Organic Letters
Letter
introducing a single palladium source/ligand combination that
has so far been exclusively applied for the copper-free
alkynylation.3,4
Pd-PyMIC (1 mol %), and 1,4-diazabicyclo[2.2.2]octane
(DABCO, 1.4 equiv) as a base in acetonitrile at room
temperature. We aimed to design the reaction conditions that
would enable cross-coupling at room temperature (22 °C). By
screening through the ligands L, only CataCXium A (L7)
provided reasonable conversion to product 3a (20%).
Triphenylphosphine-based ligands L1−L3 promoted conver-
sions of 2a into the products that were tentatively identified as
Based on the mechanistic rationale, herein we present a
novel concept for the design of palladium catalyzed copper-free
alkynylation. It features simultaneous introduction of two
different palladium (pre)catalysts into the reaction mixture,
one tuned to facilitate oxidative addition to aryl halide in Pd1-
Cycle, and another one to activate terminal alkyne in Pd2-Cycle.
Initially, we selected (PhCN)2PdCl2 as a source of palladium
to operate in the Pd1-Cycle and set a brief screening (vide
infra) through a selection of commercially available phosphine-
based ligands shown in Table 1. These Pd/ligand combina-
1
enynes as based on the H NMR resonances appearing in the
remaining unconsumed. This is not surprising because the
formation of enynes from terminal alkynes in the presence of
palladium and bulky phosphine ligands was previously
reported by Trost et al.23 and Colacot et al.10b
As indicated in the systematic investigation by Mårtensson
et al.,28 solvent composition and base are important for
copper-free alkynylation. With L7 as the ligand of choice,
under the same reaction conditions as indicated in Table 1,
1,4-dioxane afforded the highest 35% conversion to 3a over the
other tested solvents: MeCN (affording 20% of 3a), N-
methylpyrrolidone (NMP, 13%), N,N-dimethylformamide
(DMF, 17%), MeOH (15%), i-PrOH (14%), EtOAc (22%),
tetrahydrofuran (THF, 19%), and toluene (<1%). Prolonged
reaction time in 1,4-dioxane from 24 to 72 h increased the
Table 1. Phosphine Ligand Evaluation in Model
Alkynylation Reaction
a
With 1,4-dioxane as the solvent of choice, the effect of the
base was evaluated. Different organic amines (pyrrolidine,
NEt3, t-BuNH2, i-Pr2NEt, Cy2NMe, DBU, DBN, TMG, and
DABCO), organic and inorganic carbonates (KOAc, KOPiv,
K2CO3, Cs2CO3), phosphate (K3PO4), and hydroxide (KOH)
were tested under the reaction conditions from Table 1 (see
could only be achieved with DABCO and K2CO3, respectively.
By prolonging the reaction time to 72 h, K2CO3 (70%
conversion) turned out to be more effective over DABCO
(50% conversion).
Finally, the effect of the palladium source was briefly
evaluated in the reaction between 1a (2 mmol) and 2a (2.8
mmol) under the same reaction conditions as above [L7 (0.08
mmol, 4 mol %), Pd-PyMIC (0.02 mmol, 1 mol %), K2CO3
(2.8 mmol), 1,4-dioxane (1 mL), rt, 24 h] with other Pd2+
complexes (0.04 mmol, 2 mol %) including (PhCN)2PdBr2
(affording 17% yield of 3a), (MeCN)2PdCl2 (7%), Pd(OAc)2
(22%), and Pd(TFA)2 (29%), as well as Pd(dba)2 (10%) as an
example of Pd0 source. Initially selected (PhCN)2PdCl2
S2). The reduction of Pd2+ to Pd0 has been addressed
elsewhere,3,4,10b,16d,22 whereas in the Pd2-Cycle, palladium
remains in Pd2+.
To ascertain whether under the above optimized reaction
conditions the double-palladium manifold indeed plays the
anticipated role in the catalysis from Scheme 2, we conducted
the following test experiments (Table 2). A mixture of 1a, 2a,
and K2CO3 in 1,4-dioxane was exposed to 2 mol % of
(PhCN)2PdCl2, 4 mol % of L7, and 1 mol % of Pd-PyMIC for
24 h, affording 3a in 38% yield (Table 2, entry 1). An excess of
(PhCN)2PdCl2 over Pd-PyMIC (B from Scheme 2) was
employed because the formed is a precatalyst, which must
undergo several transformations before turning into the
catalytically active species A, including Pd2+ to Pd0 reduction,
ligand L7 coordination, and oxidative addition to 1a.
Repeating the reaction in the absence Pd-PyMIC but with
higher 3 mol % loading of (PhCN)2PdCl2 (and 6 mol % of L7)
a
NMR yields are reported as determined from at least two
consecutive runs.
tions have already proven to promote the formation of
catalytically active Pd0 species, subsequent oxidative addition,
and reductive elimination (RE) in a range of cross-
couplings.3,4,10b,16d,22
To build on Pd2-Cycle, we decided to avoid the phosphine-
based palladium complexes. Although their ability to activate
terminal alkynes into acetylides of type B is well-establish-
ed,10b,23 the propensity of phosphines to dissociate and
exchange24 could lead to undesired scrambling between the
reactive palladium species from both cycles, leading to
uncontrolled side reactions or even termination of the process.
Instead, we selected N-heterocyclic carbene (NHC) ligand of a
pyridine (Py) functionalized mesoionic (MIC) structure
(PyMIC),19,25 possessing coordination abilities to a metal
beyond phosphines and even NHCs. B-like NHC acetylide
Pd2+ complexes are well-documented,26 and their formation is
also evident from many Pd-NHC promoted copper-free and
copper cocatalyzed alkynylations.27 All of the above applies to
the Pd-PyMIC complex (Table 1) that has proven to have an
exceptional stability, promoting copper-free alkynylation in hot
water under aerobic conditions.19
For the test reaction, we selected 4-bromotoluene (1a, 1
equiv) and phenylacetylene (2a, 1.4 equiv) as the model
substrates (Table 1). The reaction conditions employed
(PhCN)2PdCl2 (2 mol %), phosphine ligand L (4 mol %),
B
Org. Lett. XXXX, XXX, XXX−XXX