Solvent-dependent regiochemical cyclotrimerisation of phenylacetylene with cobalt catalysts containing disulfide ligands: A case study

Article abstract of DOI:10.1002/ejoc.200800106

The intermolecular cobalt-catalysed cyclotrimerisation of phenylacetylene as a benchmark alkyne with [1,2-bis(4-methoxyphenylthio)ethane]cobalt dibromide gave the symmetrical 1,3,5-triphenylbenzene as the major product when dichloromethane was used as the solvent, whereas the unsymmetrical 1,2,4-trisubstituted product was obtained in acetonitrile in quantitative yield. Optimisation of the aromatic and aliphatic disulfide ligands in different solvents is discussed. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejoc.200800106

FULL PAPER
DOI: 10.1002/ejoc.200800106
Solvent-Dependent Regiochemical Cyclotrimerisation of Phenylacetylene with
Cobalt Catalysts Containing Disulfide Ligands: A Case Study
Gerhard Hilt,*[a] Christoph Hengst,[a] and Wilfried Hess[a]
Keywords: Cobalt / Disulfide ligands / Cyclotrimerisation / Regioselectivity / Heterogeneous catalysis
The intermolecular cobalt-catalysed cyclotrimerisation of
phenylacetylene as a benchmark alkyne with [1,2-bis(4-
methoxyphenylthio)ethane]cobalt dibromide gave the sym-
metrical 1,3,5-triphenylbenzene as the major product when
tonitrile in quantitative yield. Optimisation of the aromatic
and aliphatic disulfide ligands in different solvents is dis-
cussed.
dichloromethane was used as the solvent, whereas the un- (© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
symmetrical 1,2,4-trisubstituted product was obtained in ace-
Germany, 2008)
Introduction
fying simple cobalt complexes that are capable of forming
regioselectively symmetrical 1,3,5-trisubstituted benzene de-
rivatives from terminal alkynes,[7] especially in the light of
earlier findings that 1,2,4-trisubstituted benzene derivatives
are predominantly formed when diimine-type ligands such
as 4 are used. The potential of 1,2-bis(phenylthio)ethane
(3a) as a ligand of cobalt was identified previously when the
regioselective formation of the two possible trimerisation
products was investigated in different solvents with different
ligand types.[3]
The coordination of unsaturated ligands such as alkenes
and alkynes to transition-metal complexes with vacant co-
ordination sites is a prerequisite to induce cyclotrimeris-
ation or cycloaddition processes within the ligand sphere.
Accordingly, organic transformations that lead to carbon–
carbon bond formations, which would not be possible un-
der regular thermal conditions, can now be realised in the
presence of transition metal complexes.[1] Several new ap-
proaches to the cyclotrimerisation of alkynes besides the
traditional cyclopentadienylcobalt or carbonylcobalt com-
plexes[2] have been reported recently. Among these are the
rather simple (diimine)cobalt or (iminopyridine)cobalt
complexes,[3] and even ligand-free cobalt salts have been
used.[4] Over the last couple of years we have investigated
Results and Discussion
The use of diimine-type ligands in the cyclotrimerisation
cobalt(I)-catalysed Diels–Alder reactions between non-acti- of alkynes leads predominantly to the unsymmetrical 1,2,4-
vated terminal and internal alkynes and acyclic 1,3-dienes.[5] trisubstituted benzene derivatives 2, as was shown in a pre-
Recently, we reported a unique catalyst system that is able vious report.[3] Therein we also investigated the 1,2-bis-
to promote the formation of regioisomeric meta-disubsti- (phenylthio)ethane ligand 3a, which exhibits a very unusual
tuted dihydroaromatic compounds in excellent yields and solvent dependency. The data in Table 2 show that in aceto-
with very high regioselectivities for the first time.[6] Under nitrile product 2 is predominantly formed. On the other
these circumstances we were also very interested in identi- hand, when the reaction was performed in dichlorometh-
Scheme 1. Benchmark ligands for the regioselective cyclotrimerisation of phenylacetylene.
ane, the regioselectivity was reversed with the symmetrical
[a] Fachbereich Chemie, Philipps-Universität Marburg,
benzene derivative 1 being generated in a slight excess. We
were intrigued to find a sulfur-based ligand system, which
would predominantly generate the symmetrical products, as
Hans-Meerwein-Str., 35043 Marburg, Germany
Fax: +49-6421-2825677
E-mail: hilt@chemie.uni-marburg.de
Eur. J. Org. Chem. 2008, 2293–2297
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2293

G. Hilt, C. Hengst, W. Hess
FULL PAPER
this would be a major step towards a simple and inexpen- Table 2. Results of the cobalt-catalysed cyclotrimerisation reaction
using aromatic disulfide ligands.
sive catalyst system for the selective synthesis of either tri-
mer 1 or 2 depending on the ligand and the solvent used.
After we realised that the [1,2-bis(4-methoxyphenylthio)-
ethane]cobalt complex 3b (Table 2) exhibited a significantly
higher regioselectivity with respect to the generation of 1,
we decided to revisit the solvent dependency of [1,2-bis(4-
methoxyphenylthio)ethane]cobalt dibromide to see if
changes in the solvent should be made. Under these circum-
stances mainly chlorinated solvents were investigated to de-
termine the best combination of reactivity and selectivity
in the cyclotrimerisation reaction of phenylacetylene with
5 mol-% of the cobalt catalyst and a reaction time of 30 min
(Scheme 1).[8] The results are summarised in Table 1.
Table 1. Results for the cobalt-catalysed cyclotrimerisation reaction
with ligand 3b in different solvents.
Entry
Solvent
Yield [%] (ratio 1/2)
1
2
3
4
5
CH2Cl2
C6H5Cl
ClCH2CH2Cl
CHCl3
99 (6.2:1.0)
68 (1.5:1.0)
57 (2.8:1.0)
78 (1.4:1.0)
0 (–)
CCl4
6
7
8
9
10
11
HSCH2CH2SH
tetrahydrothiophene
C6H6
0 (–)
99 (1.0:2.8)
35 (2.8:1.0)
99 (1.6:1.0)
99 (1.0:22.9)
99 (1.0:1.7)
C6H5CH3
CH3CN
THF
The cyclotrimerisation is best conducted in dichloro-
methane under an inert gas. Other chlorinated solvents gave
inferior results, both with respect to reactivity and selectiv-
ity. In tetrachloromethane no reaction was observed at all.
Two sulfur-containing solvents were tested, which might
also act as additional ligands. The results showed that with
ethane-1,2-thiol (Entry 6) the reaction was inhibited,
whereas with tetrahydrothiophene (Entry 7), a sulfide, the
catalyst system exhibited good reactivity, but the selectivity
was only moderately in favour of product 2. The results in
acetonitrile were as expected as this solvent had given excel-
lent results with cobalt complexes of the diimine ligand
series in terms of reactivity and selectivity for the formation
of 2.[3] With benzene and toluene, only toluene gave a good
reactivity and encouraging selectivities, albeit still inferior
to that of dichloromethane. After encouraging preliminary
results with tetrahydrofuran, we decided to continue to
screen for ligands in the solvents dichloromethane, acetoni-
trile and tetrahydrofuran. In these three solvents different
aromatic sulfide ligands were tested in preformed catalysts
(see Exp. Sect.) for the cyclotrimerisation of phenylacetyl-
ene. The results of these reactions are summarised in
Table 2 and Scheme 2.
The introduction of electron-donating methoxy groups
onto the phenyl substituents (3b) significantly increased the
regioselectivity of the cyclotrimerisation reaction in dichlo- ent ligand system 3a (Entries 1 and 2). If an electron-with-
romethane in favour of product 1 and in acetonitrile in fav- drawing group was used such as in ligand 3r the desired
our of product 2 (Entries 4 and 5) compared with the par- symmetrical cyclotrimerisation product 1 was obtained as
2294
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2293–2297

Cyclotrimerisation of Phenylacetylene with Cobalt Catalysts Containing Disulfide Ligands
Table 3. Results of the cobalt-catalysed cyclotrimerisation reaction
with aliphatic disulfide ligands in dichloromethane.
Scheme 2. Optimisation of disulfide ligands in the cobalt-catalysed
cyclotrimerisation in different solvents.
the minor product in dichloromethane. Based on these re-
sults, further investigations with electron-withdrawing sub-
stituents were disregarded. On the other hand, the introduc-
tion of further methoxy groups was envisaged to further
increase the regioselectivity in favour of product 1 in dichlo-
romethane. Surprisingly, additional improvement was not
observed. On the contrary, the regioselectivity in favour of
1 was lowered with ligand 3c. Other ligands (3d–f) with the
aliphatic ethylene backbone motif derived from dibromo-
ethane and commercially available thiol derivatives were
synthesised and tested, but no significant increase in selec-
tivity towards 1 was observed in dichloromethane (En-
tries 10, 13 and 16). Nevertheless, most of the disulfide li-
gands tested demonstrated rather high reactivities and good
[a] For the alkyl-substituted disulfide ligands the reaction time was
15 h.
to excellent selectivities in acetonitrile for the formation of which gave only incomplete conversion after 15 h. As for
2. The results in tetrahydrofuran are reported for the sake the selectivity, the two ligands 3s and 3u gave essentially
of completeness as no advantage was determined in tetra- identical results with moderate preference for product 1,
hydrofuran over the other two solvents for the ligands whereas 3t gave a 1:1 mixture of the two regioisomers 1 and
tested.
2, and the ligands 3v–x preferred the formation of 2 as the
Changes to the carbon backbone on the other hand led major isomer. In a final experiment the cyclic dithioketal
to interesting results. Whereas the two-carbon backbone of 3y was tested as an additive/ligand and, although the con-
the ethylene-bearing ligands led to the preferential forma- version went to completion, product 2 was again the major
tion of 1 (Entries 1, 4, 7 and 10), the corresponding propyl- isomer formed in this reaction.
ene ligand 3g (Entry 19) and methylene-bridged bidentate
The solvent dependency can be rationalised when the re-
disulfide ligand 3h (Entry 22) favoured the unsymmetrical sults of the (diimine)cobalt-catalysed cyclotrimerisation re-
product 2. This behaviour was also observed for the aliphat- actions[3] and the results presented here are viewed in con-
ically substituted ligands discussed below.
text. The results from the (diimine)cobalt-catalysed reac-
The presence of an aromatic backbone in the form of a tions in different nitrile solvents suggest that a coordination
1,2-disubstituted benzene ring or a 2,2Ј-disubstituted 1,1Ј- site is occupied by a nitrile ligand and that the steric hin-
biphenyl derivative in ligands 3i–m (Entries 25–39) also drance of different nitrile solvents (MeCN, EtCN and
gave good selectivities for the formation of 1. However, the tBuCN) influences the regioselectivity of the reactions. This
introduction of further methoxy groups into the ligands 3k seems to also be the case in the reactions with aromatically
and 3l (Entries 31 and 34) did not result in increased selec- substituted disulfide ligands in acetonitrile with product ra-
tivity for the formation of 1. Finally, the sterically bulky tios ranging from 1:15 to 1:22 as these are comparable to
aliphatically substituted aryl substituents introduced in li- the bulky diimine ligands used in earlier studies. The very
gands 3o and 3p and the sulfoxide ligand 3n gave inferior dramatic change in selectivity observed in dichloromethane
results. The introduction of a 2-naphthyl substituent in li- can be interpreted as resulting from the different geometries
gand 3q led to lower regioselectivity relative to ligand 3a, of the active cobalt complexes. The formation of the sym-
thus further modifications of the naphthyl ring with elec- metrically substituted product 1 is only possible from a
tron-donating groups were not considered. Nevertheless, complex with a 2,4-diphenyl-substituted cobaltacycle (C),
many of the disulfide ligands gave quantitative yields of the whereas the other possible intermediates (A and B) will lead
trimerisation products, but the selectivity with respect to the to the formation of the unsymmetrical product 2 (Fig-
formation of 1,3,5-triphenylbenzene (1) was only moderate ure 1).[9] We assume that the cobaltacycle intermediates
to good for ligand 3b. contain one bidentate disulfide ligand and that the free co-
The reactions with alkyl-substituted disulfides 3s–y were ordination site is occupied by another phenylacetylene
also tested. The results are summarised in Table 3.
molecule.
The results clearly show that the ethylene-bridged ligands
By extrapolating from the collected data, it can be seen
3s and 3u yield the cyclotrimerisation products in quantita- that not only the steric effects of the phenyl substituents on
tive yield and are superior to the propylene-bridged ligands the disulfide ligand favour the formation of intermediate
3t, 3v and 3x as well as the tert-butyl-substituted ligand 3w, cobalt complex C, but that the increased electron density
Eur. J. Org. Chem. 2008, 2293–2297
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2295

G. Hilt, C. Hengst, W. Hess
FULL PAPER
Experimental Section
General Remarks: NMR spectra were recorded with a Bruker Av-
ance 300 or DRX 500 (1H: 300 or 500 MHz; 13C: 75 or 125 MHz)
spectrometer using TMS as the internal standard (δ = 0 ppm) un-
less otherwise noted. Mass and GC-MS spectra were measured
with a Hewlett-Packard 6890 GC-System equipped with a Hewlett-
Packard 5973 Mass Selective Detector. For (high-resolution) mass
spectra Finnigan MAT 95S and LTQ (ESI, HRMS) spectrometers
were used. Analytical thin-layer chromatography was performed on
Merck silica gel 60 F254. For column chromatography Merck silica
gel 60 (230–400 mesh ASTM) was used. All reactions were carried
out under nitrogen or argon using standard Schlenk techniques.
Dichloromethane was dried with phosphorus pentoxide, acetoni-
trile with CaH2 and tetrahydrofuran with sodium/benzophenone.
The analytical data of the cyclotrimerisation products 1 and 2 are
identical to those in the literature.[2,3,7] The ligands 3a, 3b, 3d–k, 3n
and 3s–y were synthesised by adapting known procedures.[10] The
cobalt complexes were prepared in THF, and after evaporation of
the solvent the residue was used without further purification.
Figure 1. Cobaltacycle intermediates and cyclotrimerisation prod-
ucts derived therefrom.
resulting from the introduction of the methoxy substituent
is also a determining factor. Computational investigations
into the causal relationship of ligand and solvent incorpora-
tion will be undertaken in due course to further elucidate
the drastic solvent effect on the cobalt/disulfide catalyst sys-
tems.
Further attempts to increase the selectivity by changing
the counterion of the cobalt salt, the reaction temperature
or the nature of the reducing agent did not result in in-
creased reactivity or selectivity.
The application of methyl propiolate in the cyclotrimeris-
ation reaction with the precatalyst [Co(3b)Br2] in dichloro-
methane led to the corresponding symmetrical triester as
the major product in 99% yield (1,3,5-derivative/1,2,4-de-
rivative = 2.6:1.0). 2-Ethynylthiophene was also converted
in quantitative yield into the trimer under identical condi-
tions. However, the cyclotrimerisation product was ob-
tained as a 1.1:1.0 mixture, showing no preference for the
symmetrical benzene derivative. Lastly, trimethylsilylacetyl-
ene was trimerised with the precatalyst [Co(3b)Br2] in
dichloromethane and gave the trimer in quantitative yield
but showed a preference for the unsymmetrical product
(1.0:1.4).
General Procedure for the Cobalt-Catalysed Cyclotrimerisation Re-
action: Phenylacetylene (0.11 mL, 1.0 mmol) was added to a sus-
pension of the cobalt complex (0.05 mmol, 5.0 mol-%), zinc dust
(7 mg, 0.1 mmol, 10.0 mol-%) and anhydrous ZnI2 (32 mg,
0.1 mmol, 10.0 mol-%) in anhydrous solvent (CH2Cl2, CH3CN or
THF) (1.0 mL) under argon, and the mixture was stirred at room
temperature. The reaction was monitored by GC-MS. After com-
plete conversion (or after 30 min) the grey reaction mixture was
passed through a pad of silica using pentane/CH2Cl2 (10:1) as the
eluent. The solvents were removed and the crude product purified
by flash chromatography using pentane/CH2Cl2 (100:1) as eluent
(for the yields, see Tables 1, 2 and 3). The ratios of regioisomers 1
1
and 2 formed were verified by integration of H NMR signals.
Analytical Data for the Ligands 3c, 3l, 3m, 3o and 3p
1,2-Bis(3,4-dimethoxyphenylthio)ethane (3c): 1H NMR (300 MHz,
CDCl3): δ = 6.94–6.88 (m, 4 H), 6.78 (d, J = 8.2 Hz, 2 H), 3.87 (s,
6 H), 3.83 (s, 6 H), 2.96 (s, 4 H) ppm. 13C NMR (75 MHz, CDCl3):
δ = 149.3, 148.9, 124.9, 123.9, 115.5, 111.7, 56.1, 56.0, 35.4 ppm.
MS (EI): m/z (%) = 366 (8) [M]+, 338 (33), 320 (6), 181 (7), 169
(100), 154 (8), 139 (6), 125 (8). HRMS: calcd. for C18H22O4S2
366.0960; found 366.0963. IR (KBr): ν = 1580, 1503, 1465, 1443,
˜
1396, 1317, 1254, 1229, 1177, 1136, 1022, 796 cm–1.
(4,5-Dimethoxy-1,2-phenylene)dithiobis(4-methoxybenzene) (3l): 1H
NMR (300 MHz, CDCl3): δ = 7.32 (d, J = 8.9 Hz, 4 H), 6.87 (d,
J = 8.9 Hz, 4 H), 6.66 (s, 2 H), 3.80 (s, 6 H), 3.68 (s, 6 H) ppm.
13C NMR (75 MHz, CDCl3): δ = 159.4, 148.8, 133.6, 129.7, 126.1,
115.0, 114.8, 56.1, 55.5 ppm. MS (EI): m/z (%) = 414 (24) [M]+,
386 (17), 276 (4), 244 (4), 207 (65), 179 (100), 164 (12), 149 (9), 115
(7), 91 (8), 57 (14). HRMS: calcd. for C22H22O4S2 414.0960; found
Conclusions
The use of disulfide ligands in the cyclotrimerisation of
phenylacetylene led to very interesting results in terms of
414.0968. IR (KBr): ν = 1590, 1491, 1435, 1288, 1248, 1204, 1169,
˜
their solvent dependency and the regioselective formation 1027, 833 cm–1.
of the symmetrical product 1, which is best achieved with
ligand 3b in dichloromethane, and unsymmetrical product
2, obtained when the reaction was performed in acetonitrile
with the same ligand 3b as the ligand of choice. In terms of
2,2Ј-Bis(4-methoxyphenylthio)-1,1Ј-biphenyl (3m): 1H NMR
(300 MHz, CDCl3): δ = 7.42–7.36 (m, 4 H), 7.23–7.19 (m, 6 H),
7.03–6.98 (m, 2 H), 6.89–6.83 (m, 4 H), 3.81 (s, 6 H) ppm. 13C
NMR (75 MHz, CDCl3): δ = 160.0, 139.4, 138.9, 136.0, 130.4,
selectivity and reactivity, the disulfide ligands are at least 128.5, 128.5, 125.5, 124.8, 115.1, 55.5 ppm. MS (EI): m/z (%) = 430
(2) [M]+, 291 (100), 276 (18), 258 (10), 247 (7), 184 (20), 139 (5).
equal to the diimine ligands investigated earlier. The reac-
HRMS: calcd. for C26H22O2S2 430.1061; found 430.1067. IR
tions with the disulfide ligands are normally complete
(KBr): ν = 1588, 1491, 1453, 1287, 1247, 1029, 825, 746, 730 cm–1.
˜
within a few minutes of applying 5 mol-% of the catalyst
system consisting of [Co(3b)Br2], zinc powder and zinc io-
dide on a 1 mmol scale.
1,2-Bis(4-tert-butyl-2-methylphenylthio)ethane (3o): 1H NMR
(300 MHz, CDCl3): δ = 7.20 (s, 2 H), 7.13 (s, 4 H), 3.01 (s, 4 H),
2296
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2293–2297

Cyclotrimerisation of Phenylacetylene with Cobalt Catalysts Containing Disulfide Ligands
2.37 (s, 6 H), 1.29 (s, 18 H) ppm. 13C NMR (75 MHz, CDCl3): δ
= 149.8, 138.3, 130.8, 129.7, 127.5, 123.5, 34.3, 32.9, 31.3,
20.8 ppm. MS (EI): m/z (%) = 386 (23) [M]+, 358 (15), 343 (9),
207 (71), 179 (100), 165 (20), 131 (7). HRMS: calcd. for C24H34S2
Chem. 2005, 4741–4767; Y. Yamamoto, Curr. Org. Chem. 2005,
9, 503–519.
[3] G. Hilt, W. Hess, T. Vogler, C. Hengst, J. Organomet. Chem.
2005, 690, 5170–5181; G. Hilt, T. Vogler, W. Hess, F. Galbiati,
Chem. Commun. 2005, 1474–1475; N. Saino, F. Amemiya, E.
Tanabe, K. Kase, S. Okamoto, Org. Lett. 2006, 8, 1439–1442;
A. Goswami, T. Ito, S. Okamoto, Adv. Synth. Catal. 2007, 349,
2368–2374.
[4] H. Bönnemann, R. Brinkmann, H. Schenkluhn, Synthesis
1974, 575–577; L. Doszczak, R. Tacke, Organometallics 2007,
26, 5722–5723; L. Doszczak, P. Fey, R. Tacke, Synlett 2007,
753–756.
[5] For leading references, see: G. Hilt, S. Lüers, K. Harms, J. Org.
Chem. 2004, 69, 624–630; G. Hilt, K. I. Smolko, Angew. Chem.
2003, 115, 2901–2903; Angew. Chem. Int. Ed. 2003, 42, 2795–
2797; G. Hilt, K. I. Smolko, B. V. Lotsch, Synlett 2002, 1081–
1084; G. Hilt, F.-X. du Mesnil, Tetrahedron Lett. 2000, 41,
6757–6761.
386.2102; found 386.2091. IR (KBr): ν = 2962, 1480, 1391, 1362,
˜
1263, 1196, 1120, 1055, 880, 834 cm–1.
1,2-Bis(5-tert-butyl-2-methylphenylthio)ethane (3p): 1H NMR
(300 MHz, CDCl3): δ = 7.35 (s, 2 H), 7.20–7.10 (m, 4 H), 3.07 (s,
4 H), 2.36 (s, 6 H), 1.28 (s, 18 H) ppm. 13C NMR (75 MHz,
CDCl3): δ = 149.6, 136.1, 133.7, 130.2, 127.6, 124.1, 34.6, 33.4,
31.5, 20.1 ppm. MS (EI): m/z (%) = 386 (23) [M]+, 207 (100), 179
(75), 149 (8), 131 (6). HRMS: calcd. for C24H34S2 386.2102; found
386.2104. IR (KBr): ν = 2962, 2868, 1487, 1464, 1385, 1361, 1261,
˜
1200, 1120, 1062, 872, 813, 716 cm–1.
[6] G. Hilt, J. Janikowski, W. Hess, Angew. Chem. 2006, 118, 5328–
Acknowledgments
5331; Angew. Chem. Int. Ed. 2006, 45, 5204–5206.
[7] For recent non-cobalt-catalysed cyclotrimerisation of phenyl-
acetylene, see: F. Goettmann, A. Fischer, M. Antonietti, A.
Thomas, New J. Chem. 2007, 31, 1455–1460; Z. Zhu, J. Wang,
Z. Zhang, X. Xiang, X. Zhou, Organometallics 2007, 26, 2499–
2500; V. Cadierno, S. E. Garcia-Garrido, J. Gimeno, J. Am.
Chem. Soc. 2006, 128, 15094–15095; M. Kakeya, T. Fujihara,
T. Kasaya, A. Nagasawa, Organometallics 2006, 25, 4131–4137;
Z. Zhu, C. Wang, X. Xiang, C. Pi, X. Zhou, Chem. Commun.
2006, 2066–2068; K. Tanaka, K. Toyoda, A. Wada, K. Shira-
saka, M. Hirano, Chem. Eur. J. 2005, 11, 1145–1156; V. Conte,
E. Elakkari, B. Floris, V. Mirruzzo, P. Tagliatesta, Chem. Com-
mun. 2005, 1587–1588; N. K. Yokomakura, T. Hattori, S. Mi-
yano, Tetrahedron Lett. 2006, 47, 1157–1161; M. Ohkubo, S.
Mochizuki, T. Sano, Y. Kawaguchi, S. Okamoto, Org. Lett.
2007, 9, 773–776; K. Yoshida, I. Morimoto, K. Mitsudo, H.
Tanaka, Chem. Lett. 2007, 36, 998–999.
We thank the German Science Foundation for financial support.
[1] For reviews, see: M. Malacria, C. Aubert, J. L. Renaud in Sci-
ence of Synthesis: Houben-Weyl Methods of Molecular Trans-
formations (Eds.: M. Lautens, B. M. Trost), Thieme, Stuttgart,
2001, vol. 1, pp. 439–530; S. Saito, Y. Yamamoto, Chem. Rev.
2000, 100, 2901–2916; I. Ojima, M. Tzamarioudaki, Z. Li, R. J.
Donovan, Chem. Rev. 1996, 96, 635–662; M. Lautens, W.
Klute, W. Tam, Chem. Rev. 1996, 96, 49–92; D. B. Grotjahn in
Comprehensive Organometallic Chemistry II (Eds.: E. W. Abel,
F. G. A. Stone, G. Wilkinson, L. Hegedus), Pergamon Press,
Oxford, 1995, vol. 12, pp. 741–770; N. E. Schore in Comprehen-
sive Organic Synthesis (Eds.: B. M. Trost, I. Fleming, L. A. Pa-
quette), Pergamon Press, Oxford, 1991, vol. 5, pp. 1129–1162;
K. P. C. Vollhardt, Angew. Chem. 1984, 96, 525–541; Angew. [8] For many disulfide ligands that show complete conversion and
Chem. Int. Ed. Engl. 1984, 23, 539–556.
quantitative yields, the reaction time can be shortened to sev-
eral minutes.
[2] For examples of the cobalt-catalysed cyclotrimerisation of
phenylacetylene, see: B.-H. Xu, D.-H. Wu, Y.-Z. Li, H. Yan,
Organometallics 2007, 26, 4344–4349; L. D. Field, A. J. Ward,
J. Organomet. Chem. 2003, 681, 91–97; L. Yong, H. But-
enschön, Chem. Commun. 2002, 2852–2853; T. Sugihara, A.
Wakabayashi, Y. Nagai, H. Takao, H. Imagawa, M. Nishizawa,
Chem. Commun. 2002, 576–577; F. Montilla, T. Aviles, T. Casi-
miro, A. A. Ricardo, M. Nunes da Ponte, J. Organomet. Chem.
2001, 632, 113–118; L. D. Field, A. J. Ward, P. Turner, Aust. J.
Chem. 1999, 52, 1085–1092; M. S. Sigman, A. W. Fatland, B. E.
Eaton, J. Am. Chem. Soc. 1998, 120, 5130–5131; H.-Y. Rhyoo,
B. Y. Lee, H. K. B. Yu, Y. K. Chung, J. Mol. Catal. 1994, 92,
41–49; S. Kotha, E. Brahmachary, K. Lahiri, Eur. J. Org.
[9] For mechanistic considerations, see: N. Agenet, V. Gandon,
K. P. C. Vollhardt, M. Malacria, C. Aubert, J. Am. Chem. Soc.
2007, 129, 8860–8871; G. Dazinger, M. Torres-Rodrigues, K.
Kirchner, M. J. Calhorda, P. J. Costa, J. Organomet. Chem.
2006, 691, 4434–4445; R. Diercks, B. E. Eaton, S. Gürtzgen, S.
Jalisatgi, A. J. Matzger, R. H. Radde, K. P. C. Vollhardt, J. Am.
Chem. Soc. 1998, 120, 8247–8248.
[10] E. D. Brown, S. M. Iqbal, L. N. Owen, J. Chem. Soc. C 1966,
415–419; R. Adams, W. Reifschneider, A. Ferretti, Org. Synth.
1962, 42, 22–25.
Received: January 29, 2008
Published Online: March 19, 2008
Eur. J. Org. Chem. 2008, 2293–2297
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2297