Kumada - Corriu coupling of Grignard reagents, probed with a chiral Grignard reagent

Article abstract of DOI:10.1039/b300033h

The Grignard reagent 2, in which the magnesium-bearing carbon atom is the sole stereogenic centre has been coupled with vinyl bromide under Pd(0) or Ni(0)-catalysis to give compound 3 with full retention of configuration. Coupling using Fe(acac)₃

Full text of DOI:10.1039/b300033h

Kumada–Corriu coupling of Grignard reagents, probed with a chiral
Grignard reagent
Bettina Hölzer and Reinhard W. Hoffmann*
Fachbereich Chemie, Philipps-Universität Marburg, Hans-Meerwein-Strasse, D-35032 Marburg, Germany.
E-mail: rwho@chemie.uni-marburg.de; Fax: +49 6421 2825677; Tel: +49 6421 2825571
Received (in Cambridge, UK) 2nd January 2003, Accepted 29th January 2003
First published as an Advance Article on the web 19th February 2003
The Grignard reagent 2, in which the magnesium-bearing
carbon atom is the sole stereogenic centre has been coupled
with vinyl bromide under Pd(0) or Ni(0)-catalysis to give
compound 3 with full retention of configuration. Coupling
using Fe(acac)3 or Co(acac)2 as catalyst was accompanied by
considerable racemisation. These findings are discussed with
respect to a dichotomy between concerted polar and stepwise
SET transmetallation pathways.
at which reagent 2 is configurationally stable. Kumada–Corriu-
couplings were effected by reducing the pre-catalyst (0.1 equiv.
rel. to total Mg) with EtMgCl at 0 °C, addition of vinyl bromide
at 278 °C and reaction with 2 for 5 d at 278 °C. Workup
furnished the coupling product 3 and the hydride-elimination
product 4. The results are compiled in Table 1.
The transition metal catalyzed cross-coupling of Grignard
reagents with vinylic and aryl halides was the subject of both the
exploratory phase1 of the cross-coupling reaction as well as of
the ground breaking studies by Corriu2 and by Kumada.3,4 After
30 years of expansion of the transition metal catalyzed cross-
coupling to other organometallic reagents,5 the Grignard cross-
coupling reactions are presently finding renewed interest.6 The
principal mechanistic scenario of the cross-coupling reaction is
by now common text-book material. However, especially the
earlier contributions provided ample evidence that free radicals
may be generated in the context of the transition metal catalyzed
cross-coupling of Grignard reagents. This raises the question
whether the generation of radicals is part of the catalytic cycle,
or rather of related or unrelated side-reactions. Radicals could
be generated, if electron motion precedes nuclear motion in the
transmetallation step. Such an SET-step could be detected on
transmetallation of an enantiomerically enriched chiral secon-
dary Grignard reagent, as the secondary alkyl radical formed in
such an oxidation step would be achiral on the time scale of the
experiment.7 Hence, racemic coupling product should result in
the case of an SET step. In the absence of such an electron
transfer the coupling products should be formed with (full)
retention of configuration.
The absolute configuration of the coupling product 3 was
established by chemical correlation with the known11 car-
boxylic acid 5. The lengthy sequence from 5 to 3 was chosen
because it did not involve any racemisation-prone inter-
mediates. It provided the product 3 of 98% e.e as indicated by
GC-analysis using a chiral stationary phase.
We recently gained access to the enantiomerically enriched
(ca. 91% e.e.) secondary Grignard reagent 2, in which the
magnesium bearing carbon atom is the sole stereogenic center.8
Reagent 2 is thus ideally suited to determine the mechanisms of
Grignard reactions.9 The reagent 2 was generated from the
sulfoxide 1 by addition of 5 equiv. of EtMgCl. We have to keep
in mind that this reaction results in a cocktail of Grignard
reagents and sulfoxides in which the reagent 2 of interest is only
a constituent.
The results in Table 1 show that both Ni(0)- and Pd(0)-
mediated coupling reactions proceeded with essentially full
retention of configuration establishing the transmetallation of 2
to Ni(II) or Pd(II) as a concerted SE2-ret process.12 The Grignard
Table 1 Kumada-type couplings of the Grignard reagent 2 of ca. 90% e.e.
with vinyl bromide in THF at 278 °C
Catalyst
3/4-ratio
3: yield (%) e.e. (%)a
1
2
3
4
5
6
NiCl2dppf
95+5
60
80
58
55
35
30
88
NiCl2(2)diop
PdCl2dppf
PdCl2(2)diop
Fe(acac)3
89+11
95+5
89b
88
86+14
> 95+ < 5
> 95+ < 5
89b
53
55
We report here on the stereochemistry of the transmetallation
of this reagent to Ni(II), Pd(II), Fe(II?), and Co(II?) in the course
of Kumada–Corriu coupling reactions with vinyl bromide. The
latter reaction can be carried out at low temperature10 (278 °C),
Co(acac)2
a Determined by GC on a Chiral-Dex column. b Racemic 2 resulted in
racemic 3 under these conditions.
732
CHEM. COMMUN., 2003, 732–733
This journal is © The Royal Society of Chemistry 2003

View Article Online
reagent 2 therefore behaves in this respect like secondary alkyl-
boron13 and alkyl-zinc reagents.14.
well as the Fonds der Chemischen Industrie for support of this
study.
In contrast, when the coupling was catalyzed by low-valent
Fe or Co generated from Fe(acac)3 or Co(acac)2 in THF–NMP15
the e.e. of the coupling product 4 was significantly reduced.
This points to a situation, in which SET between 2 and 6, e.g.
to give 8 and 9 via path (b), is involved in the transmetallation
step. As the vinyl metal species 9 (with the oxidation state
changed by one unit compared to 7) can be considered to be
persistent at the time-scale of the experiment, the persistent
radical effect16 would then guarantee an effective recombina-
tion of 8 and 9 to give (now racemic) 7. But a reaction sequence
via path (b) is by no means established, especially since the
oxidation state of the iron or cobalt species 6 in the catalytic
cycle is not known6 and, hence, it is not known whether their
oxidation potential would be high enough to oxidize a Grignard
reagent.
Notes and references
1 (a) M. S. Kharash and C. F. Fields, J. Am. Chem. Soc., 1943, 65, 504;
(b) M. S. Kharash and O. Reinmuth, Grignard Reactions of Nonmetallic
Substances, Prentice Hall, New York, 1954; (c) M. Tamura and J. K.
Kochi, J. Am. Chem. Soc., 1971, 93, 1487.
2 R. J. P. Corriu and J. P. Masse, J. Chem. Soc., Chem. Commun., 1972,
144.
3 K. Tamao, K. Sumitani and M. Kumada, J. Am. Chem. Soc., 1972, 94,
4374.
4 The historical aspects have recently been reviewed: K. Tamao, T.
Hiyama and E.- i. Negishi, J. Organomet. Chem., 2002, 653, 1; and
following papers.
5 Metal-catalyzed Cross-coupling Reactions, ed. F. Diederich and P. J.
Stang, Wiley-VCH, Weinheim, Germany, 1998.
6 A. Fürstner, A. Leitner, M. Méndez and H. Krause, J. Am. Chem. Soc.,
2002, 124, 13856.
7 D. Griller, K. U. Ingold, P. J. Krusic and H. Fischer, J. Am. Chem. Soc.,
1978, 100, 6750.
8 R. W. Hoffmann, B. Hölzer, O. Knopff and K. Harms, Angew. Chem.,
2000, 112, 3206; R. W. Hoffmann, B. Hölzer, O. Knopff and K. Harms,
Angew. Chem., Int. Ed., 2000, 39, 3072.
9 (a) R. W. Hoffmann and B. Hölzer, Chem. Commun., 2001, 491; (b) R.
W. Hoffmann, B. Hölzer and O. Knopff, Org. Lett., 2001, 3, 1945–1948;
(c) R. W. Hoffmann and B. Hölzer, J. Am. Chem. Soc., 2002, 124,
4204.
10 G. Consiglio and C. Botteghi, Helv. Chim. Acta, 1973, 56, 460.
11 W. Kirmse, P. Feyen, W. Gruber and W. Kapmeyer, Chem. Ber., 1975,
108, 1839.
12 R. E. Gawley, Tetrahedron Lett., 1999, 40, 4297.
13 (a) B. H. Ridgeway and K. A. Woerpel, J. Org. Chem., 1998, 63, 458;
(b) K. Matos and J. A. Soderquist, J. Org. Chem., 1998, 63, 461.
14 (a) A. Boudier, C. Darcel, F. Flachsmann, L. Micouin, M. Oestreich and
P. Knochel, Chem. Eur. J., 2000, 6, 2748; (b) A. Boudier, E. Hupe and
P. Knochel, Angew. Chem., 2000, 112, 2396; A. Boudier, E. Hupe and
P. Knochel, Angew. Chem., Int. Ed., 2000, 39, 2294.
15 G. Cahiez and H. Avedissian, Tetrahedron Lett., 1998, 39, 6159.
16 (a) A. Studer, Chem. Eur. J., 2001, 7, 1159; (b) H. Fischer, Chem. Rev.,
2001, 101, 3581.
Therefore an alternate mechanistic scenario should also be
discussed, in which transmetallation could occur with full
retention of configuration (path (a)). As the alkyl-cobalt or the
alkyl-iron species 7 formed in this manner has a weak carbon–
metal bond,17 reversible bond homolysis18,19 to 8 and 9
(equilibrium (c)) could then lead to racemisation of the alkyl
residue. For this scenario to account for the observed partial
racemisation in the cobalt or iron-mediated cross coupling of 2
to 3, it is necessary that spontaneous thermal bond homolysis of
7 would have to occur with a much higher rate than turnover in
the catalytic cycle. It is quite uncertain, though, whether this
condition will be met at the low temperature of 278 °C.19 For
this reason we tend to ascribe the partial racemisation observed
in the cross-coupling of 2 with vinyl bromide under iron- or
cobalt catalysis to an SET process in the transmetallation step
(i.e. path (b)).
17 (a) B. G. Daikh and R. G. Finke, J. Am. Chem. Soc., 1991, 113, 4160;
(b) J. Halpern, Acc. Chem. Res., 1982, 15, 238.
18 B. G. Daikh and R. G. Finke, J. Am. Chem. Soc., 1992, 114, 2938.
19 Walling20 presented arguments that homolytic cleavage of (RCH2)2Fe is
unimportant even for reactions of Grignard reagents with Fe(II) carried
out at room temperature. Of course, secondary (R2CH)2Fe species are
thermally much less stable21
.
20 C. Walling, J. Am. Chem. Soc., 1988, 110, 6846.
We are grateful to the Deutsche Forschungsgemeinschaft
(SFB 260 and Graduiertenkolleg Metallorganische Chemie) as
21 W. Lau, J. C. Huffman and J. K. Kochi, Organometallics, 1982, 1,
155.
CHEM. COMMUN., 2003, 732–733
733