Full Paper
hence, small or electron-rich alkynes favor a double insertion
pathway. Unsurprisingly, with unsymmetrical dialkylated acety-
lenes, a mixture of allene products was obtained with little se-
lectivity arising from the 1,2-insertion (2p and 2q in 0.7:1
ratio). In an attempt to control the selectivity of the insertion,
1-methoxyoct-2-yne gave allene 2s selectively, in line with the
1,2-insertion of an electron-rich alkyne on an electrophilic pal-
ladium species. Further extension of this methodology to relat-
ed monoaryl or ester derivatives proved impossible due to the
poor stability of the alkyne precursors.[9] Nevertheless, replac-
ing the -OMe group by a -OTBDMS or -OTHP moiety permitted
access to more advantageous allene products (2t and 2u). To
demonstrate this, a two-step deprotection/gold-catalyzed cycli-
zation was conducted starting from 2u via the allenol deriva-
tive 2v to afford the 2,5-dehydrofuran 2v’ in a promising un-
optimized yield of 30% (Scheme 5).[10]
role in controlling the cyclization reaction.[14] Consequently, in-
troduction of O- and N-Tos substituents within the ring (4c
and 4d) afforded the allene products in good yields.[15]
Mechanistic insights
As seen previously, to achieve a selective reaction, well-tuned
conditions (i.e., Pd source, phosphine ligand, base, additive,
solvent) are required to avoid alternative reaction pathways,
also observed by Heck et al., leading to products such as 2l’
and 2m’.[3]
Whereas DFOTP presumably provides the LnPd catalytic spe-
cies with the optimum combined steric and electronic factors,
we wondered what the exact role of PivOH additive in this
transformation might be (Table 1, entries 3 and 4). To answer
this question, a stepwise mechanistic investigation of the reac-
tion was performed.
As commonly accepted, it is likely that the reaction starts
with the formation of Pd0Ln followed by oxidative addition of
ArBr to generate an ArPd0BrLn intermediate. Starting from
[Pd2(dba)3] and DFOTP (2 equiv), we were pleased to isolate
the [Pd(DFTOP)2(dba)] complex A1. However, because the sep-
aration of A1 from free dba proved difficult, we envisaged
Scheme 5. Stepwise deprotection and gold-catalyzed cyclization.
a
more convenient synthesis starting from [Pd(COD)-
(CH2SiMe3)2], giving A1 as a yellow solid (91%). Complex A1
was then engaged in the oxidative addition with o-tolyl bro-
mide to afford the s-aryl-palladium(II) halide [trans-TolPdBr-
(DFOTP)2] A2 (74%), which could be characterized by X-ray dif-
fraction analysis (Scheme 7).[8,16] As expected, A1 and A2
proved to be efficient catalysts for this transformation when
submitted to standard reaction conditions, suggesting that
such species could be formed at the early stage of the reaction
and participate in the catalytic process.
Intramolecular reaction
We next envisaged an intramolecular approach leading to
a family of allene products that are otherwise difficult to
obtain, starting from easily accessible starting material
(Scheme 6).[11,12] With diester 3a, the cyclization proved unsuc-
cessful, giving less than 5% yield of the desired product, and
resulting in mostly unreacted starting material.[13] Substitution
of the ester groups by ethers in 3b permitted allene 4b to be
obtained in good yield while running the reaction at lower
temperature, indicating that the nature of the linkage located
between the aryl and the alkynyl moiety plays an important
We then wondered whether A2 would evolve through a clas-
sical Heck-type mechanism, hence through insertion of alkyne
into the PdÀaryl bond, followed by b-hydride elimination to
generate a HPdBrLn species. To probe this postulate, A2 was re-
acted with 1.2 equivalents of 4-octyne (Scheme 8) in a Young-
NMR tube following the formation of 2b from its characteristic
apparent septuplet signals at 5.15 ppm by 1H NMR spectro-
scopic analysis. Whereas no reaction was observed at room
Scheme 6. Intramolecular reaction. Reaction conditions: [Pd2(dba)3]
(2.5 mol%), L1 (10 mol%), Cs2CO3 (1.5 equiv), PivOH (30 mol%), toluene,
1008C, 2h.
Scheme 7. Catalytic activity of A1 and A2. [a] Thermal ellipsoids at 50%
probability.
Chem. Eur. J. 2014, 20, 13272 – 13278
13275
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim