Nickel-catalyzed direct alkylation of terminal alkynes at room temperature: A hemilabile pincer ligand enhances catalytic activity

Article abstract of DOI:10.1021/cs501502u

Direct coupling of alkyl halides with terminal alkynes provides an efficient and streamlined access to alkyl-substituted alkynes, which are important synthetic intermediates, biologically active molecules, and organic materials. However, until now,there h

Full text of DOI:10.1021/cs501502u

Research Article
pubs.acs.org/acscatalysis
Nickel-Catalyzed Direct Alkylation of Terminal Alkynes at Room
Temperature: A Hemilabile Pincer Ligand Enhances Catalytic Activity
Pablo M. Perez García, Peng Ren,§ Rosario Scopelliti, and Xile Hu*
́
́
Laboratory of Inorganic Synthesis and Catalysis, Institute of Chemical Sciences and Engineering, Ecole Polytechnique Fed
Lausanne (EPFL), EPFL-ISIC-LSCI, BCH 3305, Lausanne, CH 1015, Switzerland
́ ́
erale de
S
* Supporting Information
ABSTRACT: Direct coupling of alkyl halides with terminal
alkynes provides an efficient and streamlined access to alkyl-
substituted alkynes, which are important synthetic intermediates,
biologically active molecules, and organic materials. However, until
now,there have been fewer than a handful of catalytic methods
available for this reaction, and detailed mechanistic studies have not
been reported. Herein, we describe the design and development a
new nickel pincer complex that catalyzes the direct coupling of
primary alkyl halides with terminal alkynes at room temperature.
The catalysis has a good substrate scope and high functional group
tolerance. Kinetic data suggest that the new pincer ligand is
hemilabile, and the dissociation of a labile amine donor is the
turnover-determining step of the catalysis. An intermediate Ni-alkynyl species has been isolated and structurally characterized.
The reactivity of this species gives insight into the nature of the active species for the activation of alkyl halide.
KEYWORDS: nickel, Sonogashira coupling, alkynylation, pincer complex, kinetics, alkylation
1. INTRODUCTION
Alkynes are ubiquitous synthetic intermediates and precursors
to biologically active molecules and organic materials;1,2
therefore, methods that provide a streamlined access to alkynes
are highly desirable. The synthesis of aryl- and alkenyl-
substituted alkynes has become routine, largely thanks to the
development of Sonogashira coupling of aryl and alkenyl
halides with terminal alkynes;3−6 however, the synthesis of
alkyl-substituted alkynes is less straightforward. The traditional
method of reacting alkali metal acetylides, in particular lithium
acetylides, with alkyl electrophiles has a limited scope, is
intolerant to sensitive functional groups, and needs a low
temperature. Organometallic coupling reactions of alkynyl
halides with metal−alkyl reagents7,8 or alkyl halides with
metal-alkynyl reagents9−12 are more general and tolerant, but
they require the preactivation of alkynes, a reactive organo-
metallic reagent, or both. Compared with these methods, the
direct coupling of alkyl halides with terminal alkynes involves
fewer steps, is operationally simpler, and often costs less. On
the other hand, such coupling is challenging not only because
of the general difficulty in the coupling of alkyl electrophilies,
but also because of a low concentration of metal alkynyl
intermediate in the reaction mixture.
Until now, there are only four metal−ligand combinations
that are capable of catalyzing the direct alkylation of terminal
alkynes with nonactivated alky electrophiles. The groups of
Fu13 and Glorius14 introduced the first two systems (1 and 2,
Figure 1) both of which were based on Pd and a N-heterocyclic
carbene ligand. Our group reported the first Ni catalyst for this
Figure 1. Reported catalyst systems/precatalysts for direct alkylation
of alkynes.
transformation:15 the catalyst is a nickel pincer complex,
Nickamine (3). The group of Liu16 then developed a system
based on Ni(cod)2 and a pyridine bisoxazoline (Pybox) ligand
(4). This system is the most active catalyst to date, and it
catalyzes the coupling of both primary and secondary
nonactivated alky halides at room temperature.
Despite the progress in method development, the mecha-
nism of the coupling of alkyl halides with terminal alkynes
remains largely unclear. The catalysis by Nickamine seems
appropriate for in-depth mechanistic investigations because the
Received: October 1, 2014
Revised: January 9, 2015
© XXXX American Chemical Society
1164
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171

ACS Catalysis
Research Article
catalyst is well-defined. In comparison, the coordination
environment of the active catalyst in the Liu system is
unidentified.16 Nevertheless, Nickamine catalyzes the coupling
only at an elevated temperature (100−140 °C) which is
inconvenient for both synthetic applications and mechanistic
study. Herein, we describe the development of a new nickel
pincer complex that catalyzed the direct alkylation of terminal
alkynes at room temperature. The new catalytic system was
subject to a kinetics study, which revealed the hemilabile nature
of the new pincer ligand as a key factor for the improved
catalytic efficiency. A catalytically relevant Ni alkynyl complex
was synthesized and structurally characterized. The reactivity of
this Ni alkynyl complex gave important insights into the active
species for the activation of alkyl halide.
Figure 2. Structural formula for complexes 5 and 6.
replaced by a simple ethylene linker. The new NNN pincer
ligand was expected to be more labile, which rendered the
resulting Ni complex more reactive. This ligand and its Ni
complex were synthesized in a manner similar to that of the
N2N ligand and Nickamine.18 The crystal structure of 6 was
determined, showing Ni in the expected square planar ligand
environment (Figure 3). The Ni−N and Ni−Cl bond distances
in 6 are nearly identical to those in 3,19 except that the Ni1−N1
distance is 0.04 Å longer in 6.
2. RESULTS AND DISCUSSION
2.1. A New Nickel Pincer Catalyst for Direct Alkylation
of Alkynes. The coupling between 1-iodooctane and 1-octyne
was used as the test reaction for the optimization of reaction
conditions. It was reported that Nickamine catalyzed this
reaction efficiently at 100 °C.15 The reaction was best carried
out using 5 mol % 3 as catalyst, 3 mol % of CuI as the
cocatalyst, 1.4 equiv of Cs2CO3 as the base, and dioxane as the
solvent. The reaction time was 16 h. We found that replacing
Cs2CO3 with LiOtBu led to successful coupling at 60 °C (Table
S1, Supporting Information); however, the yield of the reaction
decreased to only 10% at 50 °C (entry1, Table 1). The use of
Table 1. Optimization of Conditions for Ni-Catalyzed Direct
a
Coupling of 1-Iodooctane with 1-Octyne
CuI
base (1.4
equiv)
temp
(°C)
yield
b
entry (mol %)
solvent catalyst
(%)
1
2
3
4
5
6
7
8
9
LiOtBu
LiOtBu
LiOtBu
LiOtBu
LiOtBu
LiOtBu
LiOtBu
Cs2CO3
dioxane
MeCN
DMF
DMF
DMF
DMF
DMF
DMF
3
3
3
5
6
6
6
6
50
20
20
20
20
20
20
20
10
8
9
Figure 3. Crystal structure of complex 6. Only one of the two
molecules in the asymmetric unit is shown. Hydrogen atoms are
omitted for clarity. The thermal ellipsoids are displayed in a 50%
probability. Selected lengths (Å) and angles (deg): Ni1−N1, 1.995(3);
Ni1−N2, 1.832(3); Ni2−N3, 1.962(4); Ni1−Cl1, 2.2102(11); N1−
Ni1−N2, 85.60(14); N2−Ni1−N3, 84.74(13); N3−Ni1−N1,
169.04(13).
9
43
13
74
13
74
10
9
9
0
4.5
9
a
General reaction conditions: 0.5 mmol of 1-iodooctane and other
b
reagents according to Table 1 in 2 mL of solvent. Yields are
determined by GC and are relative to alkyl halide.
To our delight, complex 6 is, indeed, a better catalyst than
Nickamine for the reaction in Table 1. At room temperature, a
coupling yield of 74% was obtained when 6 was used as the
catalyst (entry 5, Table 1). CuI was important for the coupling.
Without CuI, the yield was only 13% (entry 6, Table 1). The
loading of CuI could be decreased to 4.5 mol % while
maintaining the same yield (entry 7, Table 1). For convenience
in weighting, a 9 mol % loading of CuI was applied in small-
scale reactions. LiOtBu was the best base; use of Cs2CO3 as the
base decreased the yield to 10% (entry 8, Table 1). Under the
optimized conditions (entry 5 or 7, Table 1), the main side
products were 1-tert-butoxyoctane and 1-octene. The former
was produced by the attack of tert-butoxide anion on 1-
iodooctane, whereas the latter was produced by base-induced
HI elimination of octyl iodide or reductive elimination from a
Ni-octyl intermediate. The excess amount of 1-octyne was
unreacted.
more-polar solvents, such as acetonitrile (MeCN) or
dimethylformamide (DMF), allowed the coupling at room
temperature, but the best yield was 43% in DMF (entries 2 and
3, Table 1).
Complex 5 (Figure 2) was previously shown to have
improved efficiency for the coupling of secondary alkyl halides
with alkyl Grignard reagents;17 however, the coupling yield was
only 15% when 5 was used as catalyst (entry 4, Table 1). At the
same time, heterogeneous black particles were observed in the
reaction using 5 as catalyst. We suspected that complex 5,
having only a bidentate chelate, was unstable under the reaction
conditions and decomposed to nickel particles.
To maintain the stability of the Ni complex while potentially
increasing its activity, complex 6 was designed and synthesized.
One of the two aryl linkers in the N2N ligand of Nickamine was
1165
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171

ACS Catalysis
Research Article
2.2. Scope of Direct Alkylation of Alkynes. The
optimized conditions in Table 1 were then applied for the
coupling of more elaborate substrates (Table 2). For some
alkylation of secondary alkyl halides, which seems to be a
general limitation of the Ni(II) pincer systems. Overall, the
scope and yields for the alkynylation of primary alkyl halides are
comparable to those catalyzed by the Ni(cod)2/Pybox system
developed by Liu and co-workers.16 The current system,
therefore, is well suited for an in-depth mechanistic study
because of the defined nature of the catalyst and the mild
reaction conditions.
Table 2. Scope of Direct Alkylation of Alkynes Using 6 as
a
Catalyst
2.3. Kinetics of Direct Alkylation of Alkynes. The
reaction profile of the coupling of 1-iodooctane with 1-octyne
was measured by gas chromatography (GC). Figure 4 shows
Figure 4. Reaction profile for the coupling of 1-iodooctane with 1-
octyne. At the end of this period, the conversion of 1-iodooctane is
72%, the conversion of 1-octyne is 39%, and the yield of hexadec-7-yne
is 43%.
the conversion of reagent and formation of production over a
period of 3 h. The formation of product followed a linear line,
indicating a constant reaction rate. Because of the detection
limit of the FID detector of GC, only the data collected after 20
min could be reliably used. This gave the false impression of an
induction period in the reaction profiles produced from GC
data; however, following reactions by NMR indicates that there
is no induction period. For example, the coupling of 1-
1
iodooctane with phenylacetylene was followed by H NMR in
DMF-d7. The formation of the coupling product from the
beginning of the reaction was confirmed by NMR (Figure 5
a
For coupling of iodides, no NaI was added; for coupling of bromides,
b
20 mol % NaI was added. Isolated yields relative to alkyl halide; the
reaction was set up using 2 mL DMF for every 0.5 mmol of alkyl
c
halide. Isolated yields relative to alkyl halide; the reaction was set up
d
using 1 mL DMF for every 0.5 mmol of alkyl halide. Calibrated GC
yields are shown in parentheses.
substrates, increasing the concentrations of the reactants led to
slightly higher yields (about 10%). Alkyl, aryl, and silyl-
substituted alkynes could be coupled (entries 1−3, Table 2).
The isolated yields of some products were significantly lower
than calibrated GC yields (entry 1, Table 2), which was
probably due to the difficulty in isolation of these products.
Modest to good isolated yields were obtained for the coupling
of substrates containing acetal (entry 5, Table 2), amide (entry
6, Table 2), amine (entry 7, Table 2), furan (entry 8, Table 2),
ester (entries 9 and 10, Table 2), alkyl chloride (entry 10, Table
2), ketone (entry 11, Table 2), and nitrile (entry 13, Table 2).
The conditions could be applied to the coupling of alkyl
bromides if 20 mol % of NaI was used for a presumable Br/I
exchange (entries 7, 12−13, Table 2). The coupling of alkyl
chlorides was not successful, likely because of a more inert C−
Cl bond. Unfortunately, the system is not efficient for the
Figure 5. Comparison of time-dependent yields of coupling
determined by NMR and GC.
and Figure S2, Supporting Information). Moreover, the yields
of the coupling product determined by NMR were identical to
the yields determined by GC analysis (Figure 5). For
experimental convenience, we chose GC as the analytical tool
for further kinetic measurements.
The kinetics of the coupling of 1-iodooctane and 1-octyne
was further studied by determining the rates of the coupling
reaction using the initial rate approximation. Figure S3 shows
1166
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171

ACS Catalysis
Research Article
Figure 6. (A) Time-dependent yields of coupling at different loadings of catalyst. (B) The dependence of initial reaction rates on the loading of
catalyst. The rates were averaged over three independent measurements. The error bar represents the standard deviation of the results from three
independent measurements.
the reaction is zeroth-order in the concentration of 1-
iodooctane. An equal amount of product was formed after a
same reaction time for various concentrations of 1-iodooctane
(Figure S3A, Supporting Information). The rates of the
reaction are nearly constant at different concentrations of 1-
iodooctane (Figure S3B). Figure S4 shows that the reaction is
also zeroth order in the concentration of 1-octyne. Figure S5
shows the coupling is independent of the loading of CuI. Figure
S6 shows that the rate of coupling is also independent of the
loading of LiOtBu. On the other hand, Figure 6 shows that the
reaction is first-order in the concentration of catalyst 6. The
reaction profiles are different with different catalyst loadings
(Figure 6A), and the reaction rate is first-order with respect to
catalyst concentration (Figure 6B).
the product in 80% yield (Scheme S1). This result again
suggests that there was no significant catalyst deactivation.
2.4. Turnover-Determining Step. The kinetic studies
show that the catalysis is zeroth-order in CuI, LiOtBu, alkyl
halide, and alkyne and first-order in the catalyst. One possible
explanation is that reductive elimination is the turnover-
determining step. If this is true, the reaction rate should be
influenced by the electronic properties of the substrates.21
Thus, the reaction rates of the coupling of 1-iodooctane with
substituted phenylacetylenes were measured.22 These rates are
compared with the rate of the coupling of 1-iodooctane with 1-
octyne (Figure 7). The rates of the coupling of 1-octyne,
phenylacetylene, and 4-ethynylanisol are similar despite differ-
ence in the electronics. With electron-withdrawing CF3 and F
groups, the rates decrease. These results suggest reductive
elimination is not likely the turnover determining step because
reductive elimination would favor more electron-poor sub-
strates.
Therefore, a simple rate law can be drawn for the catalysis
(eq 1) that can be used to determine the rate constant (kcat).
Table 3 lists the values determined from a different set of
experiments. The averaged value is 1.6 × 10−3 s−1.
A possible turnover-determining step that does not involve
either substrate or Cu is an intramolecular rearrangement of the
catalyst. Judging from the ligands of complex 6, we suspected
that exchange of the flexible ethylenylamine nitrogen by
another ligand from the Ni center was likely during catalysis.
The dissociation of the nitrogen will create an open
coordination site necessary for the coupling reaction. If the
dissociation is the slowest step, then the catalysis would depend
on only Ni, not any other reagent. The different catalytic
efficiencies of complexes 3, 5, and 6 provide circumstantial
evidence for this hypothesis. Complex 3 has a more rigid pincer
ligand, so dissociation of an amine donor would be more
difficult, especially at room temperature. Complex 5 has a more
labile lutidine ligand, and its dissociation is more facile;
however, the favorable ligand dissociation also makes the
complex unstable. Complex 6 has the best catalytic activity,
which might be a result of having a stable “pincer” form and a
hemilabile amine ligand. To further support this hypothesis, the
coupling reaction of 1-iodooctane and 1-octyne was conducted
in the presence of 30 mol % coordinating ligands. Figure 8
shows that the rates of the coupling were decreased in the
presence of these ligands. Among pyridine derivatives, the least
bulky ligand pyridine slowed down the reaction the most. The
less bulky 2,4-lutidine slowed down less, and the most bulky
2,6-lutidine slowed down the least. This result is consistent with
the inhibition of catalytically active, amine-decoordinated nickel
species by the exogenous ligands. The coupling was also slowed
kOBS
[6]
d[P]
dt
= kcat[6] = kOBS → kcat
=
(1)
Table 3. Rate Constants of Coupling of 1-Iodooctane with 1-
Ocytyne
a
set of experiment
kcat (s−1
)
error (%)
varying [1-iodooctane]
varying [1-octyne]
varying [CuI]
1.8 × 10−3
1.8 × 10−3
1.5 × 10−3
1.5 × 10−3
1.8 × 10−3
1.6 × 10−3
10
7
10
10
16
varying [Catalyst]
overall reaction profile
average
a
The error is calculated as the standard deviation of three independent
measurements.
To probe possible catalyst deactivation, a reaction progress
analysis20 at the same excess conditions was performed for the
coupling of 1-iodooctane and the 1-octyne (Figure S7). The
reaction rates are nearly identical for three runs with the same
excess, precluding significant catalyst deactivation. In addition,
the catalytic system was subject to a recycling experiment. After
a catalytic run of the coupling between 1-iodooctane and 1-
octyne (yield = 80%), new samples of the same substrates and
base, but neither the Ni catalyst nor Cu cocatalyst, were added
to the reaction mixture. The coupling proceeded again to give
1167
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171

ACS Catalysis
Research Article
Figure 7. Comparison of coupling rates when different alkynes are used as coupling partners; the rates were averaged over three independent
measurements. The error bar represents the standard deviation of the results from independent measurements.
Scheme 1. Synthesis of Ni-Acetylide Complex 7 under
Catalytically Relevant Conditions
Figure 8. Comparison of coupling rates in the presence of different
exogenous ligands. Conditions: 5 mg of complex 6 (0.016 mmol, 3.5
mol %), 9 mg of CuI (0.047 mmol, 9 mol %), 56 mg of LiOtBu (0.7
mmol), 90 μL of 1-iodooctane (0.5 mmol), 30 mol % of the exogenous
ligand, and 60 μL of decane (0.31 mmol, internal standard) were
placed in a vial and 2 mL of DMF were added; 100 μL of 1-octyne (0.7
mmol) was added under magnetic stirring to the reaction mixture to
start the coupling process.
down in the presence of internal alkynes, such as hexadec-7-yne
and 4-octyne, which suggests that product inhibition might
occur to some degree. However, product inhibition has little
effect on the kinetic data that were collected at the early or
middle stage of catalysis.
2.5. Active Species for Alkylation. The literature suggests
that organometallic Ni complexes are responsible for the
oxidative addition of alkyl halides in a large number of Ni-
catalyzed cross-coupling reactions of alkyl halides.9,19,23−29 For
the reactions described here, a ligated Ni-alkynyl species might
be proposed as the active intermediate. To take advantage of
the well-defined nature of catalyst 6, we attempted to identify
and isolate such an intermediate.
Figure 9. Crystal structure of complex 7. Hydrogen atoms are omitted
for clarity. The thermal ellipsoids are displayed in a 50% probability.
Selected lengths (Å) and angles (deg): Ni1−N1, 1.965(3); Ni1−N2,
1.832(3); Ni2−N3, 1.958(3); Ni1−Cl3, 1.876(5); C13−C14,
1.219(6); N1−Ni1−N2, 85.39(13); N2−Ni1−N3, 85.13(14); N3−
Ni1−N1, 170.52(14); N2−Ni1−C13, 179.22(14); Ni−C13−C14,
175.8(3).
Ni−N2 distance is dictated by the structure of the pincer
ligands. The overall structure of 7 is also similar to that of an
analogous nickel pincer alkynyl complex.9
When complex 6 was treated with equal amounts of CuI,
LiOtBu, and phenylacetylene in DMF at room temperature, a
NNN-Ni-phenylacetylide complex 7 was formed as the only
detectable metal-containing product (Scheme 1), with an
isolated yield of 60%. Complex 7 was independently
synthesized by treating complex 6 with phenylethynylmagne-
sium bromide. The structure of 7 was determined by X-ray
crystallography (Figure 9). The replacement of Cl with
phenylacetylide leads to no significant change in the structural
parameters of the Ni-NNN fragment (compare Figures 3 and
9). The Ni−N2 bond distance is identical in complexes 6 and 7.
This is surprising, considering the different trans-influence
property of Cl and acetylide. A possible explanation is that the
Complex 7 had the same efficiency as complex 6 for the
coupling of 1-iodooctane with phenylacetylene (Scheme 2).
Furthermore, the reaction profile of the coupling using 7 as the
Scheme 2. Comparison of Catalytic Efficiency of Complexes
6 and 7
1168
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171

ACS Catalysis
Research Article
Scheme 3. Reactivity of Nickel Alkynyl Complex 7
precatalyst is identical to the reaction profile of the coupling
using 6 as the precatalyst (Supporting Information). The
reactivity of 7 toward alkyl iodide was probed. Under
catalytically relevant conditions (room temperature in DMF),
7 reacted with 1-iodooctane to give the coupling product, 1-
decyn-1-ylbenzene. However, after 6 h, the yield was only 15%,
much lower than that of catalysis (eq 2, Scheme 3). When 9
mol % CuI was added to this reaction, the yield was again only
10% (eq 3, Scheme 3). However, if this reaction was carried out
in the presence of 1 equiv of phenylacetylene and 9 mol % of
CuI, then after 6 h, the coupling yield was 76%, which was
similar to that of catalysis (eq 4, Scheme 3). When eq 4 was
conducted without LiOtBu or phenylacetylene, no product was
formed. When it was conducted without CuI, a reduced yield of
43% was obtained. These results suggested that complex 7
needs to be activated by an acetylide species to form a Ni
bis(acetylide) complex, which was the species that reacted with
alkyl halide in the catalysis. Copper acetylide was more efficient
than lithium acetylide for the formation of the bis(acetylide)
complex. When 1 equiv of 1-iodooctane, 7, and 1-octyne was
mixed together with 9 mol % of CuI and 1.4 equiv of LiOtBu,
both 1-decyn-1-ylbenzene and hexadec-7-yn were formed in
similar yields, with an overall coupling yield of 60% (eq 5,
Scheme 3). This result showed that the acetylide fragments
from complex 7 and 1-octyne could be coupled in a similar
probability.
To verify whether the phenylacetylide ligand in complex 7
could be exchanged by a different acetylide ligand ligated on Cu
under catalytically relevant conditions, 7 was reacted with 1
equiv of 1-octyne in the presence of 1 equiv of CuI and LiOtBu.
A new nickel complex (8) was formed, and it was present in a
roughly 1:1 ratio to complex 7 in the reaction mixture (eq 6,
Scheme 3). Complex 8 could be tentatively assigned to the Ni
octynyl complex. This result indicates that the acetylide ligand
on Ni is exchangeable during catalysis, which explains the
outcome of eq 5. The exchange reaction likely proceeds via the
Ni bis(acetylide) intermediate. Altogether, the outcomes of
reactions 2−6 provide indirect support for the formation of a
Ni bis(acetylide) complex, possibly connected to a Cu(I) ion,
as the active species for alkylation in the catalysis.
2.6. Tentative Catalytic Cycle. On the basis of the above
results, a tentative catalytic cycle can be drawn (Figure 10).
1169
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171

ACS Catalysis
Research Article
primary alkyl iodides and bromides with terminal alkynes at
room temperature and with good scope and group tolerance.
The mild reaction conditions and the well-defined nature of the
catalyst have facilitated the first in-depth mechanistic study of
this type of reaction. Kinetic measurements confirm the
hemilabile nature of the new pincer ligand. The decoordination
of an amine donor from a catalytic intermediate leads to the
species that activates alkyl halides. Results of kinetics and
inhibition studies are consistent with this decoordination step
being the turnover-determining step of the catalysis. A
catalytically relevant Ni-alkynyl complex has been isolated
and structurally characterized. This species is both chemically
and kinetically competent for the catalytic process. The
reactivity of this Ni-alkynyl species suggests a yet undetected
Ni bis(alkynyl) species as the essential species to activate alkyl
halide. The two alkynyl ligands in this species are exchangeable.
The work provides significant mechanistic insights into the
direct coupling of alkyl halides and alkynes, which is an efficient
and versatile method for the synthesis of alkyl-substituted
alkynes.
Figure 10. A simplified catalytic cycle highlighting the mechanistic
information obtained from this study.
ASSOCIATED CONTENT
* Supporting Information
The following files are available free of charge on the ACS
■
S
Catalyst 6 is first transformed to the alkynyl complex 9. A
further alkynylation on 9 gives a Ni bis(alkynyl) complex 10 in
which an alkynyl ligand is possibly coordinated to a Cu(I) ion.
It is possible that the Cu-alkynyl moiety is coordinated to an
amine donor of the pincer ligand rather than the Ni-center (not
drawn). Recently, Ni−Cu bimetallic alkynyl complexes in
which Cu binds to the alkyne π-system were isolated as
intermediates in Ni-catalyzed Sonogashira coupling of a vinyl
iodide.30 Complex 10 then undergoes a rate-determining
decoordination of the labile amine donor to give a more active
intermediate 11. The Cu ion likely remains coordinated to one
of the nitrogen donors of the pincer ligand because the reaction
is zeroth-order on Cu. Reaction of 11 with alkyl halide then
leads to the coupling product and regenerates complex 9.
Because activation of alkyl halide and reductive elimination
occur after the turnover-determining step, the kinetic data do
not provide information on these reactions. When the coupling
was conducted in the presence of 1 equiv of the radical
inhibitor, TEMPO, the yield decreased to 5%, suggesting the
involvement of radical intermediates. A similar effect of
TEMPO was observed in cross coupling reactions of alkyl
halides catalyzed by complex 3. Our recent studies show that
for the latter reactions, the activation of alkyl halides proceeds
via an alkyl radical in a bimetallic oxidative addition
mechanism.28,29 On the basis of the similarity of catalysts and
reactions, we propose that activation of alkyl halides in the
current catalysis has an analogous radical mechanism.
Publications website at DOI: 10.1021/cs501502u.
Experimental details and characterization data (PDF,
CIF, CIF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: xile.hu@epfl.ch.
■
Present Address
§(Peng Ren) Department of Chemistry, Princeton University,
Princeton, NJ 08544, USA
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work is supported by the EPFL and the Swiss National
Science Foundation (No. 200020_144393/1). We thank Mr.
Isuf Salihu and Mr. Nicolas Boutin (EPFL) for experimental
assistance in the synthesis of ligands and the completion of
Table 2.
REFERENCES
■
(1) Chemistry of Triple-Bonded Functional Groups; Patai, S., Ed.;
Wiley: New York, 1994.
(2) Modern Acetylene Chemistry; Stang, P. J.; Diederich, F., Eds.;
VCH: Weinheim, 1995.
The coupling reaction was monitored by NMR in an attempt
to detect the resting state of the catalysis. Diamagnetic species
6, 9, and 11 were not detected, ruling them out as the resting
states. The dominating Ni species is paramagnetic, and because
of interference of signals from other species present in the
catalysis mixture, its identification by NMR is impossible. The
formation of a paramagnetic resting state, on the other hand, is
consistent with 10 being the resting state because this 5-
coordinate species is expected to be paramagnetic.
(3) Sonogashira, K. In Handbook of Organopalladium Chemistry for
Organic Synthesis; Negishi, E., Ed.; Wiley-Interscience: New York,
2002; Vol. 1, p 493−529.
(4) Chinchilla, R.; Najera, C. Chem. Rev. 2007, 107, 874−922.
(5) Marsden, J. A.; Haley, M. M. In Metal-Catalyzed Cross-Coupling
Reactions, 2nd ed.; De Meijere, A., Diederich, F., Eds.; Wiley-VCH:
Weinheim, 2004; Vol. 1, p 317−394.
(6) Negishi, E.; Anastasia, L. Chem. Rev. 2003, 103, 1979−2017.
(7) Cahiez, G.; Gager, O.; Buendia, J. Angew. Chem., Int. Ed. 2010, 49,
1278−1281.
(8) Thaler, T.; Guo, L. N.; Mayer, P.; Knochel, P. Angew. Chem., Int.
Ed. 2011, 50, 2174−2177.
3. CONCLUSION
A new Ni pincer complex has been developed for the direct
alkylation of alkynes. This complex catalyzes the coupling of
(9) Vechorkin, O.; Godinat, A.; Scopelliti, R.; Hu, X. L. Angew.
Chem., Int. Ed. 2011, 50, 11777−11781.
1170
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171

ACS Catalysis
Research Article
(10) Hatakeyama, T.; Okada, Y.; Yoshimoto, Y.; Nakamura, M.
Angew. Chem., Int. Ed. 2011, 50, 10973−10976.
(11) Cheung, C. W.; Ren, P.; Hu, X. Org. Lett. 2014, 16, 2566−2569.
(12) Ohmiya, H.; Yorimitsu, H.; Oshima, K. Org. Lett. 2006, 8,
3093−3096.
(13) Eckhardt, M.; Fu, G. C. J. Am. Chem. Soc. 2003, 125, 13642−
13643.
(14) Altenhoff, G.; Wurtz, S.; Glorius, F. Tetrahedron Lett. 2006, 47,
2925−2928.
(15) Vechorkin, O.; Barmaz, D.; Proust, V.; Hu, X. L. J. Am. Chem.
Soc. 2009, 131, 12078−12079.
(16) Yi, J.; Lu, X.; Sun, Y.-Y.; Xiao, B.; Liu, L. Angew. Chem., Int. Ed.
2013, 52, 12409−12413.
(17) Ren, P.; Vechorkin, O.; von Allmen, K.; Scopelliti, R.; Hu, X. L.
J. Am. Chem. Soc. 2011, 133, 7084−7095.
(18) Ren, P.; Vechorkin, O.; Csok, Z.; Salihu, I.; Scopelliti, R.; Hu, X.
L. Dalton Trans. 2011, 40, 8906−8911.
(19) Csok, Z.; Vechorkin, O.; Harkins, S. B.; Scopelliti, R.; Hu, X. L.
J. Am. Chem. Soc. 2008, 130, 8156−8157.
(20) Blackmond, D. G. Angew. Chem., Int. Ed. 2005, 44, 4302−4320.
(21) Hartwig, J. F. Organotransition Metal Chemistry − From Bonding
to Catalysis; University Science Books: Sausalito, CA, 2010.
(22) See the Supporting Information.
(23) Hu, X. L. Chem. Sci. 2011, 2, 1867−1886.
(24) Terao, J.; Kambe, N. Acc. Chem. Res. 2008, 41, 1545−1554.
(25) Anderson, T. J.; Jones, G. D.; Vicic, D. A. J. Am. Chem. Soc.
2004, 126, 8100−8101.
(26) Jones, G. D.; McFarland, C.; Anderson, T. J.; Vicic, D. A. Chem.
Commun. 2005, 4211−4213.
(27) Vechorkin, O.; Proust, V.; Hu, X. L. J. Am. Chem. Soc. 2009, 131,
9756−9766.
(28) Breitenfeld, J.; Ruiz, J.; Wodrich, M. D.; Hu, X. L. J. Am. Chem.
Soc. 2013, 135, 12004−12012.
(29) Breitenfeld, J.; Wodrich, M. D.; Hu, X. L. Organometallics 2014,
33, 5708−5715.
(30) Gallego, D.; Brueck, A.; Irran, E.; Meier, F.; Kaupp, M.; Driess,
M.; Hartwig, J. F. J. Am. Chem. Soc. 2013, 135, 15617−15626.
1171
DOI: 10.1021/cs501502u
ACS Catal. 2015, 5, 1164−1171