.
Angewandte
Communications
carried out further investigation on this interesting elemen-
tary step. To prove directly the formation of CO in the
catalytic cycle, we ran the reaction in a closed vial with septum
and injected samples of the gas phase into a GC/MS. At first,
we used 5 equivalents of H2(18O) as nucleophile and indeed
we were able to detect the formation of a peak at m/z 30
corresponding to C(18O) (Figure 1). The simultaneous growth
of the peak at m/z 28 is probably due to diffusion of nitrogen
but could also arise in some part from the formation of C(16O)
out of residual non-labelled H2(16O). Nevertheless, this
experiment clearly proves the crucial role of water for the
reaction and shows that it is really the oxygen atom of the
latter that is incorporated into the eliminated carbon mon-
oxide.
Scheme 2. Dual gold-catalyzed transformation of 1a in the presence of
1 equiv of water (IPr=1,3-bis(2,6-diisopropylphenyl)imidazol-2-ylidene,
Tf =trifluoromethylsulfonyl).
The loss of one carbon atom was unambiguously proven by
a crystal structure analysis.[10,11]
Our mechanistic hypothesis for this unexpected product is
indeed based on the formation of a gold acyl complex and its
subsequent chemical transformation. We assume that, in
accordance to other reported dual gold-catalyzed transfor-
mations, the generation of gold vinylidene intermediate Ib
initiates the reaction cascade (Scheme 3). Ib can be trapped
with water as a nucleophile leading to species VI. Proto-
demetalation of the sp2-Au species and transfer of the oxygen-
bounded proton onto the conjugated system then led to gold
acyl species VIII. By extrusion of CO,[12] intermediate IX can
be formed which, after protodeauration or catalyst transfer
onto the next molecule of the starting material, completes the
catalytic cycle. Through this pathway two hydrogen atoms
from water would be added to the product, whereas the
oxygen atom together with one carbon atom from the starting
material would be eliminated in the form of CO. In total, this
would account for the observed mass loss of 10 u.
Next, we synthesized substrate [13C]-1m bearing a 13C-
label at the terminal alkyne position and followed its trans-
formation in the presence of 5 equivalents of H2(16O) in
a similar fashion (Figure 2). This time, we detected the for-
mation of a peak at m/z 29 corresponding to 13CO. Again, the
peak at m/z 28 was growing as well. To check, whether this
growth is solely due to diffusion of nitrogen or whether it
contains a contribution from 12CO, we isolated product 2m. If
12CO was produced during the reaction, the 13C-label should
still be present in the product. Since this was not the case
(neither the 13C NMR nor an MS spectrum of 2m showed any
indication of a 13C-label), this experiment clearly proves that
selectively the former terminal alkyne carbon and therefore
the vinylidene/gold acyl carbon is eliminated as CO during
the reaction and that the peak at m/z 28 can be allocated to
N2.
Since, to the best of our knowledge, no precedents for
a CO extrusion in homogeneous gold catalysis exist, we
After the proof of the postulated CO extrusion we set out
to investigate the proton transfer steps of the reaction cascade
with deuterium labeling experiments.[11]
Unfortunately, the presence of water,
which could be shown to be essential
for the reaction, led to an undesired
exchange at the terminal alkyne posi-
tion of the starting material under the
reaction conditions (confirming the
involvement of
a gold acetylide).
Therefore, no conclusions can be
drawn from these experiments and
a differentiation whether the catalytic
cycle is closed by a catalyst transfer step
or by protodeauration is not possible
either.
Next, we synthesized substrate 1b
in which the symmetry of the backbone
is broken by the introduction of an
extra methyl group. Its gold-catalyzed
transformation in the presence of
5 equivalents of D2O led to a 1.0:0.8
mixture of the two isomers 2ba and 2bb
that were assigned with the help of an
HMBC NMR spectrum (Scheme 4).
The detection of these two isomers
suggests that the displacement of the
gold catalyst in the allylic fragment of
IX can happen either directly or by an
Scheme 3. Proposed mechanism for the gold-catalyzed formation of indenes from diyne systems
in the presence of water. Counterions are omitted for clarity.
2
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!