Synthesis and structural characterization of a monomeric di-copper-substituted silicotungstate [γ-H2SiW10O36Cu2(μ-1,1-N3)2]4- and the catalysis of oxidative homocoupling of alkynes

Article abstract of DOI:10.1016/j.jcat.2008.06.004

The di-copper-substituted γ-Keggin silicotungstate with bis-μ-1,1-azido ligands TBA₄[γ-H₂SiW₁₀O₃₆Cu₂(μ-1,1-N₃)₂] (1, TBA = tetra-n-butylammonium) was synthesized in an aqueous medium. The crystal structure of the anion part of 1 was a monomer of the basal-basal end-on diazido-bridged di-copper-substituted γ-Keggin silicotungstate. The NMR and CSI-MS spectra of 1 in organic solvents, such as acetonitrile, benzonitrile, and 1,2-dichloroethane, showed that complex 1 was present as a monomer of the di-copper-substituted γ-Keggin silicotungstate. Complex 1 could act as an effective homogeneous catalyst for the oxidative homocoupling of various types of alkynes, including aromatic, aliphatic, and heteroatom-containing ones. The reaction possibly proceeds as follows: First, the ligand exchange proceeds between the azido groups in 1 and alkynyl groups to form the corresponding diyne with the reduced copper(I) species via the di-copper(II)-alkynyl intermediate, then the reduced species is reoxidized by molecular oxygen, and the oxidized species reacts with an alkyne to regenerate the alkynyl intermediate.

Full text of DOI:10.1016/j.jcat.2008.06.004

Journal of Catalysis 258 (2008) 121–130
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Synthesis and structural characterization of a monomeric di-copper-substituted
4
−
silicotungstate [γ -H SiW O Cu (μ-1,1-N ) ] and the catalysis of oxidative
2
10 36
2
3 2
homocoupling of alkynes
Kazuya Yamaguchi a,b, Keigo Kamata a,b, Syuhei Yamaguchi , Miyuki Kotani , Noritaka Mizuno
a,b,∗
b
b
a
Department of Applied Chemistry, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan
Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The di-copper-substituted γ -Keggin silicotungstate with bis-μ-1,1-azido ligands TBA [γ -H SiW10 O36Cu2-
4
2
Received 21 March 2008
Revised 29 May 2008
Accepted 3 June 2008
Available online 1 July 2008
(
μ-1,1-N3)2] (1, TBA = tetra-n-butylammonium) was synthesized in an aqueous medium. The crystal
structure of the anion part of 1 was a monomer of the basal–basal end-on diazido-bridged di-copper-
substituted γ -Keggin silicotungstate. The NMR and CSI-MS spectra of 1 in organic solvents, such as
acetonitrile, benzonitrile, and 1,2-dichloroethane, showed that complex 1 was present as a monomer of
the di-copper-substituted γ -Keggin silicotungstate. Complex 1 could act as an effective homogeneous
catalyst for the oxidative homocoupling of various types of alkynes, including aromatic, aliphatic,
and heteroatom-containing ones. The reaction possibly proceeds as follows: First, the ligand exchange
proceeds between the azido groups in 1 and alkynyl groups to form the corresponding diyne with
the reduced copper(I) species via the di-copper(II)-alkynyl intermediate, then the reduced species is
reoxidized by molecular oxygen, and the oxidized species reacts with an alkyne to regenerate the alkynyl
intermediate.
Keywords:
Alkyne
Copper
Diyne
Homocoupling
Polyoxometalate
©
2008 Elsevier Inc. All rights reserved.
1
. Introduction
photungstate KRb5[(α-PW10O37)(Ni(H2O))2(μ-1,1-N3)]·19H2O [12].
Successively, di-copper-substituted silicotungstates, KNaCs10 [{γ -
SiW10O36Cu2(H2O)(μ-1,1-N3)2}2]·26H2O (2) [13] and TEA4TBA2H2-
[(γ -SiW10O36)2Cu4(μ-1,1,1-N3)2(μ-1,1-N3)2] (3, TEA = tetraethyl-
ammonium) [14], were reported by the same group (Fig. 1).
Complexes 2 and 3 have dimeric structures and are composed
Polyoxometalates are attractive compounds used in various
fields, including analytical chemistry, medicine, electrochemistry,
photochemistry, and especially catalysis [1–8]. Recently, interest
in the catalysis of transition metal-substituted polyoxometalates,
synthesized by the introduction of substituent metal ions into the
vacant site(s) of lacunary polyoxometalates, has been growing be-
cause of their unique reactivity, which depends on the composition
and structure of the active sites [1–8]. Since the synthesis of the
6
−
of two [SiW10O36Cu (N ) ]
2
3 2
subunits, with differing connection
modes of the ligands [13,14].
Recently, many applications of 1,3-diyne derivatives (both sym-
metric and nonsymmetric ones) as key structural elements for nat-
ural product synthesis, polymer chemistry, supramolecular chem-
istry, and material science have been reported [16–18]. In addition,
electronic and optical properties of extensively π-conjugated sys-
tems have spurred research into new linear origoacetylenic com-
pounds in which 1,3-diyne derivatives can be used as precursors
[16]. Consequently, much attention has been given to the effi-
cient and selective synthesis of 1,3-diyne derivatives. One of the
most widely used synthesis procedures for synthesizing 1,3-diyne
derivatives is an oxidative alkyne–alkyne homocoupling reaction.
Stoichiometric amounts of copper salts generally have been used
under Glaser conditions [19], and, alternatively, catalytic amounts
of copper salts with appropriate nitrogen bases (ligands or sol-
vents) and molecular oxygen have been used under Hay [20] or
Eglinton conditions [21] for the oxidative alkyne–alkyne homocou-
pling reactions [22–28]. But the yields of aliphatic 1,3-diynes by
8
−
di-lacunary [γ -SiW10O36]
several di-metal-substituted polyoxometalates with the γ -Keggin
framework have been synthesized and applied to the selective
was reported by Tézé and Hervé [9],
“green” oxidation reactions with molecular oxygen or hydrogen
peroxide as the sole oxidant [1–8].
−
The azido ion (N ) is well known to be
a versatile lig-
3
and that can bridge two or more transition metals [10,11]. Re-
cently, various transition metal-substituted polyoxometalates with
azido ligands have been synthesized [12–15]. Mialane et al. re-
ported
a mono-μ-1,1-azido-bridged di-nickel-substituted phos-
Corresponding author at: Department of Applied Chemistry, School of Engineer-
*
ing, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan. Fax:
+81 3 5841 7220.
E-mail address: tmizuno@mail.ecc.u-tokyo.ac.jp (N. Mizuno).
0
021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.06.004

122
K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
lution states. Furthermore, the scope of the 1-catalyzed oxidative
alkyne–alkyne homocoupling is extended, and 13 new entries are
added [36]. We also discuss the possible reaction mechanism for
the present 1-catalyzed homocoupling in more detail [36].
(1)
2. Experimental
2
.1. General
IR spectra were measured with a Jasco FT/IR-460 Plus de-
vice using KBr disks. NMR spectra were recorded at 298 K on
1
13
29
JEOL JNM-EX-270 spectrometer ( H, 270 MHz; C, 67.8 MHz; Si,
3.45 MHz; 1 W, 11.20 MHz). Chemical shifts (δ) were reported
83
5
1
13
in ppm downfield from the internal TMS for H and
C, ex-
Si, and external Na2WO4 (in D2O)
W. UV–vis spectra were recorded on a Jasco V-570 spec-
2
9
ternal TMS (in CDCl3) for
for
183
trometer with a Unisoku thermostatic cell holder (USP-203). GC
analyses were performed on Shimadzu GC-2014 with a flame ion-
ization detector equipped with a TC-5 capillary column (internal
diameter = 0.25 mm, length = 60 m) or a TC-1 capillary col-
umn (internal diameter = 0.25 mm, length = 30 m). Mass spec-
tra were recorded on Shimadzu GCMS-QP2010 equipped with a
TC-5HT capillary column (internal diameter = 0.25 mm, length =
3
0 m). CSI-MS spectra were recorded on JEOL JMS-T100LC. Typical
measurement conditions were orifice voltage, −95 V; spray tem-
perature, 263 K; and ion source temperature, room temperature.
Na2WO4·2H2O, NaN3, and copper salts were obtained from Wako
or Kanto (reagent grade) and used as received. Copper(I) pheny-
lacetylide was obtained from Alfa Aesar (reagent grade) and used
as received. Solvents and alkynes were obtained from Tokyo Kasei
or Aldrich (reagent grade) and purified before use [37]. The di-
copper-substituted silicotungstate 1 was synthesized as described
previously [36] (see Supporting information).
2
.2. Procedure for oxidative alkyne–alkyne homocoupling
The oxidative alkyne–alkyne homocoupling was carried out as
Fig. 1. Polyhedral representations of the anion parts of (a) TBA4[γ -H2Si-
W10 O36Cu2(μ-1,1-N3)2] (this study) (1) [36], (b) KNaCs10 [{γ -SiW10 O36Cu2-
follows. First, 1 (2.2 mol% with respect to alkyne), alkyne (1 mmol),
and benzonitrile (1 mL) were successively placed into a glass re-
actor. The reaction mixture was stirred at 373 K under 1 atm of
molecular oxygen. The yield was periodically determined by GC
analysis. After the reaction was completed, 20 mL of diethylether
was added to the solution. The precipitated catalyst was recovered
by filtration (95% recovery), washed with diethylether (ca. 50 mL),
and dried in vacuo before being recycled. Isolation and purification
of diynes were carried out by column chromatography on silica
gel, using n-hexane as an eluent. All products were confirmed by
comparing their GC retention times, mass, and NMR spectra with
those of authentic samples. The purity of isolated products was de-
termined by 1H NMR and was >95% in all cases.
(H2O)(μ-1,1-N3)2}2]·26H2O (2) [13], and (c) TEA4TBA2H2[(γ -SiW10 O36)2Cu4-
(μ-1,1,1-N3)2(μ-1,1-N3)2] (3) [14]. Gray octahedra and blue tetrahedra show the
[WO6] and [SiO4] units, respectively. Blue, green, and light blue spheres show the
copper, nitrogen (azido), oxygen (water) atoms, respectively. (For interpretation of
the references to color in this figure legend, the reader is referred to the web ver-
sion of this article.)
the homocoupling reactions of aliphatic alkynes under the Hay
or Eglinton conditions in the presence of copper salts generally
have been only low to moderate [19–28]. Although the oxidative
homocoupling of aliphatic alkynes with combined catalysts of cop-
per and palladium salts can proceed efficiently, these systems have
several disadvantages, including (i) the use of expensive palladium
catalysts, (ii) the need for bases and co-catalysts (copper salts), and
3
. Results and discussion
(iii) difficult catalyst recovery [29–35]. Thus, the development of
the widely usable efficient alkyne–alkyne homocoupling systems
remains a challenge. The scope and limitations of copper- and/or
palladium-mediated homocoupling systems have been reviewed in
more detail previously [16].
Recently, we reported that a di-copper-substituted γ -Keggin
silicotungstate TBA4[γ -H2SiW10O36Cu2(μ-1,1-N3)2] (1) acts as an
effective homogeneous catalyst for the oxidative alkyne–alkyne ho-
mocoupling reactions [Eq. (1)] [36]. In this paper, we report details
of the synthesis procedure and structural characterization of the
di-copper-substituted γ -Keggin silicotungstate 1 in solid and so-
3
.1. Synthesis and characterization of
TBA4[γ -H2SiW10O36Cu2(μ-1,1-N3)2] (1)
The TBA salt of a di-copper-substituted silicotungstate with bis-
μ-1,1-azido ligands (1) was obtained by the reaction of K8[γ -
SiW10O36] with two equivalents of CuCl2 and an excess amount
of NaN3 in aqueous solution at room temperature, followed by
the addition of TBABr. The detailed procedure is as follows. First,
CuCl2 (0.090 g, 0.67 mmol, two equivalents with respect to [γ -
SiW10O36]8 ) in water (20 mL) was added to a suspension of
−

K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
123
Fig. 2. ORTEP drawing of the anion part of 1 drawn at 30% probability level.
K8[γ -SiW10O36]·12H2O (1.00 g, 0.335 mmol) in water (20 mL). Af-
Table 1
Selected structural parameters for 1
ter 1 min, NaN3 (0.114 g, 1.75 mmol, 5.2 equivalents with respect
to [γ -SiW10O36]8 ) in water (20 mL) was added, followed by the
immediate addition of TBABr (2.00 g, 6.20 mmol) in water (20 mL).
The pH of the mixture was adjusted to 6.0 with 1.0 M aqueous
HNO3 solution, followed by stirring for 20 min at room temper-
ature. The green precipitate of 1 was collected by filtration and
washed with an excess amount of water (0.78 g, 64% yield based
on K8[γ -SiW10O36]·12H2O).
−
Length (Å)
Length (Å)
Cu(1)–O(1)
Cu(1)–N(1)
Cu(2)–O(2)
Cu(2)–N(1)
Cu(1)···Cu(2)
W=Oav
1.89(2)
1.94(3)
1.93(2)
1.93(3)
2.812(5)
1.68
Cu(1)–O(3)
Cu(1)–N(4)
Cu(2)–O(4)
Cu(2)–N(4)
1.95(2)
1.99(3)
1.93(2)
1.96(3)
W–OCu av
W–OW av
1.85
1.92
W–OSi av
2.31
1.63
Si–Oav
The elemental analysis of 1 demonstrated a Si:Cu:W ratio of
1:2:10. The IR spectrum of 1 showed bands of the asymmet-
Angle (deg.)
Angle (deg.)
ric and symmetric stretching vibrations of the μ-1,1-azido lig-
Cu(1)–N(1)–Cu(2)
O(1)–Cu(1)–O(3)
O(1)–Cu(1)–N(4)
O(3)–Cu(1)–N(4)
O(2)–Cu(2)–O(4)
O(2)–Cu(2)–N(4)
O(4)–Cu(2)–N(4)
93(1)
Cu(1)–N(4)–Cu(2)
O(1)–Cu(1)–N(1)
O(3)–Cu(1)–N(1)
N(1)–Cu(1)–N(4)
O(2)–Cu(2)–N(1)
O(4)–Cu(2)–N(1)
N(1)–Cu(2)–N(4)
90(1)
92(1)
173(2)
77(1)
96(1)
170(1)
78(1)
−1
ands at 2078 and 1274 cm , respectively [38]. Complex 1 was
found to be soluble in various organic solvents, including acetoni-
trile, benzonitrile, dichloromethane, 1,2-dichloroethane, DMF, and
DMSO, and the single crystal suitable for X-ray analysis was suc-
cessfully obtained by recrystallization in a mixed solvent of 1,2-
dichloroethane and dichloromethane with vapor diffusion of di-
ethylether. The molecular structure of 1 is illustrated in Fig. 2. As
shown, the anion part of 1 is a monomer of the basal–basal end
on diazido-bridged di-copper-substituted γ -Keggin silicotungstate.
The selected bond lengths and angles are summarized in Table 1.
The Cu(1)···Cu(2) distance, dCu(1)···Cu(2), was 2.812(5) Å, shorter
than those of dimeric complexes 2 [dCu(1)···Cu(2) = 2.979(1) Å] [13]
93.9(9)
170(1)
95(1)
92.4(9)
170(1)
91(1)
copper-substituted γ -Keggin silicotungstate under the present con-
ditions.
The existence of four TBA cations per the anion part of 1 im-
plies a −4 charge of the cluster anion. The bond valence sum (BVS)
[39,40] values of copper (1.84–1.87), tungsten (5.88–6.53), and sil-
icon (3.91) indicate that the respective valences in 1 were +2, +6,
and +4. Thus, based on the charge balance, two protons were asso-
ciated with the cluster anion. The BVS values of 1.24 for O(12) and
◦
and 3 [2.875(3) Å] [14]. The average Cu–N–Cu angle, θav, was 91.5 ,
smaller than those of 2 (96.7 ) [13] and 3 (94.6 ) [14].
◦
◦
Recently, two di-copper-substituted silicotungstate dimers 2 [13]
and 3 [14] were reported by Mialane et al. (Figs. 1b and 1c). Com-
plex 2 was synthesized in aqueous media with hydrophilic alkali
metals (K, Na, and Cs) as counter cations [13]. Complex 3 was
synthesized in organic solvents (an acetonitrile–methanol mixture)
with lyophilic alkylammoniums (TEA and TBA) as mixed counter
cations [14]. Whereas the connection modes of the ligands dif-
1.35 for O(13) were lower than those for the other oxygen atoms
in 1 (1.54–2.27), suggesting that O(12) and O(13) were monoproto-
nated. These results and elemental analysis show that the formula
of 1 was TBA4[γ -H2SiW10O36Cu2(μ-1,1-N3)2]. The formation of 1
anion is expressed by the following equation:
[γ -SiW10O36 8
]
−
+ 2Cu2+ + 2N−3 + 2H
+
fered, both conditions gave the dimer of the [SiW10O36Cu2(N3)2]6
−
4
−
→
[γ -H2SiW10O36Cu2(μ-1,1-N3)2]
(2)
subunit in the solid states. In sharp contrast, the synthesis in
aqueous media with lyophobic TBA as a counter cation afforded
the monomer form of the basal–basal end on diazido-bridged di-
To investigate the state of complex 1 in solution, NMR spec-
tra were measured in organic solvents. The 29Si NMR spectrum

124
K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
8
−
field (ꢀδ = 39.4 ppm) to that of [γ -SiW10O36] , and the line was
broadened. This upfield shift and line broadening are likely due
to the effect of the paramagnetic d9 copper(II) species. The
183
W
NMR spectrum exhibited two signals at −130.6 and −136.3 ppm,
with a respective intensity ratio of 1:2 (Fig. 3b). The signals of
four tungsten atoms, W(1), W(2), W(3), and W(4), bound to the
183
copper units were not observed; a similar disappearance of
W
NMR signals has been reported for a copper(II)-substituted polyox-
9
−
[41]. This is likely due to the
ometalate [α2-P2W17 O61Cu(II)Br]
9
paramagnetic coupling with d copper(II) species [41]. The struc-
tural symmetry of tungsten atoms (C2v ) suggests that the cluster
framework of 1 maintained the γ -Keggin structure in acetoni-
trile.
Next, the cold-spray ionization mass (CSI-MS) spectra of 1
were measured in organic solvents. It has been reported that
the alkali metal salts of polyoxometalates are soluble only in
water and give very complex MS spectra due to the side reac-
tions [42]. In contrast, the TBA salts of polyoxometalates in or-
ganic solvents give well-defined MS spectra, related to the acces-
sible protonation sites of the clusters in solution [42]. The CSI-
MS spectrum of the acetonitrile solution of 1 exhibited mainly
a +1-charged ion peak at m/z = 3866, which can be assigned
+
to {TBA5[H2SiW10O36Cu2(N3)2]} (Fig. 4). Ion peaks corresponding
to dimer species and non-copper-substituted silicotungstates were
hardly observed in the CSI-MS spectrum. The CSI-MS spectrum of
1
in benzonitrile and 1,2-dichloroethane also exhibited similar ion
peaks that can be assigned to the monomer species. All of these
NMR and CSI-MS data suggest that complex 1 was present as a
monomer of the di-copper-substituted γ -Keggin silicotungstate in
these solutions.
Fig. 3. (a) 29Si and (b) 183 W NMR spectra of 1 in CD3CN.
3
.2. Scope of the 1-catalyzed oxidative alkyne–alkyne homocoupling
The 1-catalyzed oxidative homocoupling of ethynylbenzene (4a)
of the acetonitrile (CD3CN) solution of 1 gave a chemical shift at
122.8 ppm, suggesting that compound 1 is a single species in
acetonitrile (Fig. 3a). The Si NMR resonance was shifted to up-
−
2
9
to 1,4-diphenyl-1,3-butadiyne (5a) was carried out in various sol-
+
Fig. 4. CSI-MS spectra of 1; (a) m/z = 2000–8000 and (b) m/z = 3800–3925; m/z = 3866 {TBA5[H2SiW10 O36Cu2(N3)2]} (A). The lines in (c) are the calculated pattern of
+
{
TBA5[H2SiW10 O36Cu2(N3)2]}
.

K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
125
Table 2
Effect of solvents on the 1-catalyzed oxidative homocoupling of 4a
a
Entry
Solventb
Additive
Time (h)
Yield of
c
5a
(%)
1
2
Benzonitrile
Benzonitrile
DMF
DMF
DMF
DMF
DMF
DMSO
DMSO
Acetonitrile
1,2-Dichloroethane
Nitrobenzene
Toluene
–
–
–
–
3
6
3
24
24
3
3
3
3
3
91
86
39
63
88
9
90
39
86
15
4
d
3
4
5
6
7
8
9
Benzonitrile (0.5 mmol)
Water (1 mmol)
MS4A (200 mg)
–
MS4A (200 mg)
1
0
–
–
–
–
–
1
1
3
3
3
3
1
2
3
2
2
1
3
Scheme 1. Possible reaction mechanism for the copper-mediated alkyne–alkyne ho-
mocoupling proposed by Bohlmann and co-workers [43]. This mechanism has been
well accepted [16]. (N = nitrogen bases, X = Cl−, OAc , etc.)
14
Acetic acid
−
a
Reaction conditions: 1 (2.2 mol% with respect to 4a), 4a (1 mmol), solvent
(
1 mL), 373 K, O2 (1 atm).
b
DMF = N,N-dimethylformamide. DMSO = dimethyl sulfoxide. MS4A =
and that various kinds of solvents can be used for the present ho-
mocoupling reaction in the presence of water scavengers.
Molecular Sieve 4A (pretreated at 423 K).
c
Determined by GC analysis using naphthalene as an internal standard.
The reaction was carried out in air (1 atm).
d
The effect of the catalysts was investigated (Table 3). In the
absence of catalysts or in the presence of simple copper(I) and cop-
per(II) salts, the reaction hardly proceeded under the present con-
ditions (entries 5–11). The mono-copper-substituted silicotungstate
TBA4[α-H2SiW11 CuO39] (entry 2), the non-copper-substituted sili-
cotungstate TBA4[γ -SiW10O34(H2O)2] (entry 3), and a mixture of
TBA4[γ -SiW10O34(H2O)2] and CuCl2 (entry 4) were almost inac-
tive. This demonstrates that the di-copper core {Cu2(μ-1,1-N3)2}
in 1 played an important role in the present oxidative alkyne–
alkyne homocoupling. Bohlmann et al. [43] proposed that the
copper(II)-catalyzed alkyne–alkyne homocoupling reaction pro-
ceeds via the formation of the di-copper(II)–alkynyl intermediate
Table 3
a
Oxidative homocoupling of 4a by various catalysts
Entry
Catalyst
Yield of
b
5a
(%)
1
2
3
1
91
2
TBA4[α-H2SiW11 CuO39]
c
TBA4[γ -SiW10 O34(H2O)2]
<1
5
10
4
d
4
TBA4[γ -SiW10 O34(H2O)2] + CuCl2
5
6
7
8
9
Cu(OAc)2
CuCl2
CuCl
7
{
Cu(II)2(μ-C≡CR)2}, which would collapse directly to the 1,3-diyne
CuI
2
[Cu(CH3CN)4]PF6
Cu(I) phenylacetylide
none
<1
<1
<1
products, as shown in Scheme 1. This reaction mechanism gen-
erally has been accepted [16]. The present homocoupling likely
proceeded via the formation of similar alkynyl species (see Sec-
tion 3.3).
1
1
0
1
a
Reaction conditions: catalyst (Cu: 4.4 mol% with respect to 4a), 4a (1 mmol),
benzonitrile (1 mL), 373 K, 3 h, O2 (1 atm).
b
In the present study, we also investigated the scope of the 1-
catalyzed oxidative homocoupling reaction. We recently reported
on the 1-catalyzed oxidative alkyne–alkyne homocoupling [36]. In
the present work, we extended the scope of the investigation and
included 13 new entries, as shown in Table 4. The catalytic ox-
idative homocoupling reactions of aromatic alkynes (4a–4m) with
electron-donating as well as electron-withdrawing substituents
gave the corresponding aromatic diynes in excellent yields (entries
Determined by GC analysis using naphthalene as an internal standard.
c
2
.2 mol%.
d
Mixture of TBA4[γ -SiW10 O34(H2O)2] (2.2 mol%) and CuCl2 (4.4 mol%) was used.
vents (Table 2). Among the solvents tested (with no additives),
benzonitrile gave the corresponding diyne 5a in the highest yield,
1% (entry 1). In benzonitrile, the reaction proceeded efficiently
9
1–18). The positions of the substituents on aromatic rings did not
even in air (1 atm) (entry 2). Polar aprotic solvents, such as DMF,
DMSO, and acetonitrile, gave 5a in moderate yields (entries 3, 8,
and 10), whereas nonpolar and protic polar solvents, such as 1,2-
dichloroethane, toluene, and acetic acid, gave low yields of 5a (en-
tries 11–14). In the 1-catalyzed homocoupling of 4a in benzonitrile,
an equimolar amount of benzamide (formed through the hydration
of benzonitrile) with respect to the corresponding diyne 5a was
formed as a co-product. A separate experiment confirmed that the
homocoupling reaction was strongly inhibited by the presence of
water (entry 6 vs entry 3). These findings suggest that benzoni-
trile likely acted as a scavenger to remove the water formed by
the homocoupling reaction. The amount of benzonitrile can be re-
duced. When the oxidative homocoupling of 4a in DMF was carried
out in the presence of stoichiometric amounts of benzonitrile with
respect to the amount of water formed, an 88% yield of 5a was
obtained over 24 h (entry 5) (cf. 63% yield without benzonitrile;
entry 4). In addition, the reaction of 4a in DMF and DMSO with
Molecular Sieve 4A (200 mg, pretreated at 423 K) gave 5a in 90%
and 86% yields, respectively, over 3 h (entries 7 and 9). The forego-
ing results demonstrate that benzonitrile is not a special solvent,
affect the reaction: The reaction rates for the oxidative coupling
of 2-, 3-, and 4-methylethynylbenzenes were almost the same (en-
tries 8–10). No dehalogenation occurred in the case of halogen-
substituted ethynylbenzenes (4i–4k) (entries 14–16). The reaction
of the nitrogen- (4l, 4n, and 4o) and sulfur-containing (4p) alkynes
also proceeded smoothly and gave the corresponding diynes in
high yields (entries 17, 19–21). Of note, the less reactive aliphatic
terminal alkynes (4q–4v) [19–28] were efficiently oxidized to give
the corresponding aliphatic diynes in high yields (entries 22–27).
Propargylic alcohols (4w–4y) and amine (4z) also gave the corre-
sponding diynes (entries 28–31). These findings indicate that the
present system has very wide applicability, and that various kinds
of structurally diverse alkynes (including aromatic, aliphatic, and
heteroatom-containing ones) can be selectively converted to the
corresponding diynes in the presence of 1.
Complex 1 can be applied to larger-scale diyne production. For
example, in a 20-mmol scale reaction of 4a (1; 0.2 mol%, PhCN;
20 mL, 373 K, 18 h), the turnover number (TON = amount of
1a consumed/amount of 1) reached up to 468, the highest value
reported for the copper-mediated oxidative alkyne–alkyne homo-

126
K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
Table 4
a
Oxidative alkyne–alkyne homocoupling catalyzed by 1
Entry
Alkyne
Time (h)
Yield of
diyneb
(
%)
1
Ethynylbenzene (4a)
4a
4a
4a
4a
3
3
3
3
3
91 (88)
94
82
88
80
c
2
c
3
c
4
c
5
c
6
4a
3
78
7
8
9
Ethynylbenzene-d1 (4b)
2-Methylethynylbenzene (4c)
3-Methylethynylbenzene (4d)
4-Methylethynylbenzene (4e)
4-n-Penthylethynylbenzene (4f)
3-Methoxyethynylbenzene (4g)
4-Methoxyethynylbenzene (4h)
4-Fluoroethynylbenzene (4i)
3-Chloroethynylbenzene (4j)
4-Bromoethynylbenzene (4k)
3-Aminoethynylbenzene (4l)
1-Ethynylnaphthalene (4m)
3-Ethynylpyridine (4n)
4-Ethynylpyridine (4o)
3-Ethynylthiophene (4p)
1-Hexyne (4q)
3
3
3
3
3
3
3
3
3.5
4.5
5
4
2
2
4
5
5
87
95
97 (93)
95 (97)
96 (64)
91 (90)
97 (82)
96 (89)
90 (93)
94 (90)
90
93
95 (82)
83
92
91 (80)
90
1
0
1
1
1
2
1
3
1
4
1
5
1
6
1
7
1
8
1
9
20
2
1
22
23
24
25
26
3,3-Dimethyl-1-butyne (4r)
1-Octyne (4s)
1-Decyne (4t)
5
7
8
91 (86)
92 (89)
76
1-Hexadecyne (4u)
2
7
6-Chloro-1-hexyne (4v)
2-Propyn-1-ol (4w)
2-Methyl-3-butyn-2-ol (4x)
1-Ethynylcyclohexanol (4y)
N,N-Dimethyl-2-propyn-1-amine (4z)
7
18
2
4
4
76
85
>99 (99)
90 (82)
91
28
29
30
31
a
Reaction conditions: 1 (2.2 mol% with respect to alkyne), alkyne (1 mmol), ben-
zonitrile (1 mL), 373 K, O2 (1 atm).
b
Determined by GC analysis using naphthalene as an internal standard. The val-
ues in the parentheses were the isolated yields.
c
These experiments used a recycled catalyst; first recycle (entry 2), second recy-
cle (entry 3), third recycle (entry 4), fourth recycle (entry 5), and fifth (entry 6). The
reaction was carried out under the same as those in entry 1.
Fig. 5. Reaction profiles for the 1-catalyzed homocoupling of 4a under (a, b) O2 and
c) Ar atmosphere. Reaction conditions for (a) and (c): 1 (20 mM), 4a (900 mM),
(
coupling reactions [19–35]. Furthermore, 4.5 g of 5a (89% isolated
yield, 99% purity by 1H NMR) was obtained in a 50-mmol scale re-
action (1; 1.1 mol%, PhCN; 30 mL, 373 K, 6 h). After the reaction
was completed, the catalyst could be easily recovered by adding
an excess of diethylether (the precipitation method; see Section 2).
The recovered catalyst could be recycled at least five times with
no significant loss of catalytic activity (entries 2–6).
benzonitrile (1 mL), 373 K, and O2 or Ar (1 atm). Reaction conditions for (b): 1
(4 mM), 4a (200 mM), benzonitrile (1 mL), 373 K, and O2 (1 atm). In the case of
(
c), the reaction was initially carried out under Ar atmosphere. After 180 min, 1 atm
of molecular oxygen was introduced as indicated by an arrow.
action of 1 with an alkyne to form the catalytically active species.
Consequently, the reactivity of 1 with an alkyne (4a) was evalu-
ated. To the benzonitrile solution of 1 (4 mM), 50 equivalents of
4
3
a with respect to 1 were added, and the solution was heated to
73 K under Ar atmosphere. Then the UV–vis spectra of the so-
lution and the production of 5a (by GC analysis) were monitored
Fig. 6). An induction period (approximately 50 min) was observed
3
.3. Mechanistic studies
The addition of radical scavenger (2,6-di-tert-butyl-4-methyl-
(
phenol, two equivalents with respect to 1) and radical initiator
ꢀ
for the formation of 5a (Fig. 6d) in a similar way as that for cat-
alytic oxidative homocoupling (O2 atmosphere) under the same
conditions (Fig. 5b). During the induction period, the intensity of
(
2,2 -azobis(isobutyronitrile), two equivalents with respect to 1)
did not change the reaction rates (profiles) or the product selec-
tivity for the 1-catalyzed oxidative homocoupling of 4a under the
conditions in Table 4. This suggests that free-radical intermediates
−
the absorption band at 360 nm assignable to the N3 → Cu(II)
charge-transfer [44–46] of 1 gradually decreased, and the band al-
most disappeared after approximately 50 min (Figs. 6a and 6c).
The IR spectrum of 1 recovered after treatment with 4a at 373 K
showed that the bands of the asymmetric and symmetric stretch-
ing vibrations of the μ-1,1-azido ligands had disappeared. These
findings indicate that the azido groups were eliminated from 1
during the induction period.
(e.g., alkynyl radicals) were not involved in the present oxidative
homocoupling reactions.
The reaction profile for the 1-catalyzed oxidative homocoupling
of 4a under the conditions in Fig. 5a ([1], 20 mM; [4a], 900 mM)
showed an induction period of approximately 5 min. The length
of the induction period depended on the concentrations of 1 and
4
a; a decrease in the concentrations of 1 and 4a prolonged the
induction period. When the homocoupling of 4a was carried out
under the conditions in Fig. 5b ([1], 4 mM; [4a], 200 mM), the
induction period was approximately 50 min. The induction period
disappeared on pretreatment of 1 with 4a under Ar atmosphere at
In the presence of 1, no reactions of internal alkynes, such
as 2-octyne and diphenylacetylene, proceeded. Therefore, the cop-
per species in 1 is terminally bound to an alkyne. Both the cat-
alytic homocoupling of 4a with Cu(I) phenylacetylide (see entry
10 in Table 3) and the stoichiometric reaction with Cu(I) pheny-
373 K, suggesting that the induction period was caused by the re-

K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
127
Fig. 6. Changes of UV–vis spectra ((a) ∼50 min and (b) 50–100 min) for the reaction of 1 with 4a under Ar atmosphere, and time profiles of (c) the absorbance at 360 nm
and 700 nm and (d) formation of 5a. Reaction conditions: 1 (4 mM), 4a (200 mM), benzonitrile (1 mL), 373 K, and Ar (1 atm).
lacetylide did not proceed, suggesting that formation of Cu(I)
acetylide species was not involved in the present catalytic cycle. As
mentioned earlier, the di-copper core in 1 played an important role
in the present oxidative alkyne–alkyne homocoupling, and it has
been proposed that the copper(II)-catalyzed alkyne–alkyne homo-
coupling reaction proceeds via the formation of the di-copper(II)–
alkynyl intermediate {Cu(II)2(μ-C≡CR)2} [16,43]. Thus, it is likely
that the present homocoupling proceeded via the formation of
similar di-copper(II)–alkynyl species formed by the ligand ex-
change between the azido groups in 1 and alkynyl groups, and
that the induction period corresponded to formation of the catalyt-
ically active di-copper(II)–alkynyl species [Eq. (3)] [36]. Whereas
the absorption bands of copper alkynyl species has been observed
at around 370 nm [47], here the absorption band could not be de-
tected because of the overlap with the intense absorption bands of
suggest that the copper(II) species in 1 was reduced to copper(I)
species [Eq. (4)] [36]. The d–d transition band of the copper(II)
species again appeared when molecular oxygen was introduced to
the solution, indicating that the reduced copper(I) species in 1 was
reoxidized to copper(II) species by molecular oxygen [Eq. (5)] [36].
When molecular oxygen was introduced to the system after the
treatment of 1 with 4a at 373 K under Ar atmosphere, no induc-
tion period was observed, and the homocoupling of 4a proceeded
at almost the same rate as that under the catalytic turnover con-
ditions [profile (a) vs profile (c) in Fig. 5].
(4)
1
, solvent, and products.
(5)
Monitoring the formation of water (benzamide) for the 1-
catalyzed homocoupling of 4a with molecular oxygen revealed that
the amount of water produced was the same as that of 5a pro-
duced, as shown in Fig. 7a. Molecular oxygen uptake during the
homocoupling of 4a also was measured. As shown in Fig. 7b, the
amount of molecular oxygen consumed was half that of 5a pro-
duced. These respective 1:1 (water formed/diyne formed) and 1:2
(
3)
After the induction period, an almost equimolar amount of 4a
(
8.8± 1.0 mM) with respect to the copper(II) species in 1 (8.0 mM)
was converted into 5a, with a concomitant decrease in the absorp-
tion band at 700 nm, which can be assigned to the d–d transi-
tion of the copper(II) species in 1 (Figs. 6b–6d) [48]. These results

128
K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
Fig. 7. (a) Relationship between amounts of 5a and benzamide (water) formed, and (b) relationship between amounts of 5a formed and uptake of molecular oxygen for the
1
-catalyzed homocoupling of 4a. Reaction conditions were the same as those in Table 3. Slope of (a) (water formed/5a formed) = 1.11. Slope of (b) (O2 uptake/5a formed) =
0.56.
Fig. 8. Dependences of the reaction rate (R0) on (a) the concentration of 1, (b) the concentration of 4a, and (c) the partial pressure of molecular oxygen for the 1-catalyzed
homocoupling of 4a. Reaction conditions: 1 (7.0–28.7 mM), 4a (0.7–2.0 M), benzonitrile (1 mL), 373 K, and O2 (∼1 atm).

K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
129
Acknowledgments
This work was supported by the Core Research for Evolutional
Science and Technology (CREST) program of the Japan Science and
Technology Agency (JST) and the Grants-in-Aid for Scientific Re-
search from Ministry of Education, Culture, Sports, Science and
Technology. The authors thank T. Katayama, University of Tokyo for
his help with the experiments.
Supporting information
The online version of this article contains additional supple-
mentary material.
Please visit DOI: 10.1016/j.jcat.2008.06.004.
References
[1] M.T. Pope, Heteropoly and Isopoly Oxometalates, Springer-Verlag, Berlin, 1983,
pp. 1–165.
[
[
[
[
[
[
[
2] C.L. Hill, C. Chrisina, M. Prosser-McCartha, Coord. Chem. Rev. 143 (1995) 407.
3] T. Okuhara, N. Mizuno, M. Misono, Adv. Catal. 41 (1996) 113.
4] N. Mizuno, M. Misono, Chem. Rev. 98 (1998) 199.
5] R. Neumann, Prog. Inorg. Chem. 47 (1998) 317.
6] I.V. Kozhevnikov, Chem. Rev. 98 (1998) 171.
Fig. 9. Arrhenius plots for the 1-catalyzed homocoupling of 4a. Line fit:
ln(kobs) = 24.4 − 8296.5/T .
7] N. Mizuno, K. Yamaguchi, K. Kamata, Coord. Chem. Rev. 249 (2005) 1944.
8] N. Mizuno, K. Kamata, K. Yamaguchi, in: R. Richards (Ed.), Surface and
Nanomolecular Catalysis, Taylor and Francis Group, New York, 2006, pp. 463–
492.
(molecular oxygen consumed/diyne formed) stoichiometries sup-
[9] A. Tézé, G. Hervé, Inorg. Synth. 27 (1990) 85.
port the overall reactions shown in Eqs. (4) and (5).
The initial reaction rate (R0) for the homocoupling of 4a
showed a first-order dependence on the concentration of 1 (7.0–
[
10] J. Ribas, A. Escuer, M. Monfort, R. Vicente, R. Cortés, L. Lezama, T. Rojo, Coord.
Chem. Rev. 193–195 (1999) 1027.
[
11] S. Cenini, E. Gallo, A. Caselli, F. Ragaini, S. Fantauzzi, C. Piangiolino, Coord.
2
8.7 mM, Fig. 8) and was almost independent of both the con-
Chem. Rev. 250 (2006) 1234.
[
12] P. Mialane, A. Dolbecq, J. Marrot, F. Sécheresse, Angew. Chem. Int. Ed. 43 (2004)
274.
centration of 4a (0.7–2.0 M, Fig. 8) and the partial pressure of
molecular oxygen (>0.7 atm, Fig. 8). Good linearity of the Arrhe-
nius plots (observed rate constant kobs vs T ; Fig. 9) was seen,
2
[
13] P. Mialane, A. Dolbecq, J. Marrot, E. Rivière, F. Sécheresse, Chem. Eur. J. 11
(2005) 1771.
[14] P. Mialane, C. Duboc, J. Marrot, E. Rivière, A. Dolbecq, F. Sécheresse, Chem. Eur.
J. 12 (2006) 1950.
15] P. Mialane, A. Dolbecq, F. Sécheresse, Chem. Commun. (2006) 3477.
16] P. Siemsen, R.C. Livingston, F. Diederich, Angew. Chem. Int. Ed. 39 (2000) 2632.
−
1
−
1
and the activation energy (Ea) was 68.9 kJ mol , lower than that
−1
of the homocoupling under Hay’s conditions (Ea = 87.9 kJ mol
)
[
[
[
[35]. From the kinetic data, R0 for the 1-catalyzed homocoupling
of 4a can be expressed by the following equation:
17] L. Hansen, P.M. Boll, Phytochemistry 25 (1986) 285.
[
18] H. Matsunaga, M. Katano, H. Yamamoto, H. Fujito, M. Mori, K. Takata, Chem.
Pharm. Bull. 38 (1990) 3480.
10
3
1
0
0
O2
R0 = 3.93 × 10 exp(−8.30 × 10 /T )[1] [alkyne(4a)] p
.
(6)
[19] C. Glaser, Ber. Dtsch. Chem. Ges. 2 (1869) 422.
A kinetic isotope effect was not observed for the oxidative ho-
mocoupling of ethynylbenzene (4a) and ethynylbenzene-d1 (4b)
under the conditions in Table 4 (kH/kD = 1.0), indicating that the
C–H bond cleavage is not included in the rate-limiting step [49].
These results demonstrate that the formation of a diyne from di-
copper(II)–alkynyl species [Eq. (4)] is the rate-limiting step.
[20] A.S. Hay, J. Org. Chem. 27 (1962) 3320.
21] I.D. Campbell, G. Eglinton, Org. Synth. 45 (1965) 39.
[
[
[
[
22] S.M. Auer, M. Schneider, A. Baiker, J. Chem. Soc. Chem. Commun. (1995) 2057.
23] J. Li, H. Jiang, Chem. Commun. (1999) 2369.
24] R. Salazar, L. Fomina, S. Fomine, Polym. Bull. 47 (2001) 151.
[25] A. Sharifi, M. Mirzaei, M.R. Naimi-Jamal, J. Chem. Res. (2002) 628.
[26] J.S. Yadav, B.V.S. Reddy, K.B. Reddy, K.U. Gayathri, A.R. Prasad, Tetrahedron
Lett. 44 (2003) 6493.
[
27] L. Wanga, J. Yana, P. Lia, M. Wanga, C. Su, J. Chem. Res. (2005) 112.
4
. Conclusion
[28] X. Lu, Y. Zhang, C. Luo, Y. Wang, Synth. Commun. 36 (2006) 2503.
[
[
29] R. Rossi, A. Carpita, C. Bigelli, Tetrahedron Lett. 26 (1985) 523.
30] Q. Liu, D.J. Burton, Tetrahedron Lett. 38 (1997) 4371.
In this work, the monomeric di-copper-substituted γ -Keggin
[
31] W.A. Herrmann, V.P.W. Böhm, C.V.K. Gstöttmayr, M. Grosche, C.-P. Reisinger, T.
Weskamp, J. Organomet. Chem. 618 (2001) 616.
silicotungstate with bis-μ-1,1-azido ligands 1 was synthesized, and
the molecular structure was successfully determined by the X-ray
crystallographic analysis. Complex 1 was present as a monomer,
and no dimerization occurred in organic solvents, including ace-
tonitrile, benzonitrile, and 1,2-dichloroethane. Complex 1 could act
as an effective homogeneous catalyst for the oxidative alkyne–
alkyne homocoupling. Various structurally diverse alkynes, includ-
ing aromatic, aliphatic, and heteroatom-containing ones, could be
converted into the corresponding diynes in high to excellent yields.
The catalyst could be easily separated from the reaction mixture,
and the recovered catalyst could be recycled with no significant
decrease in catalytic performance. Our findings indicate that the
di-copper core in 1 plays an important role and that the formation
of the di-copper(II)–alkynyl species {Cu(II)2(μ-C≡CR)2} is the key
step in the oxidative alkyne–alkyne homocoupling.
[32] A. Lei, M. Srivastava, X. Zhang, J. Org. Chem. 67 (2002) 1969.
[33] I.J.S. Fairlamb, P.S. Bäuerlein, L.R. Marrison, J.M. Dickinson, Chem. Commun.
(2003) 632.
[
[
34] J. Li, Y. Liang, X.-D. Zhang, Tetrahedron 61 (2005) 1903.
35] J.-H. Li, Y. Liang, Y.-X. Xie, J. Org. Chem. 70 (2005) 4393.
[36] K. Kamata, S. Yamaguchi, M. Kotani, K. Yamaguchi, N. Mizuno, Angew. Chem.
Int. Ed. 47 (2008) 2407.
[37] D.D. Perrin, W.L.F. Armarego (Eds.), Purification of Laboratory Chemicals, third
ed., Pergamon Press, Oxford, 1988, pp. 80–388.
[38] L. Casella, O. Carugo, M. Gullotti, S. Garofani, P. Zanello, Inorg. Chem. 32 (1993)
2056.
[39] I.D. Brown, D. Altermatt, Acta Crystallogr. Sect. B: Struct. Sci. 41 (1985) 244.
40] N.E. Brese, M. O’Keeffe, Acta Crystallogr. Sect. B: Struct. Sci. 47 (1991) 192.
[
[
41] D.K. Lyon, W.K. Miller, T. Novet, P.J. Domaille, E. Evitt, D.C. Johnson, R.G. Finke,
J. Am. Chem. Soc. 113 (1991) 7209.
[42] D.-L. Long, C. Streb, Y.-F. Song, S. Mitchell, L. Cronin, J. Am. Chem. Soc. 130
(2008) 1830.

130
K. Yamaguchi et al. / Journal of Catalysis 258 (2008) 121–130
[
43] F. Bohlmann, H. Schönowsky, E. Inhoffen, G. Grau, Chem. Ber. 97 (1964) 794.
[46] M. Allevi, M. Bonizzoni, L. Fabbrizzi, Chem. Eur. J. 13 (2007) 3787.
[47] V.W. Yam, K.K. Lo, W.K. Fung, C. Wang, Coord. Chem. Rev. 17 (1998) 17.
[48] B.J. Hathaway, D.E. Billing, Coord. Chem. Rev. 5 (1970) 143.
[49] K.A. Connors, Chemical Kinetics, The Study of Reaction Rates in Solution, VCH
Publishers, New York, 1990, pp. 292–305.
[44] J.E. Pate, P.K. Ross, T.J. Thamann, C.A. Reed, K.D. Karlin, T.N. Sorrell, E.I. Solomon,
J. Am. Chem. Soc. 111 (1989) 5198.
[45] S. Sen, S. Mitra, D.L. Hughes, G. Rosair, C. Desplenches, Polyhedron 26 (2007)
1740.