,l,m),
aryl (6g), ether (6c–e), amine (6 f), and halide (6h–i)
substituents on the aromatic nucleus were tolerated. Inter-
estingly, no noticeable loss of reactivity was observed with 6 f,
bearing a Lewis basic dimethylamino group which susceptible
to coordination with the gold catalyst. It is also noteworthy
that this chemistry could be applied to heteroaromatic
substrates (6k–m), thus representing a useful alternative to
the previously reported hazardous use of a-diazo acetonyl
derivatives.[17] A limitation was found with the chloro
derivative 6i, for which a modified catalytic system must be
used.[18]
[a] Substrate concentration: 0.2m. Yields of isolated products. Reaction
time in parentheses. [b] Reaction run with 8c (2 equiv), MsOH
(1.2 equiv) at 208C. [c] Reaction run at 208C.
Table 3: Optimization of reaction conditions with 6n and 6s.[a]
Entry Substrate
L
8 (equiv)
n
T
t
Yields [%][b]
[8C] [h]
1
2
3
4
5
6n
6n
6n
6n
6n
L3 8e (1.2)
L3 8c (2)
L3 8j (2)
L3 8k (2)
L1 8k (2)
0
60
3
56
–
–
64
–
(62)
(56)
(61)
(67)
(53)
1.2 20 0.5
1.2 20 0.5
1.2 20 0.5
1.2 20
2
6
7
8
9
6s
6s
6n
6s
6s
L3 8e (1.2)
L3 8c (2)
L3 8j (2)
L3 8k (2)
L1 8k (2)
0
60 12
0
–
–
67
–
(0)
We also attempted to react substrates which were either
mono- or disubstituted at the propargylic position. Under the
previously optimized catalytic conditions, the monomethyl-
substituted substrate 6n reacted slowly (3 h) to furnish 7n in
a moderate yield of 56% upon isolation (Table 3, entry 1),
while the dimethyl-substituted substrate 6s did not react at
all. This reduced or even complete loss of reactivity was
tentatively attributed to the bulky 8e, which would disfavor
its addition to the gold-activated alkyne when substituents are
present on the propargylic position (Scheme 1b). We there-
fore decided to employ a less bulky pyridine oxide (2 equiv)
and methanesulfonic acid (1.2 equiv) to trap the pyridine by-
product and therefore disfavor its coordination with the
catalyst.[1c,e,i,p,q] Among the different pyridine oxides screened
(Table 3) 8k, possessing a bromo substituent at the 3-position,
gave the best results both in terms of yield and reaction rate
1.2 60
1.2 60
1.2 60 0.5
1.2 60
6
1
(21)
(59)
(70)
(62)
10
1
[a] Substrate concentration: 0.2m. [b] Yield of isolated product. Yield as
determined by NMR spectroscopy is given within parentheses.
(compare entries 6–9). Gratifyingly, the indan-2-ones 7n and
7s could be thus obtained after only 0.5 hours in 64 (at 208C)
and 67% yield (at 608C), respectively.[19] Again, the phos-
phonite-based [L3Au]NTf2 catalyst gave better results than
Angew. Chem. Int. Ed. 2013, 52, 1 – 7
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!