Fast, ligand- and solvent-free synthesis of 1,4-substituted buta-1,3-diynes by cu-catalyzed homocoupling of terminal alkynes in a ball mill

Article abstract of DOI:10.1002/chem.201100604

A method for the Glaser coupling reaction of alkynes by using a vibration ball mill has been developed. The procedure avoids the use of ligands and solvents during the reaction. Aryl- and alkyl-substituted terminal alkynes undergo homocoupling if coground with KF-Al₂O₃ and CuI as a milling auxiliary and catalyst. Furthermore, an alternative protocol has been developed incorporating 1,4-diazabicyclo[2.2.2]octane (DABCO) as an additional base allowing the use of KF-Al₂O₃ with a lower KF loading. Besides Cu salts, the homocoupling of phenylacetylene is also catalyzed by Ni or Co salts, as well as by PdCl₂. TMS-protected phenylacetylene could be directly converted into the homocoupling product after in situ deprotection of the alkyne by fluoride-initiated removal of the trimethylsilyl group.

Full text of DOI:10.1002/chem.201100604

FULL PAPER
DOI: 10.1002/chem.201100604
Fast, Ligand- and Solvent-Free Synthesis of 1,4-Substituted Buta-1,3-diynes
by Cu-Catalyzed Homocoupling of Terminal Alkynes in a Ball Mill
[
a]
[a]
[a]
[a]
Robert Schmidt, Rico Thorwirth, Tony Szuppa, Achim Stolle,*
[
a]
[b]
Bernd Ondruschka, and Henning Hopf
Abstract: A method for the Glaser
coupling reaction of alkynes by using a
vibration ball mill has been developed.
The procedure avoids the use of li-
gands and solvents during the reaction.
Aryl- and alkyl-substituted terminal al-
kynes undergo homocoupling if co-
ground with KF–Al O and CuI as a
more, an alternative protocol has been
developed incorporating 1,4-
diazabicyclo[2.2.2]octane (DABCO) as
KF–Al O with a lower KF loading.
2 3
Besides Cu salts, the homocoupling of
phenylacetylene is also catalyzed by Ni
AHCTUNGTRENNUNG
an additional base allowing the use of
or Co salts, as well as by PdCl . TMS-
2
protected phenylacetylene could be di-
rectly converted into the homocoupling
product after in situ deprotection of
the alkyne by fluoride-initiated remov-
al of the trimethylsilyl group.
Keywords: alkynes · ball milling ·
copper · cross-coupling · homocou-
pling · materials science
2
3
milling auxiliary and catalyst. Further-
[
9]
Introduction
Glaser reaction in the absence of a solvent. The best re-
sults occurred in the presence of KF–Al O (40 w% KF)
2
3
The first homocoupling of terminal alkynes was reported by
and CuCl . Within very short reaction times they achieved a
2
[1]
Glaser in 1869. He discovered that the dimerization of
phenylacetylene to form 1,4-diphenylbuta-1,3-diyne pro-
ceeds with high selectivity in the presence of copper salts
and air as the oxidant. Similar products with a buta-1,3-
diyne moiety are now widely applied within materials sci-
ence and molecular electronics, for the synthesis of poly-
75% yield for the model reaction with phenylacetylene, but
an excess of the copper salt (3.7 equiv) was necessary. Shari-
fi et al. optimized this reaction by reducing the requirement
for the copper salt to 5 mol% and replacing the fluoride
[10a]
loaded alumina with neutral Al O .
However, morpho-
2
3
line was then required for this procedure. They also report-
ed a coupling reaction with good yield by stirring the alkyne
[2,3]
mers, supramolecular materials, and drugs.
In general, this reaction requires an organic solvent, for
example, methanol or pyridine, to take place in noticeable
quantities. Focused research to improve the classical proce-
dure has led to interesting variations of the original method.
Thus, the quantity of volatile or hazardous organic solvents
with morpholine, KF–Al O , and Cu AHCTUNGTNERNUG( OAc) at room temper-
2 3 2
[10a]
ature for 3 h.
In comparison with the microwave en-
hanced route, the yield of 96% is an improvement, but the
reaction time (3 h) is greatly increased.
The step from a solvent-free synthesis in a microwave
oven to comminution in ball mills is often a small one. Ball
mills have a range of applications in materials science, in
which they are used to grind and homogenize, for example,
minerals, or in the manufacturing of catalysts and nanoparti-
[4]
can be reduced by the employment of supercritical CO ,
2
[
5]
[6]
ionic liquids, (near-critical) water, or polyethylene glycol
PEG) as the solvent. However, protocols are also known
[7]
(
[8–12]
that totally omit the solvent.
The first solvent-free proce-
[13]
dure for an alkyne homocoupling involved a CuCl –pyridine
cles. However, ball mills have recently become interesting
tools for organic synthesis because of the relatively mild re-
2
[8]
complex as the catalyst. Kabalka et al. took advantage of
dielectric heating in a microwave oven to perform the
[14]
action conditions they allow. Examples of the use of ball
[
15]
mills are found for carbonyl reactions,
in fullerene
[16]
[17]
[18]
chemistry,
and Sonogashira cross-coupling reactions.
the synthesis of peptide building blocks and heterocycles by
and in Suzuki–Miyaura,
Mizoroki–Heck,
[
19]
[
a] R. Schmidt, R. Thorwirth, Dr. T. Szuppa, Dr. A. Stolle,
Prof. B. Ondruschka
Additionally,
Institute for Technical Chemistry and Environmental Chemistry
Friedrich-Schiller University Jena
Lessingstrasse 12, 07743 Jena (Germany)
Fax : (+49)3641-948402
[
20]
this type of procedure has been reported.
metal catalysis, Pd is generally applied as the catalytically
Regarding
[17–19]
active component.
Only in a few cases are other metals
E-mail: Achim.Stolle@uni-jena.de
employed; examples of these are 1) InCl for naphthopyrane
3
[
b] Prof. H. Hopf
[21]
synthesis, 2) Al flakes for intermolecular Aldol reactions
of valeraldehyde, 3) copper or CuI as a cocatalyst in So-
nogashira cross-coupling reactions that use K CO as the
and 4) copper salts as catalysts in Huisgen 1,3-dipo-
lar cycloaddition reactions between alkynes and azides.
Institute for Organic Chemistry
Technical University Braunschweig
Hagenring 30, 38106 Braunschweig (Germany)
[22]
2
3
[19a]
base,
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201100604.
[23]
Chem. Eur. J. 2011, 17, 8129 – 8138
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8129

Interestingly, the last two, which involve Cu-(co)catalyzed
reactions, lack the occurrence of homocoupling products, de-
spite of the fact that Cu is present and may provoke this re-
action pathway.
Table 1. The correlation between the milling auxiliary and the yield of
[
a]
1
,4-diphenylbuta-1,3-diyne (2a; Scheme 1).
[
b]
[c]
Entry Type of solid reagent system
Yield of 2a [%]
[
d]
1
2
3
4
5
6
7
8
9
0
1
commercial KF–Al
commercial KF–Al
commercial KF–Al
commercial KF–Al
2
2
2
2
O
O
O
O
3
3
3
3
42
96
89
99
0
0
0
0
0
31
76
53
0
[
d,e]
[
e]
Results and Discussion
a-Al
2
O
3
(acidic)
(neutral)
(basic)
g-Al
g-Al
2
O
O
3
Recently, a procedure for the Sonogashira cross-coupling re-
action in a ball mill was reported by using a Pd catalyst and
2
3
quartz sand
silica gel 60
1
,4-diazabicyclo
ACHTUNGTRENNUNG[ 2.2.2]octane (DABCO) as the base for de-
1
1
KF–Al
KF–Al
KF–Al
2
2
2
O
O
O
3
3
3
(a; acidic)
(g; neutral)
(g; basic)
[19b]
protonation of the terminal alkyne.
Solution-based proce-
I
dures for this reaction often employ Cu as the cocatalyst,
yielding Cu–acetylides as the reaction intermediates. In con-
trast, the ball mill variation runs in the absence of copper.
12
13
14
15
16
KF–quartz sand
KF–silica gel 60
0
[f]
KF–Al
KF–Al
2
O
O
3
(g; neutral), impregnation method 91
(g; neutral), calcined
I
Furthermore, the presence of Cu salts yielded product mix-
[f]
2
3
46
tures showing a decreased selectivity for the cross-coupled
product. Instead, an increase in the amount of alkyne homo-
coupling product (the Glaser reaction) was reported for the
reaction in ball mills in the presence of strong bases like
[
a] Reaction conditions: 1a (2 mmol), CuI (5 mol%), 4 g reagent system,
ZrO beaker (35 mL), 12ꢁZrO milling balls (10 mm), nosc =30 Hz, t=
2
2
10 min. [b] Unless otherwise stated the KF content was 40 w%. [c] Found
by GC-FID analysis after intermediate workup. Product selectivity
>99%. [d] KF content=32 w%. [e] Reaction was carried out in the pres-
ence of DABCO (2.5 mmol). [f] Impregnation and calcination performed
by following to reference [24c].
[19b]
DABCO.
In the absence of strong bases no effect on the
selectivity for the products due to homocoupling was repor-
[
19a,23]
ted.
These results motivated us to develop a reaction protocol
for the Glaser reaction in a ball mill. Thus, the challenge
was to find catalysts that allowed the transformation of ter-
minal alkynes (1) into the corresponding 1,4-substituted
buta-1,3-diynes (2, Scheme 1) in a time- and energy-efficient
achieved with these reagents if the reactions were carried
out in the presence of DABCO (2.5 mmol) as an additional
organic base (entries 2 and 4, Table 1). The application of
KF–Al O in solvent-free reactions appears prominently for
2
3
[17d–e,24]
many types of transformation.
The present results
confirm the general ability of KF–Al O reagent systems to
2
3
perform this type of reaction. In comparison to the homo-
coupling reaction procedure involving grinding the reactants
[10b]
by using a stirring bar,
morpholine has been replaced by
the less toxic base DABCO or the reaction occurs without
additional base. Microwave-assisted reaction protocols for
the Glaser reaction that run with KF–Al O in the absence
Scheme 1. Generic reaction scheme for the Glaser coupling reaction of
terminal alkynes (1) to form 1,4-substituted buta-1,3-diynes (2) initiated
by ball milling (model reaction: 1a to 2a, R=Ph).
2
3
of an organic base suffer from the fact that the catalyst
(
CuCl ) has to be applied in greater than stoichiometric
2
[9]
quantities.
manner. The reactions were carried out in a ball mill by
The general applicability of KF–Al O for the synthesis of
2 3
using zirconia (ZrO ) as the material for the milling beakers
2a has thus been demonstrated. However, the commercial
2
and balls. The reactions were carried out by using the maxi-
mum oscillation frequency (nosc =30 Hz) of the ball mill
used in the experiments.
reagents are based on different alumina modifications. Thus,
physical mixtures of acidic a-, neutral g-, and basic g-Al O3
2
with a KF content of 40 w% and similar particle size distri-
butions have been applied in the same reaction. Neutral g-
Al O furnished the highest yields under the experimental
Influence of the base and milling auxiliary: The first chal-
2
3
lenge was to identify the best base for the homocoupling re-
conditions (entries 10–12, Table 1). Rather than simply
[10a]
action. In the presence of CuI as the catalyst,
the reac-
mixing KF and Al O with 1a and the catalyst (CuI), KF–
2
3
tion of phenylacetylene (1a) to form 1,4-diphenylbuta-1,3-
diyne (2a) was used as the model system to find bases suit-
Al O can also be prepared by an impregnation method.
2 3
This material was prepared by suspending neutral g-Al O3
2
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
able for the reaction (Table 1). Oxides like silica and alumi-
in an aqueous KF solution with subsequent removal of the
[9,10b,17c–e,24c,25]
na, which showed high activity in the Sonogashira reaction
in presence of DABCO,
both in the presence and absence of DABCO (entries 5–9,
Table 1). The use of commercially available KF–Al O with
solvent in vacuo.
Simple physical mixing of the
[19b]
failed in the model reaction
reactants afforded lower yields for both concentrations of
KF than commercial KF–Al O , but the reaction with KF–
2
3
Al O prepared by the impregnation method yielded 91%
2
3
2
3
3
2 or 40 w% KF furnished 2a in 42 and 89% yield, respec-
2a after 10 min of ball milling, which is in the same range as
with the respective commercial reagent (89%). Calcina-
[26]
tively (entries 1 and 3, Table 1). Quantitative yields were
8130
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8129 – 8138

Solvent-Free Glaser Coupling
FULL PAPER
tion of the synthesized KF–Al O for 2 h at 3008C prior to
Table 2. Screening of bases for the Glaser reaction of phenylacetylene
2
3
[
a]
(
1a; Scheme 1).
Base
without additional base
,4-diazabicyclo[2.2.2]octane
triethylamine
application in the homocoupling of 1a to form 2a
Scheme 1) resulted in moderate yields (46%). This result is
in accordance with the deactivation reported for Pd(OAc) -
catalyzed Suzuki–Miyaura cross-coupling reactions in a
[
b]
Yield of 2a [%]
(
A
H
U
G
R
N
U
G
42
96
51
0
2
1
ACHTUNGTRENNUNG
[24c]
planetary ball mill.
tetrabutylammonium bromide
Considering the mass content of 40 w% KF in the KF–
Al O reagent and the applied amount in the reaction (4 g),
KOH
K CO
3
54
43
2
3
2
27.5 mmol of fluoride is available to deprotonate the termi-
[
a] Reaction conditions: 1a (2 mmol), base (2.5 mmol), CuI (5 mol%),
nal alkyne and initiate the reaction. Replacing the alkali
cation and keeping the fluoride-content constant by using
CsF as the fluoride source gave a 32% yield of 2a, whereas
KF–Al O (4 g, 32 w%, commercial), ZrO2 beaker (35 mL), 12ꢁZrO2
milling balls (10 mm), nosc =30 Hz, t=10 min. [b] Found by GC-FID anal-
ysis after intermediate workup. Product selectivity >99%.
2
3
the application of NaF and MgF failed to facilitate the re-
2
[27]
action. The employment of more complex fluoride anions,
such as NH [BF ] and Na [AlF ] (kryolithe), was also unsuc-
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
A
H
U
G
R
N
U
G
The positive effect of N-coordinating bases on this reaction
4
4
3
6
cessful. Thus, the combination of KF and neutral g-Al O
seems to be required for the Glaser reaction in a ball mill,
is also stressed by mechanistic DFT calculations at the
2
3
[29]
B3LYP/LACVP level of theory.
These calculations indi-
since the reagent combination KF–SiO (entries 13 and 14,
cated that no solvent molecules are involved in the Cu-cata-
lyzed homocoupling of alkynes. The total inactivity of tetra-
butylammonium bromide (TBAB) and the unobserved addi-
2
Table 1) also failed. As indicated previously, the alumina
[
24c]
modification
is as important to the reaction as the KF
[
17c,24c,28]
content if working with KF–Al O reagents.
The cor-
relation between the yield of homocoupling product 2a
Y ) and the mass fraction of KF (wKF) in the solid reagent
tional effect of triethylamine (NEt ) can be traced back to
2
3
3
[11]
different origins. Since NEt is a liquid, its incorporation
3
(
in the reaction is less successful than that of the solid base
2a
system is illustrated in Figure 1. A clear linear relationship
DABCO. In contrast to DABCO and Et N, TBAB has no
3
free coordination site on the nitrogen atom, making it a less
active base and hindering its interaction with both copper
and the alkyne in the transition state.
The influence of oxygen and catalyst: The generally accept-
ed reaction mechanism for the catalytic cycle of the Glaser
reaction involves oxidative regeneration of the Cu catalyst
[3a,29,30]
after reductive elimination of the 1,3-diyne product.
Generally, the reactions are carried out in air either under
active mass transport of the oxidant into the solution (sol-
vent saturation) or by passive exposure to air. The latter
[8–12]
technique is favored in solvent-free methods.
One limi-
tation of ball milling technology is the restricted volume of
the reactor. Assuming material densities for KF (60 w%), g-
À3
Al O , and zirconia (milling balls) of 2.5, 4.0, and 6 gcm ,
2
3
Figure 1. The correlation between KF content and yield of 2a (see
Scheme 1) resulting from homocoupling of 1a (for reaction conditions,
respectively, a free reaction volume in the milling beaker of
8 mL can be calculated, excluding the amount of catalyst
and substrate. Due to preparation of the reaction mixtures
in an air atmosphere, approximately 0.25 mmol of oxygen
2
see Table 1; reagent system: KF–Al
up).
2 3
O (g; neutral), intermediate work-
(
as O ) are present for reoxidation of the catalyst. Thus, an
2
between Y2a and wKF is shown over a data range that almost
reaches quantitative yield with a KF loading of 60 w%. This
behavior is also in line with the application of the commer-
cial reagent systems listed in Table 1, entries 1 and 3.
oxidant deficiency is normally present that will prevent the
reaction from proceeding to completion under standard con-
ditions. Opening of the reactor after milling exposes the re-
action mixture to air and incorporates oxygen into the reac-
tion mixture. This fast post-milling process probably ac-
counts for the high yields discussed in Tables 1 and 2, as
well as in Figure 1. To guarantee reproducibility, the reac-
tion of 1a to form 2a (Scheme 1) was performed with differ-
ent metal(II) chlorides (M=Cu, Ni, Co, Pd), as well as with
CuI. Samples of the reaction mixture were withdrawn from
the milling beakers immediately after opening and after var-
[31]
However, the application of low loadings of KF–Al O re-
2
3
quired the presence of an additional organic base to drive
the reaction to completion in a reasonable time. Table 2 dis-
plays the influence of bases on the reaction of 1a to form
2
a with commercial KF–Al O (32 w%). Alternative organ-
2 3
ic (NEt ) and inorganic (KOH, K CO ) bases to DABCO
3
2
3
showed only minor effects on product yield compared to the
reaction with only KF–Al O . However, the selectivity for
ious air-exposure times (Figure 2). Apart from PdCl , the
2
3
2
2
a was almost completely independent of the base applied.
metal salts revealed similar behavior. The samples taken im-
Chem. Eur. J. 2011, 17, 8129 – 8138
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8131

A. Stolle et al.
inactive in the homocoupling process. Further exposure to
air leads to an exponential increase in the yield of 2a. Com-
pared to other Pd-catalyzed homocoupling reactions, the re-
[33]
action proceeded in the absence of a cocatalyst and did
[11]
not need the addition of p-iodophenol. After 120 min of
exposure to air no starting material was detected and 2a
was formed in only 40% yield. Instead, oligomerization of
1
a was detected. Two trimers of 1a have been identified
from the reaction mixture: 1,2,4-triphenyl- (3) and 1,3,5-tri-
phenylbenzene (4; Scheme 2). Identification of 3 and 4 was
Figure 2. Dependence of the yield of 2a (see Scheme 1) on the time the
reaction mixture is exposed to air for different metal salts as the catalyst
(
phenylacetylene (2 mmol), catalyst (5 mol%; ^ =CuCl
CoCl , ꢀ =PdCl , *=CuI), KF–Al (4 g, g-Al , neutral, 60 w% KF),
ZrO beaker (35 mL), 12ꢁZrO milling balls (10 mm); nosc =30 Hz, t=
0 min).
2 2
, & =NiCl ,
~ =
Scheme 2. The cyclotrimerization products of phenylacetylene (1a) iden-
tified in the product mixture of the reaction carried out in the presence
of PdCl (for the reaction conditions, see Figure 2).
2
2
2
2
O
3
2 3
O
2
2
1
difficult, since their concentration in the reaction mixture
was relatively low and isolation was therefore not possible.
However, comparison of mass spectra (see the Supporting
Information) and retention times in GC-MS, as well as GC-
FID, analyses with those from reference compounds was
successful. Reference compound 3 was synthesized from
mediately after opening the reaction vessel indicate that Y2a
ranges from 28 (NiCl ) to 58% (CuCl ). After exposure to
2
2
air for 30 min, CuCl and CuI provided quantitative yields,
2
whereas reactions with NiCl and CoCl needed three times
2
2
as long to reach a similar level of conversion.
In contrast to 1,3-dipolar cycloaddition between alkynes
and azides, the oxidation state of the Cu catalyst seems to
be essential for the homocoupling in a ball mill. Comparing
1,2,4-tribromobenzene and phenylboronic acid by Suzuki–
[23]
[34]
Miyaura cross-coupling (Scheme 3).
The gas chromato-
gram of the crude reaction mixture from the PdCl -catalyzed
2
the initial values for CuCl (58%) and CuI (32%) proves
reaction after 120 min of exposure to air is pictured in
Figure 3. The results indicate that 3 and 4 are the only cyclic
trimers formed.
2
the influence of the oxidation state of the copper salt.
Cupric chloride converts twice the amount of reactant
before it is reduced to Cu and requires reoxidation for fur-
ther reaction cycles (see also Table 3 shown later). Similar
experiments with ferric and ferrous chloride led to no con-
[32]
version at all over the time range. For the Cu-, Ni-, and
Co-catalyzed reactions a pseudo-zero-order reaction rate
was determined concerning the formation of 2a. This is typi-
cal for reactions suffering from mass-transport limitations,
such as diffusion of the oxidant towards the catalyst.
Experiments with CuI in the presence of different inor-
ganic oxidants afforded significantly higher yields than
under only the enclosed air. The addition of Oxone
Scheme 3. Reference synthesis for the identification of compound 3 in
the Pd-catalyzed reaction of 1a (see Figure 2).
The ratio of 3 to 4 is 5:4 (Figure 3) showing that forma-
tion of the symmetrical structural isomer is disfavored. The
formation of cyclotrimerization products from terminal al-
(
0.5 mmol), (NH ) S O (1 mmol), or aqueous H O (100 mL,
4 2 2 8 2 2
6
2
0 w%) to the reaction mixture before ball milling furnished
a in 64, 72, and 98% yield, respectively. These results are
0
II
[35]
kynes catalyzed by Pd and Pd is well studied. Besides
trimerization, oligo- and polymerization have been reported
in clear contrast to the inactivity of CuI in combination with
Na BO ” as the oxidant in the microwave-assisted homo-
“
with [Pd ACHTUNGTRNENUG( dba) ] (dba=dibenzylideneacetone) and in the ab-
2 3
2
4
[9]
[35a]
coupling of 1a, in which the application of the reagents in
both 3.7:1 and 7.4:1 ratios compared to 1a provided only
sence of phosphine ligands.
In the present case, a thermal
[2+2+2] cycloaddition can be ruled out since reaction in the
absence of the catalyst did not show any conversion of 1a.
Other metal salts tested for the reaction (see Table 3) did
not yield side-products 3 and 4. In contrast to the reactions
with the Cu, Ni or Co salts (Figure 2), the reaction in the
[9]
9
% 2a compared to 17% in the absence of the oxidant.
As already demonstrated in similar Sonogashira reactions,
PdCl is inactive in the transformation of terminal alkynes
2
in a ball mill if the reaction mixture is subsequently worked
[
19b]
up after the reaction (Figure 2).
Thus, only small amounts
presence of PdCl is of pseudo-first order with regard to for-
2
0
of 2a are formed, along with the formation of Pd , which is
mation of 2a.
8132
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8129 – 8138

Solvent-Free Glaser Coupling
FULL PAPER
Table 3. The influence of different metal salts employed as catalysts on
[
a]
the yield of 1,4-diphenylbuta-1,3-diyne (2a; Scheme 1).
[
b]
[e]
Yield of 2a [%] for workup procedure:
[
c]
[d]
Catalyst
A
B
C
Cu
0
40
39
32
50
59
58
45
84
4
28
14
27
23
11
50
29
0
0
90
69
0
CuCl
CuBr
CuI
CuOAc
CuCl
Cu
Cu
Cu
CuO
NiCl
NiBr
62
64
96
66
73
72
74
89
4
37
32
40
40
45
57
37
–
>99
97
>99
92
69
>99
14
87
67
72
56
44
70
49
0
2
A
H
U
G
R
N
U
G
2
A
H
N
T
N
U
G
2
A
H
U
T
E
U
G
3 2 2
) ·3H O
2
Figure 3. Gas chromatogram of the crude reaction mixture resulting from
2
the PdCl
conditions see Figure 2) and the identification of cyclic trimers 3 and 4
Scheme 2).
2
-catalyzed Glaser reaction of 1a (exposure time=120 min; for
Ni
Ni
Ni
A
T
N
T
E
U
G
2 2
·4H O
A
H
U
G
R
N
U
G
2
(
A
H
U
G
R
N
N
3 2 2
) ·6H O
CoCl
Co(acac)
FeCl
FeSO
ferrocene
FeCl
[Fe(CN)
PdCl
RuCl
2
A
H
U
G
R
N
N
3
2
4
0
0
0
0
2
0
0
0
0
0
4
0
–
–
–
–
–
0
Screening of catalysts: Based on the aforementioned evalua-
tion of the reagent system (KF content, fluoride source,
Al O modification), the presence of an additional base and
3
K
3
6
]
2
3
the air-exposure time after ball milling, three different reac-
tion procedures have been used for screening different
metal catalysts (Table 3). Procedures A and B were carried
out with KF–Al O (g, neutral, 4 g, 60 w%); in the former
2
3
·H
2
O
[
a] Reaction conditions: 1a (2 mmol), catalyst (5 mol%), KF–Al (4 g,
2
O
3
g-Al , neutral, 60 w% KF), ZrO beaker (35 mL), 12ꢁZrO milling
2
O
3
2
2
2
3
balls (10 mm), nosc =30 Hz, t=10 min; acac=acetylacetone. [b] Found by
GC-FID analysis. Product selectivity >99%. [c] A: immediate work up
after the reaction. [d] B: after milling the mixture was exposed to air for
30 min without further mixing. [e] C: the reaction was carried out with
the product mixture is immediately worked up, whereas in
the latter the mixture is exposed to air for a further 30 min.
Procedure C involves the employment of KF–Al O (com-
2
3
2 3
KF–Al O (32 w%, commercial) and DABCO (2.5 mmol) and subse-
mercial, 4 g, 32 w%) and DABCO (2.5 mmol). Following
procedure A the mixture is extracted immediately after
opening the milling beaker. Common metal salts were
tested for their applicability as a catalyst in the homocou-
pling of phenylacetylene (1a) to form 1,4-diphenylbuta-1,3-
diyne (2a; Scheme 1) in a mixer ball mill. Results for this
variation are summarized in Table 3. The general trend, al-
ready discussed above in connection with Figure 2, was also
observed in the case of metal salts with other anions, provid-
ed that the reaction worked at all. Thus, procedure B lead
to higher yields than A. Compared to the reaction in the
presence of the diamine DABCO (C), method B lead to su-
perior results in most cases.
quently worked up.
at all with Ni and Co salts. Thus, the first solvent-free proce-
dures for the homocoupling of 1a in the presence of these
catalysts have been found. Additionally, the reaction did not
require Cu salts as cocatalysts, nor the presence of ligands,
and the catalyst concentrations are relatively low, compared
[36]
[37]
to literature procedures working with Ni, Co, and Pd
[11,33b,c]
catalysts.
Substrate screening: Based on the screening reactions
(Scheme 1; variation of milling auxiliary, catalyst, and base),
the scope of the reaction was then extended to the coupling
of different alkynes. Thus the influence of substituents on
the course of the reaction has been studied (Table 4) by
using procedures B and C. Although other aryl- and alkyl-
substituted terminal alkynes (1a–i) were suitable for Glaser
coupling following both procedures, 3-ethynylpyridine (1j)
showed no conversion if the reaction was carried out in the
presence of DABCO (procedure C). However, the effects of
electron-donating substituents like methyl or methoxy
groups on conversion were recognizable. With these sub-
strates, lower yields of buta-1,3-diynes were observed. By
doubling the amount of CuI (10 mol%) and the reaction
time (20 min), the less reactive alkynes (1b, c, e, h, and i)
could also be converted quantitatively. However, all of the
alkynes preserve excellent selectivity (>96%) for the homo-
Table 3 demonstrates that the formation of 2a by homo-
coupling of 1a (Scheme 1) can be catalyzed by various Cu,
Ni, and Co salts, whereas no reaction occurred in the pres-
ence of ferrous and ferric salts or RuCl . PdCl afforded
3
2
only low yields. The reactivity of the different metals de-
creases in the order: Cu>NiꢀCo@Pd. Generally, the selec-
tivity for the Glaser coupling product (2a) was nearly
100%, with no side-products being found regardless of the
type of catalyst. Thus, the obvious differences in yields are
traced back to different degrees of conversion. The applica-
tion of CuO or copper metal (powder) afforded very low
conversion. In situ comproportionation of CuO and Cu in
equal amounts, furnishing Cu O, yielded minor amounts of
2
2
a. Despite the fact that CuI and Cu
ACHTUNGTNERNU(GN NO ) are the most
3 2
suitable catalysts, it is astonishing that the reaction proceeds
Chem. Eur. J. 2011, 17, 8129 – 8138
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8133

A. Stolle et al.
Table 4. Screening of different terminal alkynes (1) in the homocoupling
the presence of DABCO increased the reactivity of the mix-
ture, as already shown in Table 1.
[
a]
furnishing buta-1,3-diynes (2; Scheme 1).
Workup procedure B
Yield Selectivity
%]
[
b]
[c]
Workup procedure C
R group
2
Yield
[%]
Selectivity
[
[%]
[%]
Table 5. Cryogenic milling of phenylacetylene (1a) to yield 1,4-diphenyl-
[
a]
buta-1,3-diynes (2a; Scheme 1).
C
C
C
6
6
6
H
H
H
5
5
5
2a
2a
2a
2b
2c
2d
2e
2 f
2g
2h
2i
>99
3
>99
98
–
96
13
1
>99
99
99
[
[
d]
e]
Exposure
time [min]
KF–Al
2
O
3
(60 w%)
KF–Al
DABCO (2.5 mmol)
2 3
O (32 w%),
[b]
–
[
f]
o-Me-C
p-Me-C
o-MeO-C
m-MeO-C
p-MeO-C
p-F-C
C
C
6
H
H
4
88
88
72
80
68
90
92
89
88
97
95
89
91
90
98
93
91
92
99
77
92
98
70
93
96
89
0
>99
>99
>99
>99
>99
>99
>99
>99
–
Yield [%]
Selectivity [%]
Yield [%]
Selectivity [%]
6
4
0
0
0
3
12
55
>99
>99
98
13
38
88
>99
98
90
[
[
f]
f]
6
H
H
4
1
3
6
4
6 4
H
[
a] Reaction conditions: 1a (2 mmol), CuI (5 mol%), KF–Al
2 3
O (4 g, g-
6
H
4
[
[
f]
f]
Al O , neutral), ZrO beaker (35 mL), 12ꢁZrO2 milling balls (10 mm),
H
17
2
3
2
8
closed milling beakers were cooled to the temperature of liquid nitrogen
prior to milling, nosc =30 Hz, t=10 min. [b] Reaction was carried out with
KF–Al O (32 w%, commercial).
2 3
10
H
21
3
-pyridyl
2j
[
a] Reaction conditions: 1 (2 mmol), catalyst (5 mol%), KF–Al (4 g,
2
O
3
g-Al , neutral, 60 w% KF), ZrO beaker (35 mL), 12ꢁZrO milling
2
O
3
2
2
balls (10 mm), nosc =30 Hz, t=10 min. [b] B: after milling the mixture was
exposed to air for 30 min without further mixing. [c] C: reaction was car-
ried out with KF–Al O (32 w%, commercial) and DABCO (2.5 mmol)
2 3
and subsequently worked up. [d] Cryogenic milling: milling beakers were
cooled with liquid nitrogen prior to ball milling. [e] Cryogenic milling
under constant cooling with liquid nitrogen. [f] Catalyst (10 mol%); t=
An attempt to cross-couple two different alkynes (1a and
h; Scheme 4) resulted in a mixture of all three possible
products in comparable yields, although the homocoupled
2
0 min.
coupling product. Considering the fact that only 5 mol% of
copper was used, the solvent-free conversion in the ball mill
is much more efficient than its solvent-free counterpart in a
microwave envisioned by Kabalka et al., for which
3
.7 equivalents of copper salt were required to achieve simi-
Scheme 4. The coupling reaction of phenylacetylene (1a) and decyne
(1h; reaction procedure C, see Table 4; full conversion).
[9]
lar yields.
The alkynes summarized in Table 4 are liquids at room
temperature (1a–e, h, and i) or possess a melting point
[35]
below 508C (1 f=28–29, g=26–27, and j=39-408C). Due
products 2a and h were slightly preferred. Similar reactions
to the fact that high-energy collisions of milling balls result
in a microwave preferentially gave rise to the cross-coupling
[17e]
[9,32,37c]
in heating of the reaction mixture up to 50–608C
these
product 2k.
cedure (CoBr , Zn, nitrobenzene, P
In comparison with the Co-catalyzed pro-
[37c]
reagents remain liquid or are adsorbed onto the solid mill-
ing auxiliary. Another option to solidify liquid reagents is
cryogenic milling, that is, the comminution of reactants
A
H
U
G
R
N
N
(OiPr) ),
the absence
2
3
of a defined coordination sphere allows the formation of the
statistical mixture of products only. These results indicate
that either greater than stoichiometric amounts of copper in-
itiate the cross-coupling or both modes of activation pass
through different activation pathways. In the presence of
high concentrations of Cu the possibility for the formation
of Cu–alkyne species is much higher. These species can un-
dergo oxidative coupling to the mixed product more easily.
In the present case it seems more plausible that both reac-
tion partners are coordinated to the same Cu center during
the course of the reaction. Extrusion of the metal in the
[15e,39]
cooled down to the temperature of liquid nitrogen
or
the performance of reactions in temperature-controlled re-
[40]
action vessels. The first method was applied to the synthe-
sis of 2a from 1a, yielding the product in 3 and 13% yields,
working with 60 w% KF–Al O or with the commercial
2
3
32%-loaded reagent with DABCO (2.5 mmol), respectively
(
Table 4). The lowered yield is in agreement with cryogenic
[15e]
milling experiments using the Knoevenagel condensation
[39]
and Tishchenko reaction procedures in a mixer and vibra-
tion ball mill, respectively. If the reaction is carried out
under constant cooling of the milling beakers with liquid ni-
trogen the reaction is totally suppressed and only 1% of 2a
was identified after ball milling. Prolonged standing at room
temperature resulted in unfreezing of the reactants. Thus,
reversible freezing of molecular motion does not inhibit the
particle refinement and the formation of new, active surfa-
ces at which the reaction takes place. The recovered molecu-
lar flexibility allowed the reaction to proceed (Table 5) and
[3a,29,30]
final step then provides the buta-1,3-diyne.
In addition, it was found that the nature of the alkyne and
the scope of the reaction can be further extended. It is, for
example, possible to use trimethylsilyl (TMS)-protected phe-
nylacetylene (5) directly in the coupling reaction
[9]
(Scheme 5). It is known that the TMS group can be
[41]
cleaved by KF deprotection procedures in solution. Mill-
ing of 5 in the presence of KF–Al O furnished the free
2
3
alkyne in situ and the Cu catalyst converted it to the homo-
8134
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8129 – 8138

Solvent-Free Glaser Coupling
FULL PAPER
[10a]
were mixed in a mortar and pestle prior to the reaction.
This step is not considered in the calculation of the catalyst
activity. Ball mills inherently have a very high mixing effi-
ciency. Thus, diffusion and mass transport problems are
avoided due to the fast movement of the milling balls
[
13c,d,44]
Scheme 5. In situ deprotection and Glaser coupling of TMS–acetylene (3;
reaction procedure C, see Table 4; 100% selectivity).
during the process.
Besides the practical advantages associated with the appli-
cation of ball mills to solvent-free organic syntheses, the
energy efficiency of such processes is also higher than for
comparable reactions carried out in microwaves or under
coupling product. Analysis indicates that no free phenylace-
tylene is present in the reaction mixture. However, conver-
sion of 5 was lower than that of free phenylacetylene (1a),
although the selectivity for the coupling product was still
high (95%). Thus, it can be concluded that the removal of
p-TMS is the rate-determining step. As well as deprotection
of 5, this reaction variant was possible with p-(TMS-ethy-
nyl)toluene and 4-(TMS-ethynyl)trifluoromethylbenzene,
but resulted in considerably lower yields of the homocou-
pling products. The importance of the fluoride ion in this re-
action is underlined when comparing the results with those
[
15e,17e,45]
the influence of ultrasound.
Performing the model
reaction with 2 mmol of 1a (Scheme 1) in a mixer ball mill
following protocol C furnished 2a in 96 and 13% yield for
normal and cryogenic milling, respectively (Table 6). Calcu-
[17e]
lation of the energy intensity
resulted in values of 8.9 and
À1
63.2 kWhmol . Scaling up of the reaction in a ball mill re-
quired a change of apparatus and thus a change in the type
of motion from oscillation (mixer ball mill) to rotation
(planetary ball mill). However, results indicated that scal-
ing up the reaction to 3 and 4.5 mmol has little effect on the
yield and TOF (Table 6). Keeping the reaction time the
same gave 78 and 80% yield of 2a. The energy intensity in-
creased slightly to 17.6 and 22.5 kWhmol for reactions on
a 3 and 4.5 mmol scale, respectively. Similar results have
also been found for the scaling up of Suzuki–Miyaura cross-
[46]
[19a]
reported recently by Mack and co-workers.
TMS-acety-
lene was applied as a substrate for Sonogashira reactions in
a ball mill by using K CO as the base instead a fluoride re-
2
3
À1
agent. Thus, no deprotection was reported after ball milling
for 17 h.
[17e]
Comparison of methods and energy entry: Comparison of
different reaction conditions that have been published for a
solvent-free Glaser reaction revealed the pre-eminence of
coupling reactions in ball mills.
tion of ball mills in particle refinement processes,
Compared to the applica-
[13c,d]
the
size of the milling balls did not affect the experimental re-
sults. Performing the reaction with 10ꢁ10 or 6ꢁ12 mm mill-
ing balls resulted in similar yields since the masses of the
milling balls are equal, which has also been proven for
[
9,10a]
reactions carried out in microwave ovens
with respect to reaction time (Table 6).
or ball mills
[
8–12]
Apart from the
solvent-free syntheses of 2a presented herein (Table 6), two
protocols exist wherein the formation of 2a is reported as a
[
17d]
[47]
Suzuki–Miyaura
ball mills.
and oxidation reactions carried out in
[
42]
side-reaction in Cu-catalyzed
metal-free
or DABCO-mediated
[43]
Sonogashira cross-coupling reactions. Maxi-
mum yields of 56 and 16% are achieved, respectively. Cal-
culation of the turn-over frequencies (TOF) for the different
protocols in Table 6 indicate that ball milling is as powerful
Conclusion
[10a]
À1
as microwave irradiation,
giving TOFs of 115 and 98 h ,
A new ligand- and solvent-free method for the Glaser cou-
pling reaction in a vibration ball mill has been developed.
Assessment of different metal salts identified Cu, Ni, and
respectively. However, it must be noted that in the micro-
À1
wave reaction protocol with a TOF of 98 h the reactants
Table 6. A comparison of different reaction conditions for solvent-free protocols for the Glaser reaction of phenylacetylene (1a) to form 1,4-diphenylbu-
[
a]
ta-1,3-diyne (2a; Scheme 1).
Reference
Reaction conditions
Catalyst
Base
Additive
n
1a
n
1a/ncatalyst
t
Yield of 2a
[%]
TON
TOF
À1
A[ mmol]
H
U
G
R
N
N
A[ min]
H
U
G
R
N
N
A[ h
C
H
T
U
N
G
T
R
E
N
N
U
N
G
]
[
b]
[
[
[
[
[
9]
microwave irradiation
microwave irradiation
CuCl
CuI
2
–
KF–Al
Al
KF–Al
p-iodophenol
–
KF–Al
KF–Al
KF–Al
KF–Al
2
O
O
3
3
1
1
1
1
1
2
2
3
0.27
20
5
100
33
20
20
20
20
8
10
180
720
360
10
10
10
10
75
82
96
84
96
96
13
78
80
<1
16
5
84
32
19
3
1.5
98
[
a]
10a]
10b]
11]
12]
morpholine
morpholine
Et
Et
DABCO
DABCO
DABCO
DABCO
2
O
3
[
a]
stirring
stirring
stirring
mixer ball mill
mixer ball mill
Cu
[Pd]
A
H
U
G
R
N
N
(OAc)
2
2
1.6
7.0
5.3
115
16
[
c]
3
N
N
CuCl
CuI
CuI
CuI
CuI
2
3
[
[
d]
this work
this work
this work
this work
2
2
2
2
O
O
O
O
3
3
3
3
d,e]
[
[
f]
planetary ball mill
planetary ball mill
16
16
94
96
g]
4.5
[
a] Reactants were preground with a mortar and pestle prior to reaction; n1a =amount of 1a, ncatalyst =amount of catalyst. [b] Due to greater than stoi-
chiometric employment of CuCl the metal salt is not a catalyst in this case. [c] Immobilized Pd complex. [d] Reaction variant C (see Table 4). [e] Cryo-
genic milling: milling beakers were cooled with liquid nitrogen prior to ball milling. [f] Batch size: 1.5ꢁvariant C in Table 4, ZrO beaker (45 mL), 18ꢁ
2
2
ZrO
2
milling balls (10 mm), nrot =13.3 Hz, t=10 min. [g] Batch size: 2.25ꢁvariant C in Table 4, ZrO
2
beaker (80 mL), 30ꢁZrO
2
milling balls (10 mm),
n
rot =13.3 Hz, t=10 min.
Chem. Eur. J. 2011, 17, 8129 – 8138
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8135

A. Stolle et al.
Co compounds as potential catalysts for application in the
homocoupling of phenylacetylene, although CuI was identi-
fied as the most effective catalyst. In combination with KF–
was evaporated in vacuo, the crude products were dried and analyzed by
GC-FID and GC-MS. Analytical samples for NMR investigation were
isolated by column chromatography using n-hexane/toluene mixtures.
1
Products were identified according to literature data. Both H and
Al O as the base, high selectivity and conversion have been
13
2
3
C NMR spectra are provided in the Supporting Information.
General reaction procedure C: The milling beakers were filled with the
milling balls and then KF–Al (32 w% KF, 4 g), the alkyne (2 mmol),
achieved in the reactions of various terminal alkynes with
hetero)aromatic, as well as aliphatic, substituents. Variation
of the KF content in the KF–Al O reagent and alteration
(
2 3
O
DABCO (2.5 mmol, 280 mg), and CuI (5 mol%, 18 mg) were added se-
quentially. Milling was performed at 30 Hz for 10 min. After cooling to
room temperature, the crude products were extracted on a frit with a
thin silica layer by using chloroform (3ꢁ10 mL). The solvent was evapo-
rated in vacuo, the crude products were dried and analyzed by GC-FID
and GC-MS. Analytical samples for NMR investigation were isolated by
column chromatography using n-hexane/toluene mixtures. Products were
identified as described above.
2
3
of the alumina modification revealed a major influence of
both variables on the product yield, although the selectivity
remained unaffected. The use of 60 w% KF and neutral g-
alumina allowed the reaction to be performed in the ab-
sence of additional base, whereas lower loadings of these re-
agents required the presence of DABCO as an additional
base. The complications connected with the lack of oxidant
present in the closed reaction vessels were vanquished by al-
lowing post-reaction exposure to air. Additionally, the appli-
cation of peroxides as oxidants led to similar results. Besides
the homocoupling of terminal alkynes, the method is also
suitable for the in situ deprotection of TMS-protected al-
kynes.
[
48]
1
1
7
,4-Diphenylbuta-1,3-diyne (2a):
H NMR (200 MHz, CDCl
3
): d=
):
13
.55–7.47 (4H, m), 7.37–7.27 ppm (6H, m); C NMR (50 MHz, CDCl
3
d=132.50, 129.20, 128.44, 121.83, 81.53, 73.90 ppm; GC-MS: m/z (%):
202 (100).
[
49] 1
1,4-Bis
.47 (2H, d, J=7.5 Hz), 7.29–7.07 ppm (6H, m); C NMR (50 MHz,
CDCl ): d=141.61, 132.90, 129.55, 129.08, 125.64, 121.70, 81.11, 77.22,
0.70 ppm; GC-MS: m/z (%): 230 (100).
A
H
U
G
R
N
U
G
3
H NMR (200 MHz, CDCl ): d=
1
3
7
3
2
[48] 1
1
,4-Bis
A
H
U
G
E
N
N
(p-tolyl)buta-1,3-diyne (2c):
3
H NMR (200 MHz, CDCl ): d=
Compared with other solvent-free reaction protocols for
alkyne homocoupling, the reaction conditions are mild with
respect to the reaction time and temperature. Thus, the
method is competitive with microwave-assisted protocols
with the advantage that the reactions can be accomplished
without the need for mixing the reagents prior to the reac-
tion.
7
.39 (4H, d, J=8.2 Hz), 7.09 (4H, d, J=7.9 Hz), 2.34 ppm (6H, s);
1
3
C NMR (50 MHz, CDCl ): d=139.47, 132.38, 129.20, 118.80, 81.53,
3
73.45, 21.58 ppm; GC-MS: m/z (%): 230 (100).
[
49]
1
1,4-Bis(o-methoxyphenyl)buta-1,3-diyne (2d):
H NMR (200 MHz,
CDCl
4H, m), 3.86 ppm (6H, s); C NMR (50 MHz, CDCl
34.39, 130.52, 120.46, 111.32, 110.74, 78.63, 77.99, 55.78 ppm; GC-MS:
m/z (%): 262 (100).
3
): d=7.44 (2H, d, J=7.5 Hz), 7.28 (2H, t, J=8.0 Hz), 6.92–6.81
1
3
(
1
3
): d=161.33,
[
49]
1
1
,4-Bis(m-methoxyphenyl)buta-1,3-diyne (2e):
H NMR (200 MHz,
CDCl
3
): d=7.18 (2H, t, J=8.0 Hz), 7.08 (2H, d, J=7.6 Hz), 7.00 (2H, s),
1
3
6
.88–6.81 (2H, d, J=8.1 Hz), 3.76 ppm (6H, s); C NMR (50 MHz,
Experimental Section
3
CDCl ): d=159.31, 129.48, 125.06, 122.66, 117.08, 115.99, 81.45, 73.62,
5
5.28 ppm; GC-MS: m/z (%): 262 (100).
,4-Bis(p-methoxyphenyl)buta-1,3-diyne (2 f):
): d=7.43 (4H, d, J=8.7 Hz), 6.82 (4H, d, J=8.7 Hz), 3.79 ppm
General information: All chemicals were purchased from Sigma–Aldrich
or Alfa Aesar and were used as received. Reactions were accomplished
in a Retsch MM 301 vibration ball mill by using milling beakers (35 mL)
[48]
1
1
H NMR (200 MHz,
CDCl
3
13
(
6H, s); C NMR (50 MHz, CDCl
3
): d=160.26, 134.00, 114.09, 113.92,
made from yttrium-stabilized zirconia (Y–ZrO
2
). Either 10ꢁ10 or 6ꢁ
8
1.19, 72.92, 55.28 ppm; GC-MS: m/z (%): 262 (100).
1
2 mm milling balls were applied that consisted of yttrium- and magne-
[
49]
1
1
,4-Bis(p-fluorophenyl)buta-1,3-diyne (2g):
): d=7.53–7.42 (4H, m), 7.00 ppm (4H, t, J=8.8 Hz); C NMR
50 MHz, CDCl ): d=165.56, 160.55, 134.58, 134.42, 117.83, 117.76,
16.11, 115.66, 80.40, 73.51 ppm; GC-MS: m/z (%): 238 (100).
H NMR (200 MHz,
sia-stabilized zirconia, respectively. GC-FID measurements were per-
formed on an HP-6890 GC machine and GC-MS measurements were re-
corded on an Agilent Technologies GC 6890N machine with MS detector
1
3
CDCl
(
1
3
3
5
1
2
973. Conditions for GC-FID: HP 5, 30 mꢁ0.32 mmꢁ0.25 mm;
0 psi; temperature program: 708C (hold for 3 min), 15 Kmin up to
808C (hold for 10 min); injector temperature: 2808C; detector tempera-
H
2
:
À1
[49] 1
Icosa-9,11-diyne (2h):
3
H NMR (200 MHz, CDCl ): d=2.19 (4H, t, J=
1
3
6.9 Hz), 1.58–1.18 (12H, m), 0.83 ppm (6H, t, J=6.4 Hz); C NMR
(50 MHz, CDCl ): d=77.48, 65.21, 31.76, 29.06, 28.99, 28.78, 28.29, 22.57,
19.15, 14.01 ppm; GC-MS: m/z (%): 274 (100).
ture: 3008C. Conditions for GC-MS: HP 5, 30 mꢁ0.32 mmꢁ0.25 mm; He:
3
À1
1
2
0 psi; temperature program: 708C (hold for 3 min), 15 Kmin up to
808C (hold for 7 min); injector temperature: 2808C; detector: electron
[
49] 1
Tetracosa-11,13-diyne (2i):
t, J=7.0 Hz), 1.56–1.15 (16H, m), 0.83 ppm (6H, t, J=6.5 Hz); C NMR
(50 MHz, CDCl ): d=77.47, 65.20, 31.83, 29.49, 29.41, 29.23, 29.03, 28.79,
28.30, 22.61, 19.15, 14.04 ppm; GC-MS: m/z (%): 330 (100).
3
H NMR (200 MHz, CDCl ): d=2.19 (4H,
1
3
impact (70 eV). NMR spectra were recorded with a Bruker Avance
2
2
00 MHz system at room temperature in chloroform-[ H]
General reaction procedure A: The milling beakers were filled with mill-
ing balls and then KF–Al (60 w% KF, 4 g), phenylacetylene (1a;
mmol, 204 mg), and CuI (5 mol%, 18 mg) were added sequentially.
3
(CDCl
3
).
3
[
50]
2
O
3
Synthesis of 1,2,4-triphenylbenzene (3): A mixture of 1,2,4-tribromo-
benzene (1.89 g, 6 mmol), Pd(OAc) (67.5 mg, 0.3 mmol), triphenylphos-
2
ACHTUNGTRENNUNG
2
Milling was performed at 30 Hz for 10 min. After cooling to room tem-
perature, the crude products were extracted on a frit with a thin silica
layer by using chloroform (3ꢁ10 mL). The solvent was evaporated in
vacuo; the crude products were dried and analyzed by GC-FID and GC-
MS.
phine (157.5 mg, 0.6 mmol), phenylboronic acid (3.30 g, 27 mmol), tolu-
À1
ene (90 mL), methanol (30 mL), and aqueous K CO (24 mL, 2 molL
2
3
)
was heated at 758C for 24 h while maintaining intensive stirring. After
the reaction mixture was cooled to room temperature, it was quenched
with water (30 mL) and extracted with dichloromethane (3ꢁ50 mL). The
General reaction procedure B: The milling beakers were filled with mill-
unified organic phase was dried with anhydrous Na
was removed in vacuo.
2 4
SO and the solvent
ing balls and then KF–Al
CuI (5 mol%, 18 mg) were added sequentially. Milling was performed at
0 Hz for 10 min. After cooling to room temperature, the milling beaker
2 3
O (60 w% KF, 4 g), the alkyne (2 mmol), and
3
was opened and allowed to stand in an air atmosphere for 30 min. Fol-
lowing this post-reaction oxidation, the crude products were extracted on
a frit with a thin silica layer by using chloroform (3ꢁ10 mL). The solvent
8136
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8129 – 8138

Solvent-Free Glaser Coupling
FULL PAPER
Acknowledgements
T. D. Clark, S. T. Marcum, J. Mack, Green Chem. 2010, 12, 209–211;
i) A. Baron, J. Martinez, F. Lamaty, Tetrahedron Lett. 2010, 51,
6246–6249.
This work was funded by the German Federal Environmental Foundation
DBU; grant No. 27281–31) and the Fond der Chemischen Industrie. R.S.
(
[16] a) T. Braun, Fullerene Sci. Technol. 1997, 5, 1291–1311; b) G.-W.
Wang in Encyclopedia of Nanoscience and Nanotechnology (Ed.:
H. S. Nalwa), American Scientific, Stevenson Ranch, 2004, pp. 557–
565; c) K. Komatsu, Top. Curr. Chem. 2005, 254, 185–206; d) H. Wa-
tanabe, E. Matsui, Y. Ishiyama, M. Senna, Tetrahedron Lett. 2007,
48, 8132–8137.
and A.S. would like to acknowledge the help of G. Gottschalt in the
screening of the catalysts. The authors are thankful to M. Friedrich (Insti-
tute of Inorganic and Analytical Chemistry, FSU Jena) for NMR analy-
ses. Special thanks go to Retsch GmbH for providing the equipment for
cryogenic milling under constant liquid nitrogen cooling (CryoMill).
[
17] a) S. Feldbæk Nielsen, D. Peters, O. Axelsson, Synth. Commun.
000, 30, 3501–3509; b) L. M. Klingensmith, N. E. Leadbeater, Tet-
2
rahedron Lett. 2003, 44, 765–768; c) F. Schneider, B. Ondruschka,
ChemSusChem 2008, 1, 622–625; d) F. Schneider, A. Stolle, B. On-
druschka, H. Hopf, Org. Process Res. Dev. 2009, 13, 44–48; e) F.
Schneider, T. Szuppa, A. Stolle, B. Ondruschka, H. Hopf, Green
Chem. 2009, 11, 1894–1899; f) J. Gꢆlvez, M. Gꢆlvez-Llompart, R.
Garcꢃa-Domemech, Green Chem. 2010, 12, 1056–1061.
[
[
1] C. Glaser, Ber. Dtsch. Chem. Ges. 1869, 2, 422–424.
2] a) P. Siemsen, B. Felber in Handbook of CÀH Transformations: Ap-
plications in Organic Synthesis (Ed.: G. Dyker), Wiley-VCH, Wein-
heim, 2005, pp. 53–62; b) F. Cataldo, Y. Keheyan in Polyynes: Syn-
thesis, Properties, and Applications (Ed.: F. Cataldo), CRC, Boca
Raton, 2006, pp. 493–498; c) W. E. Lindsell, C. Murray, P. N. Pres-
ton, T. A. J. Woodman, Tetrahedron 2000, 56, 1233–1245; d) T. R.
Hoye, P. R. Hanson, Tetrahedron Lett. 1993, 34, 5043–5046.
[
[
[
18] a) E. Tullberg, F. Schacher, D. Peters, T. Frejd, J. Organomet. Chem.
2
004, 689, 3778–3781; b) E. Tullberg, F. Schachter, D. Peters, T.
Frejd, Synthesis 2006, 1183–1189.
[
3] a) P. Siemsen, R. C. Livingston, F. Diederich, Angew. Chem. 2000,
19] a) D. A. Fulmer, W. C. Shearouse, S. T. Medonza, J. Mack, Green
Chem. 2009, 11, 1821–1825; b) R. Thorwirth, A. Stolle, B. On-
druschka, Green Chem. 2010, 12, 985–991.
1
12, 2740–2767; Angew. Chem. Int. Ed. 2000, 39, 2632–2657; b) M.
Brønsted Nielsen, F. Diederich, Chem. Rec. 2002, 2, 189–198; c) M.
Kivala, F. Diederich, Pure Appl. Chem. 2008, 80, 411–428; d) J. Liu,
J. W. Y. Lam, B. Z. Tang, Chem. Rev. 2009, 109, 5799–5867; e) B. W.
Gung, Compt. Rend. Chim. 2009, 12, 489–505; f) R. Gleiter, D. B.
Werz, Chem. Rev. 2010, 110, 4447–4488.
20] a) V. Declerck, P. Nun, J. Martinez, F. Lamaty, Angew. Chem. 2009,
1
21, 9482–9485; Angew. Chem. Int. Ed. 2009, 48, 9318–9321;
b) M. A. P. Martins, C. P. Frizzo, D. N. Moreira, L. Buriol, P. Macha-
do, Chem. Rev. 2009, 109, 4140–4182.
[
[
4] a) J. Li, H. Jiang, Chem. Commun. 1999, 2369–2370; b) H.-F. Jiang,
J.-Y. Tang, A.-Z. Wang, G.-H. Deng, S.-R. Yang, Synthesis 2006,
[
[
[
[
21] Y.-W. Dong, G.-W. Wang, L. Wang, Tetrahedron 2008, 64, 10148–
1
0154.
1
6
155–1161; c) A.-Z. Wang, H.-F. Jiang, Yougi Huaxue 2007, 27,
19–622 [A.-Z. Wang, H.-F. Jiang, Chem. Abstr. 2007, 148, 238761].
22] A. S. Heintz, J. E. Gonzalez, M. J. Fink, B. S. Mitchell, J. Mol. Catal.
A: Chem. 2009, 304, 117–120.
23] R. Thorwirth, A. Stolle, B. Ondruschka, A. Wild, U. S. Schubert,
Chem. Commun. 2011, 47, 4370–4372.
24] a) B. E. Blass, Tetrahedron 2002, 58, 9301–9320; b) B. Basu, P. Das,
S. Das, Curr. Org. Chem. 2008, 12, 141–158; c) F. Bernhardt, R.
Trotzki, T. Szuppa, A. Stolle, B. Ondruschka, Beilstein J. Org. Chem.
5] a) J. S. Yadav, B. V. S. Reddy, K. Bhaskar Reddy, K. Uma Gayathri,
A. R. Prasad, Tetrahedron Lett. 2003, 44, 6493–6496; b) B. C. Ranu,
S. Banerjee, Lett. Org. Chem. 2006, 3, 607–609; c) C. Ye, J.-C. Xiao,
B. Twamley, A. D. LaLonde, M. G. Norton, J. M. Shreeve, Eur. J.
Org. Chem. 2007, 5095–5100.
[
[
6] a) P.-H. Li, J.-C. Yan, M. Wang, L. Wang, Chin. J. Chem. 2004, 22,
2
010, 6, No. 7.
2
2
19–221; b) S.-N. Chen, W.-Y. Wu, F.-Y. Tsai, Green Chem. 2009, 11,
69–274.
[
[
25] J. Yamawaki, T. Ando, Chem. Lett. 1979, 8, 755–758.
26] It is assumed that the 32 and 40 w% commercial KF–Al O reagents
2 3
are prepared with the impregnation method, see reference [25].
27] a) G. W. Kabalka, R. M. Pagni, C. M. Hair, Org. Lett. 1999, 1, 1423–
7] X. Lu, Y. Zhang, C. Luo, Y. Wang, Synth. Commun. 2006, 36, 2503–
511.
2
[
[
[
8] F. Toda, Y. Tokumaru, Chem. Lett. 1990, 19, 987–990.
9] G. W. Kabalka, L. Wang, R. M. Pagni, Synlett 2001, 0108–0110.
1
425; b) B. Basu, P. Das, M. M. H. Bhuiyan, S. Jha, Tetrahedron
Lett. 2003, 44, 3817–3820.
[
10] a) A. Sharifi, M. Mirzaei, M. R. Naimi-Jamal, J. Chem. Res. 2002,
28–630; b) A. Sharifi, M. Mirzaei, M. R. Naimi-Jamal, Chem.
[
28] a) T. Ando, J. Yamawaki, T. Kawate, S. Sumi, Z. Hanafusa, Bull.
Chem. Soc. Jpn. 1982, 55, 2504–2507; b) J. H. Clark, D. G. Cork,
M. S. Robertson, Chem. Lett. 1983, 12, 1145–1148; c) M. Verziu, M.
Florena, S. Simon, V. Simon, P. Filip, V. I. Parvulescu, C. Hardacre,
J. Catal. 2009, 263, 56–66.
6
Monthly 2006, 137, 213–217.
[
[
[
11] M. Bandini, R. Luque, V. Budarin, D. J. Macquarrie, Tetrahedron
2
005, 61, 9860–9868.
12] D. Wang, J. Li, N. Li, T. Gao, S. Hou, B. Chen, Green Chem. 2010,
2, 45–48.
13] a) C. C. Koch, Nanostruct. Mater. 1993, 2, 109–129; b) E. Gaffet, F.
Bernard, J. C. Niepec, F. Charlot, C. Gras, G. Le Caꢂr, J. L. Gui-
chard, P. Delcroix, A. Mocellin, O. Tillement, J. Mater. Chem. 1999,
[
29] L. Fomina, B. Vazquez, E. Tkatchouk, S. Fomine, Tetrahedron 2002,
1
5
8, 6741–6747.
[30] N. Mizuno, K. Kamata, Y. Nakagawa, T. Oishi, K. Yamaguchi,
Catal. Today 2010, 157, 359–363.
9
, 305–314; c) C. Suryanarayana, Prog. Mater. Sci. 2001, 46, 1–184;
[31] The time between the end of the ball milling process and extraction
of the reaction mixture prior to analysis was approximately 3–5 min.
[32] X. Meng, C. Li, B. Han, T. Wang, B. Chen, Tetrahedron 2010, 66,
4029–4031.
d) L. Takacs, Prog. Mater. Sci. 2002, 47, 355–414.
[
14] a) G. Kaupp, Top. Curr. Chem. 2005, 254, 95–183; b) B. Rodrꢃguez,
A. Bruckmann, T. Rantanen, C. Bolm, Adv. Synth. Catal. 2007, 349,
2
213–2233; c) G. Kaupp, J. Phys. Org. Chem. 2008, 21, 630–643;
[33] a) E. Negishi, A. Alimardanov in Handbook of Organopalladium
Chemistry for Organic Synthesis, Vol. 1 (Ed.: E. Negishi), Wiley-
VCH, New York, 2002, p. 989; b) A. S. Batsanov, J. C. Collings,
I. J. S. Fairlamb, J. P. Holland, J. A. K. Howard, Z. Lin, T. B. Marder,
A. C. Parsons, R. M. Ward, J. Zhu, J. Org. Chem. 2005, 70, 703–706;
c) J.-H. Li, Y. Liang, X.-D. Zhang, Tetrahedron 2005, 61, 1903–1907.
[34] Z. H. Li, M. S. Wong, Y. Tao, Tetrahedron 2005, 61, 5277–5285.
[35] a) Y. Yamamoto, A. Nagata, H. Nagata, Y. Ando, Y. Arikawa, K.
Tatszmi, K. Itoh, Chem. Eur. J. 2003, 9, 2469–2483; b) N. Agenet, O.
Buisine, F. Slowinski, V. Gandon, C. Aubert, M. Malacria, Org.
React. 2007, 68, 1–302.
d) T. Fri sˇ cˇ i cˇ , W. Jones, Cryst. Growth Des. 2009, 9, 1621–1637; e) A.
Stolle, T. Szuppa, S. E. S. Leonhardt, B. Ondruschka, Chem. Soc.
Rev. 2011, 40, 2317–2329.
[
15] a) Z. Zhang, Y.-W. Dong, G.-W. Wang, K. Komatsu, Synlett 2004,
6
1–64; b) Z. Zhang, Y.-W. Dong, G.-W. Wang, K. Komatsu, Chem.
Lett. 2004, 33, 168–169; c) B. Rodrꢃguez, A. Bruckmann, C. Bolm,
Chem. Eur. J. 2007, 13, 4710–4722; d) E. M. C. Gꢄrad, H. Sahin, A.
Encinas, S. Brꢅse, Synlett 2008, 2702–2704; e) R. Trotzki, M. M.
Hoffmann, B. Ondruschka, Green Chem. 2008, 10, 767–772; f) E.
Colacino, P. Nun, F. M. Colacino, J. Martinez, F. Lamaty, Tetrahe-
dron 2008, 64, 5569–5576; g) A. Bruckmann, B. Rodriguez, C.
Bolm, CrystEngComm 2009, 11, 404–407; h) D. C. Waddell, I. Thiel,
[36] a) P. H. Li, L. Wang, M. Wang, J. C. Yan, Chin. Chem. Lett. 2004, 15,
1295–1298; b) W. Yin, C. He, M. Chen, H. Zhang, A. Lei, Org. Lett.
Chem. Eur. J. 2011, 17, 8129 – 8138
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8137

A. Stolle et al.
2
009, 11, 709–712; c) J. D. Crowley, S. M. Goldup, N. D. Gowans,
D. A. Leigh, V. E. Ronaldson, A. M. Z. Slawin, J. Am. Chem. Soc.
010, 132, 6243–6248.
[42] C.-L. Deng, Y.-X. Xie, D.-L. Yin, J.-H. Li, Synthesis 2006, 3370–
3376.
2
[43] R. Luque, D. J. Macquarrie, Org. Biomol. Chem. 2009, 7, 1627–
1632.
[44] P. Balꢆ zˇ , Mechanochemistry in Nanoscience and Minerals Engineer-
ing, Springer, Berlin, 2008, Chapter 2, pp. 103–132.
[
[
37] a) M. E. Krafft, C. Hirosawa, N. Dalal, C. Ramsey, A. Stiegman, Tet-
rahedron Lett. 2001, 42, 7733–7736; b) M. Mayer, W. M. Czaplik,
A. J. von Wangelin, Synlett 2009, 2919–2923; c) G. Hilt, C. Hengst,
M. Arndt, Synthesis 2009, 395–398.
38] The application of the solid alkyne p-nitrophenylacetylene (melting
point=148–1508C) failed and yielded the corresponding product in
lower than 1% yield.
39] D. C. Waddell, J. Mack, Green Chem. 2009, 11, 79–82.
40] a) J. Mokhtari, M. R. Naimi-Jamal, H. Hamzeali, M. G. Dekamin, G.
Kaupp, ChemSusChem 2009, 2, 248–254; b) M. R. Naimi-Jamal, J.
Mokhtari, M. G. Dekamin, G. Kaupp, Eur. J. Org. Chem. 2009,
[
[
[
45] R. Thorwirth, F. Bernhardt, A. Stolle, B. Ondruschka, J. Asghari,
Chem. Eur. J. 2010, 16, 13236–13242.
46] For the detailed mechanism and schematic pictures of ball mills, see
references [13c], [14b], and [44].
47] T. Szuppa, A. Stolle, B. Ondruschka, W. Hopfe, Green Chem. 2010,
[
[
1
2, 1288–1294.
[
[
48] J.-H. Li, Y. Liang, Y.-X. Xie, J. Org. Chem. 2005, 70, 4393–4396.
49] T. Oishi, T. Katayama, K. Yamaguchi, N. Mizuno, Chem. Eur. J.
2
009, 15, 7539–7542.
3
567–3572.
41] a) W. B. Austin, N. Bilow, W. J. Kellegan, K. S. Y. Lau, J. Org. Chem.
981, 46, 2280–2286; b) A. Arcadi, S. Cacchi, M. D. Rasario, G. Fab-
rizi, F. Marinelli, J. Org. Chem. 1996, 61, 9280–9288.
[
50] X. Bu, Z. Zhang, X. Zhou, Organometallics 2010, 29, 3530–3534.
[
1
Received: February 24, 2011
Published online: May 30, 2011
8138
www.chemeurj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 8129 – 8138