Oxidative homocoupling of aryl, alkenyl, and alkynyl Grignard reagents with TEMPO and dioxygen

Article abstract of DOI:10.1002/anie.200804197

(Formula Presented) Without a transition metal, unsaturated organomagnesium compounds undergo homocoupling in the presence of a substoichiometric amount of the oxidant 2,2,6,6-tertramethylpiperidine-N-oxyl radical (TEMPO) and oxygen. Alkynyl Grignard reagents can even be coupled to give the corresponding diynes with dioxygen in the absence of a catalyst (see scheme).

Full text of DOI:10.1002/anie.200804197

Angewandte
Chemie
DOI: 10.1002/anie.200804197
Synthetic Methods
Oxidative Homocoupling of Aryl, Alkenyl, and Alkynyl Grignard
Reagents with TEMPO and Dioxygen**
Modhu Sudan Maji, Thorben Pfeifer, and Armido Studer*
The treatment of various alkyl organometallic compounds
Knochel, Mayr, and co-workers showed recently that the
homocoupling of Grignard reagents proceeds efficiently in
the presence of a stoichiometric amount of the organic
oxidant 3,3’,5,5’-tetra-tert-butyl-4,4’-diphenoquinone.[5]
For initial studies, we chose the conversion of PhMgBr
into biphenyl (1a) with TEMPO (1.08 equiv) in THF
(Table 1). When the reaction was carried out at room
À
R M (M = Li, Mg, Zn, Cu, Sm, Ti) with 2 equivalents of
2,2,6,6-tertramethylpiperidine-N-oxyl radical (TEMPO) has
been reported to lead to the corresponding alkoxyamines
TEMPO–R.[1] In these processes, one equivalent of TEMPO
is required for oxidation of the organometallic species to the
corresponding C-centered radical, which is subsequently
trapped by the second equivalent of TEMPO to provide an
alkoxyamine (Scheme 1).[2] Alkyl catecholboranes have been
Table 1: Oxidative homocoupling of RMgBr with TEMPO (1.08 equiv) in
THF under different conditions.
Entry
Product
R
T [8C]
t [min]
Yield [%]
1[a]
2[a]
3[a]
4[a]
5[b]
6
1a
1a
1a
1a
1a
1b
1c
1d
1e
1 f
1g
1h
1i
1j
1k
1l
1m
1m
1n
1o
1p
1q
1r
C6H5
C6H5
C6H5
C6H5
20
20
20
66
66
66
66
66
66
66
66
66
66
66
66
66
20
66
66
66
66
66
66
66
10
20
40
5
5
20
10
15
15
30
10
30
20
10
25
25
4320
240
360
300
300
300
300
300
78
91
92
98
98
86
86
87
77
84
87
81
79
96
50[d]
64[d]
55
90
86
94
94
76
65
72
C6H5
4-CH3C6H4
4-CH3O-C6H4
4-(CH3)2NC6H4
4-FC6H4
3-CH3C6H4
3-CH3O-C6H4
2-CH3C6H4
2-CH3OC6H4
b-naphthyl
C6H5CH CH
n-C6H13CH CH
C6H5C C
C6H5C C
4-CH3OC6H4C C
4-CF3C6H4C C
n-C6H13C C
C6H11C C
TMS-C C
1-c-hexenyl-C C
7[a]
8
9
10
11[a]
12
13
14
15[c]
16[c]
17
18
19
20
21
22
23
24
Scheme 1. Reaction of TEMPO with organometallic compounds.
shown by Renaud and co-workers to react with TEMPO in a
similar way.[3] To our knowledge, the reactivity of TEMPO
towards aryl, alkenyl, and alkynyl magnesium compounds has
not been investigated previously. As aryl, alkenyl, and alkynyl
radicals are destabilized, oxidation of the corresponding
magnesium compounds by TEMPO should not occur. We
therefore expected a different reaction outcome. Herein we
report a highly efficient homocoupling of organomagnesium
compounds with TEMPO. Such reactions are usually con-
ducted with transition-metal catalysts.[4] In pioneering studies,
=
=
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
1s
[a] The aryl magnesium bromide was used as received from Acros after
titration. [b] The reaction was conducted with C6H5MgCl (Acros). [c] The
=
Grignard reagent trans-R’CH CHMgX was prepared by the transmeta-
[*] M. S. Maji, Prof. Dr. A. Studer
=
lation of trans-R’CH CHI with iPrMgCl; see Ref. [7]. [d] E,E/E,Z/Z,Z>
NRW Graduate School of Chemistry
99:1:0.
Organisch-Chemisches Institut, Westfꢀlische Wilhelms-Universitꢀt
Corrensstrasse 40, 48149 Mꢁnster (Germany)
Fax: (+49)251-83-36523
E-mail: studer@uni-muenster.de
temperature for 10 min, 1a was obtained in 78% yield
(Table 1, entry 1). The yield was improved by extending the
reaction time to 40 min (92%; Table 1, entries 2 and 3). An
even better result was observed when the reaction was carried
out at reflux in THF (98%; Table 1, entry 4). Compound 1a
was also obtained in 98% yield from PhMgCl. Thus, the
halide ion of the Grignard reagent does not appear to
influence the reaction (Table 1, entry 5). All subsequent
experiments were conducted in THF at reflux with organo-
magnesium bromides. To study the scope and limitations of
T. Pfeifer
Institut fꢁr Anorganische und Analytische Chemie
Westfꢀlische Wilhelms-Universitꢀt
Corrensstrasse 30, 48149 Mꢁnster (Germany)
[**] A.S. thanks the Novartis Pharma AG for financial support (Novartis
Young Investigator Award). We thank the NRW Graduate School of
Chemistry for support of our work (scholarship to M.S.M.).
TEMPO=2,2,6,6-tertramethylpiperidine-N-oxyl.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200804197.
Angew. Chem. Int. Ed. 2008, 47, 9547 –9550
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9547

Communications
the homocoupling, various organomagnesium compounds
TEMPO (14 mol%) for 10 min in THFat reflux. The reaction
mixture was then purged with dioxygen for another 10 min.
The renewed addition of PhMgBr (14 mol%, 10 min reaction
time) was followed by purging with O2 (10 min).[9] The
reaction sequence (addition of PhMgBr and treatment with
O2), which takes about 20 min, was then repeated a further six
times.[10] With in situ recycling by this procedure, 1a was
obtained in 81% yield along with phenol in 8% yield
(Table 2, entry 1). At room temperature under similar con-
ditions (with 15 mol% TEMPO), biphenyl was formed in
74% yield (Table 2, entry 2). A decrease in the amount of
TEMPO used to 10 mol% led to the formation of 1a in
slightly lower yield (Table 2, entry 3).
were treated under the optimized conditions with TEMPO.
Commercially available Grignard reagents were used as
received after titration. All other Grignard reagents were
freshly prepared from the corresponding bromides with
magnesium turnings and a catalytic amount of I2 and titrated
before use.
Aryl Grignard reagents with electron-rich and electron-
poor substituents in the para position reacted to give the
corresponding biphenyls 1b–e in high yield (77–87%; Table 1,
entries 6–9). Similar results were obtained with meta-substi-
tuted aryl Grignard compounds (Table 1, entries 10 and 11),
and ortho-substituted congeners were transformed into the
corresponding biphenyls 1h and 1i in good yield (Table 1,
entries 12 and 13). The best result was observed for the
oxidative coupling of b-naphthylmagnesium bromide (96%;
Table 1, entry 14). Moreover, we found that the oxidative
homocoupling of vinyl Grignard reagents to provide dienes is
possible with TEMPO: b-Styrylmagnesium chloride and
Table 2: Homocoupling of RMgBr with catalytic amounts of TEMPO.
Entry Product
R
TEMPO [mol%] t [min] Yield [%]
1
1a
1a
1a
1b
1c
1d
1e
1 f
1g
1h
1i
C6H5
C6H5
C6H5
14
15
10
14
14
15
15
15
15
15
14
15
18
10
25
10
15
10
15
10
15
10
25
25
10
25
81
74
76
80
81
84
61
63
57
44
70
81
64[c]
2[a]
3
=
C6H13CH CHMgCl were transformed into the dienes 1k
and 1l, respectively (Table 1, entries 15 and 16). Alkynyl
magnesium compounds also underwent TEMPO-mediated
coupling;[6] however, longer reaction times (4–6 h) were
necessary. At room temperature, the homocoupling of
4
5
6
7
8
9
10
11
4-CH3C6H4
4-CH3OC6H4
4-(CH3)2NC6H4
4-FC6H4
3-CH3C6H4
3-CH3OC6H4
2-CH3C6H4
2-CH3OC6H4
b-Naphthyl
ꢀ
C6H5C CMgBr was very slow (Table 1, entry 17); at reflux
ꢀ
in THF, however, C6H5C CMgBr and other aryl alkynyl
organomagnesium compounds underwent highly efficient
homocoupling to give the corresponding products 1m–o in
86–94% yield (Table 1, entries 18–20). The diynes 1p and 1q
were formed in good yield from the corresponding alkyl
alkynyl Grignard reagents (Table 1, entries 21 and 22), and
trimethylsilylethynyl magnesium bromide was transformed
into 1r in 65% yield (Table 1, entry 23). Enynes also under-
went homocoupling under these conditions (Table 1,
entry 24).
12
1j
1k
13[b]
C6H5CH CH
=
[a] The reaction was conducted at room temperature. [b] The Grignard
=
reagent was prepared from trans-C6H5CH CHBr with Mg turnings.
[c] The product was obtained as a mixture of isomers: E,E/E,Z/Z,Z
18:4:1.
Importantly, we found that TEMPO–MgBr (2), which was
formed as a by-product, was reoxidized readily to TEMPO
with dioxygen in refluxing THF within 10 min, as indicated by
TLC [Eq. (1)]. Following this observation, we investigated
the TEMPO-catalyzed aerobic oxidation of aryl Grignard
reagents.[8]
We tested the aerobic TEMPO-mediated homocoupling
under the optimized conditions with other substrates. For less
reactive Grignard reagents, we increased the reaction time
(up to 25 min; see Table 2). The para-substituted biphenyl
derivatives 1b–e were isolated in moderate to good yield (61–
84%) when a catalytic amount of TEMPO was used (14 or
15 mol%). Thus, the yields were slightly lower than for the
equivalent reactions in the presence of a stoichiometric
amount of TEMPO (see Table 1), but still good. Aryl
Grignard compounds with meta and ortho substituents
underwent homocoupling in the presence of substoichiomet-
ric amounts of TEMPO in 44–70% yield (Table 2, entries 8–
11). A very good result was observed with b-naphthylmagne-
sium bromide (Table 2, entry 12), and dienes were also
accessible by this method (Table 2, entry 13).
Since the oxidation of alkynyl magnesium bromide
derivatives with TEMPO was slow, we did not believe that
the procedure for the in situ regeneration of TEMPO would
be efficient with these systems. To our surprise, we found that
alkynyl magnesium bromide derivatives could transformed
into the corresponding diynes with dioxygen without the
addition of TEMPO. Thus, when O2 was bubbled through a
When O2 was bubbled through a solution of PhMgBr at
room temperature in the absence of TEMPO, biphenyl (1a)
was obtained in 5% yield. The same experiment at reflux in
THF gave 1a in 12% yield. We repeated this experiment in
the presence of TEMPO (20 mol%): Biphenyl (1a) was
isolated in 61% yield along with phenol (24%). Thus, the
reoxidation of TEMPO–MgBr with O2 is too slow to
completely suppress the reaction of PhMgBr with O2. We
therefore developed a procedure for the in situ recycling of
TEMPO. To this end, PhMgBr (14 mol%) was treated with
ꢀ
solution of C6H5C CMgBr in THF at reflux for 2 h, the diyne
1m was obtained in 60% yield (Table 3, entry 2). The same
9548 www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9547 –9550

Angewandte
Chemie
Table 4: Trace analysis (all data in ppm= mggÀ1).
ꢀ
Table 3: Oxidative homocoupling of RC CMgBr with dioxygen.
Entry
Product
R
T [8C]
t [h]
Yield [%]
Metal
Table 1
entry 10
Table 3
entry 2
TEMPO
iPrMgCl
ꢀ
1
2
1m
1m
1m
1n
1o
1p
C6H5C C
C6H5C C
C6H5C C
4-CH3OC6H4C C
4-CF3C6H4C C
n-C6H13C C
C6H11C C
20
66
66
66
66
66
66
2
2
2
2
2
4
4
10
60
–
62
52
46
41
55Mn
56Fe
7.78Æ0.28
8.35Æ0.39
1.19Æ0.49
<LOQ[a]
21.5Æ0.1
11.2Æ1.6
2.12Æ1.03
<LOQ[a]
0.28Æ0.03
26.6Æ2.7
<LOQ[a]
12.6Æ0.03
1.81Æ1.22
1.14Æ0.61
<LOQ[a]
ꢀ
3[a]
4
ꢀ
63Cu
105Pd
ꢀ
<LOQ[a]
ꢀ
5
6
7
ꢀ
[a] LOQ=limit of quantification.
ꢀ
1q
[a] The reaction was conducted in the absence of O2.
(13 ppm). Another source of transition metals, particularly of
Fe (27 ppm), is TEMPO. We also analyzed the solvent used
(THF). However, all metals tested for were below the limit of
quantification. In the reported oxidative iron- and manga-
nese-catalyzed homocoupling reactions, 5 mol% of the metal
had to be used.[8] Copper-mediated homocoupling reactions
of aryl Grignard reagents are generally conducted with a
stoichiometric amount of a copper salt,[12] and copper-
catalyzed acetylide homocoupling reactions require around
5 mol% of the copper catalyst.[13] Therefore, we believe that it
is unlikely that the trace amounts of transition metals present
are able to catalyze these coupling reactions efficiently. As
also suggested by Knochel, Mayr, and co-workers, we
currently assume that these coupling reactions proceed
without the aid of a transition metal.
In conclusion, we have described “transition-metal-free”
homocoupling reactions of various organomagnesium com-
pounds in the presence of commercially available TEMPO as
an organic oxidant. The reactions could be conducted with
15 mol% of TEMPO by using dioxygen as the terminal
oxidant. Moreover, we found that alkynyl magnesium com-
pounds underwent homocoupling to provide the correspond-
ing diynes upon treatment with dioxygen at higher temper-
atures in the absence of a catalyst.
experiment at room temperature provided 1m in only 10%
yield (Table 3, entry 1). In the absence of O2 under otherwise
identical conditions, 1m was not identified in the reaction
mixture (Table 3, entry 3). To our knowledge, the high-
yielding homocoupling of Grignard compounds with dioxy-
gen in the absence of a transition-metal catalyst is unprece-
dented.[11] Aryl alkynyl magnesium compounds with para
substituents also underwent homocoupling under aerobic
conditions in satisfactory yield (Table 3, entries 4 and 5).
Lower yields were observed for the oxidative homocoupling
of alkyl alkynyl organomagnesium derivatives (Table 3,
entries 6 and 7).
The mechanism of the TEMPO-mediated oxidation of
organomagnesium compounds is not yet known. Since the
treatment of the Grignard reagent 3 with TEMPO afforded
exclusively the homocoupling product 4 (76%), without any
trace of products derived from a 5-exo radical cyclization, we
exclude free aryl radicals as intermediates at present
[Eq. (2)].
Received: August 25, 2008
Published online: November 3, 2008
À
Keywords: biaryls · C C coupling · dienes · diynes · nitroxides
.
As the mechanism was not clear, and aryl Grignard
reagents are known to undergo homocoupling in the presence
of Mn and Fe catalysts under oxidative conditions,[8] we
decided to measure the concentrations of transition metals in
our reaction mixtures (Table 4; see also the Supporting
Information). We analyzed for the presence of Fe and Mn,
but also for traces of Pd and Cu, as these transition metals are
also known to mediate the homocoupling of Grignard
reagents. The measurements were performed on reaction
mixtures for an aryl coupling and an alkynyl coupling.
Palladium was detected in the reaction mixtures, but could
not be quantified (< 0.01 ppm). In the reaction mixture for
the aryl homocoupling, we found Mn and Fe, each at a
concentration of around 8 ppm, and Cu (1 ppm). Slightly
larger amounts of these transition metals were detected in the
crude reaction mixture for the alkynyl homocoupling (Mn:
22 ppm, Fe: 11 ppm, Cu: 2 ppm). The alkynyl Grignard
reagent was prepared by using commercially available
iPrMgCl, in which we identified a significant amount of Mn
[1] G. M. Whitesides, T. L. Newirth, J. Org. Chem. 1975, 40, 3448; T.
Nagashima, D. P. Curran, Synlett 1996, 330; P. I. Dalko, Tetrahe-
dron Lett. 1999, 40, 4035.
[2] For reviews on the use of TEMPO in synthesis, see: a) T. Vogler,
A. Studer, Synthesis 2008, 2163; b) A. Studer, T. Schulte, Chem.
Rec. 2005, 5, 27; c) A. Studer, Chem. Soc. Rev. 2004, 33, 267;
d) A. Studer, Chem. Eur. J. 2001, 7, 1159; for the use of TEMPO
À
as an oxidant in rhodium-catalyzed C H arylation reactions, see:
e) T. Vogler, A. Studer, Org. Lett. 2008, 10, 129.
[3] A.-P. Schaffner, P. Renaud, Eur. J. Org. Chem. 2004, 2291; for a
review, see: C. Ollivier, P. Renaud, Chem. Rev. 2001, 101, 3415.
[4] J. Hassan, M. Sꢀvignon, C. Cozzi, E. Schulz, M. Lemaire, Chem.
Rev. 2002, 102, 1359.
[5] A. Krasovskiy, A. Tishkov, V. del Amo, H. Mayr, P. Knochel,
Angew. Chem. 2006, 118, 5132; Angew. Chem. Int. Ed. 2006, 45,
5010.
[6] For the coupling of alkynyl metal compounds with dinitrogen
tetroxide, see: C. J. Woltermann, H. Shechter, Helv. Chim. Acta
2005, 88, 354.
[7] a) P. Knochel, W. Dohle, N. Gommermann, F. F. Kneisel, F.
Kopp, T. Korn, I. Sapountzis, V. A. Vu, Angew. Chem. 2003, 115,
Angew. Chem. Int. Ed. 2008, 47, 9547 –9550
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 9549

Communications
4438; Angew. Chem. Int. Ed. 2003, 42, 4302; b) H. Ren, A.
[11] For transition-metal-catalyzed aerobic oxidations, see: J. Piera,
J.-E. Bꢁckvall, Angew. Chem. 2008, 120, 3558; Angew. Chem. Int.
Ed. 2008, 47, 3506.
[12] a) B. H. Lipshutz, K. Siegmann, E. Garcia, Tetrahedron 1992, 48,
2579; b) R. S. Coleman, E. B. Grant, Tetrahedron Lett. 1993, 34,
2225; c) C. M. P. Kronenburg, C. H. M. Amijs, P. Wijkens, J. T. B.
Jastrzebski, G. van Koten, Tetrahedron Lett. 2002, 43, 1113.
[13] A. S. Hay, J. Org. Chem. 1962, 27, 3320.
Krasovskiy, P. Knochel, Org. Lett. 2004, 6, 4215.
[8] G. Cahiez, A. Moyeux, J. Buendia, C. Duplais, J. Am. Chem. Soc.
2007, 129, 13788.
[9] It was not necessary to stir the solution to remove the remaining
O2 (after purging with O2) prior to the renewed addition of
PhMgBr. (The product was obtained in the same yield with or
without this extra step.)
[10] The following amounts of PhMgBr were added: 13 mol% in the
third and fourth cycles, 12 mol% in the fifth and sixth cycles, and
11 mol% in the seventh and eighth cycles.
9550
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 9547 –9550