A well-defined low-valent cobalt catalyst Co(PMe3)4 with dimethylzinc: a simple catalytic approach for the reductive dimerization of benzyl halides

Article abstract of DOI:10.1039/c6nj03265f

Herein, we report the first catalytic version of a cobalt-catalysed reductive homocoupling of benzyl halides which proceeds with low catalyst loadings (0.5 to 5 mol%). By synthetizing each cobalt intermediate we demonstrate that reaction proceeds through two single electron transfers (SET) and that dimethylzinc is only involved in the regeneration of the catalytic species.

Full text of DOI:10.1039/c6nj03265f

NJC
View Article Online
View Journal
LETTER
A well-defined low-valent cobalt catalyst Co(PMe3)4
with dimethylzinc: a simple catalytic approach for
the reductive dimerization of benzyl halides†
Cite this:DOI: 10.1039/c6nj03265f
Received (in Montpellier, France)
18th October 2016,
Accepted 14th November 2016
Brendan J. Fallon, Vincent Corce, Muriel Amatore, Corinne Aubert, Fabrice Chemla,
´
Franck Ferreira, Alejandro Perez-Luna and Marc Petit*
DOI: 10.1039/c6nj03265f
www.rsc.org/njc
Herein, we report the first catalytic version of a cobalt-catalysed in about 10% yield.10 Having recently demonstrated the ability
reductive homocoupling of benzyl halides which proceeds with low of low valent cobalt complexes (Co(PMe3)4 and HCo(PMe3)4) to
catalyst loadings (0.5 to 5 mol%). By synthetizing each cobalt participate in C–H functionalisation and hydrosilylation
intermediate we demonstrate that reaction proceeds through two reactions,11 we began to investigate their utility for the homo-
single electron transfers (SET) and that dimethylzinc is only involved coupling of benzyl halides. As before, we examined the use of
in the regeneration of the catalytic species.
Me2Zn to regenerate the active catalytic species. Gratifyingly,
the addition of 0.6 equivalent of dimethylzinc allowed the
The formation of C–C bonds is a fundamental and powerful product of homocoupling to be formed efficiently. The use of
approach in organic synthesis to design the skeletal framework sub-stoichiometric zinc in this reaction is in sharp contrast to
for organic molecules.1 In particular, dibenzyl derivatives are key previously reported systems in which a large excess of metal or
intermediates in the synthesis of pharmaceuticals, agrochemicals reducing reagents based on Mn,12a–c Mg,12c–e Zn10 etc. is required
and polymers.2 In addition, many natural products are dimers or to achieve catalytic turnover.13 The disclosed reaction proceeds
pseudodimers and the ability of dimers to modulate proteins or under mild reaction conditions (40 to 66 1C, no acid nor base), at
DNA is well known.3 While C–C bond formation involving Csp low catalyst loadings (0.5 to 5 mol%) and in a short reaction time
and Csp2 has been extensively explored, homocoupling reactions (10 to 60 min). In addition the procedure is easily scalable to
between two sp3-hybridised carbons remain comparatively rare.4 10 mmol without any decrease in yield. It is worth noting the
Many of these homocoupling reactions have been carried out in reaction proceeds at room temperature in 18 h with only a slight
the presence of both stoichiometric and catalytic amounts of decrease in yield. Me2Zn, Et2Zn and Me3Al were all investigated as
metal such as nickel,5 rhodium,6 or palladium7 among others. transmetallation reagents with Me2Zn proving as the most efficient.
Many of the current catalytic systems suffer several drawbacks The dimerisation is compatible with a variety of solvents: it proceeds
such as: (i) use of costly precious metals,7a (ii) use of additives5e or most rapidly in toluene however direct substitution of the benzyl
(iii) harsh and long reaction times.5f Homocoupling using cobalt halide by Me2Zn is an issue. This problem can be completely
catalysis has mainly been reported under oxidative conditions.8 eradicated on switching the reaction medium to THF.
The use of stoichiometric amounts of ClCo(PPh3)3 is known to
With our optimal conditions in hand (Scheme 1), we applied
promote the reductive homocoupling of benzyl and allyl halides, them to a variety of benzyl halides and we were pleased to
however, no catalytic activity of this complex has been reported.9 observe that the functional group tolerance was broad (Table 1).
Inspired by previous results based on rhodium, the use of dialkylzinc
reagents to perform a double transmetallation followed by a
reductive elimination appeared an attractive approach to
achieve a catalytic turnover using ClCo(PPh3)3.6 However, the
use of ClCo(PPh3)3 (10 mol%) and Me2Zn (1.2 equiv.) with
benzyl bromide only yielded the desired homocoupled product
´
´
Sorbonne Universites, UPMC Univ Paris 06, Institut Parisien de Chimie Moleculaire,
UMR CNRS 8232, Case 229, 4 Place Jussieu, 75252 Paris Cedex 05, France.
E-mail: marc.petit@upmc.fr
Scheme 1 Optimal conditions for the catalytic reductive homocoupling
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6nj03265f
of benzyl halides.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016
New J. Chem.

View Article Online
Letter
NJC
Table 1 Scope of the reductive dimerization of benzyl halidesa
Entry
Substrate
Product
Yieldb (%)
1
2
3
4
5
6
7
8
R1 = H
4-Me
4-F
4-iPr
4-CF3
4-Br
4-NO2
4-CO2Me
4-OMe
3,5-t-Bu
2-I
2-Br
3-Cl
2a
2b
2c
2d
2e
2f
2g
2h
2i
2j
2k
2l
2m
90
95
95
Quant (95)c
91
72d
80
96
77
92
78
65
90
9
10
11
12
13
14
2n
85
15
16
2o
2p
Quant
83
17
18
R2 = Me
CO2Et
2q
2r
92e
60e
19
2s
90
20
21
R1 = 4-F
H
2t
2u
55 f
80 f
a
b
c
d
e
Reaction conditions: 66 1C, 10 min, THF. Isolated yields. Yield on 10 mmoles scale. Reaction was run at 40 1C. Product isolated as a 50/50
f
mixture of meso/dl. Reaction conditions: Co(PMe3)4 5 mol%, 66 1C, 60 min, THF.
para-Substituted benzyl bromides dimerized smoothly, both halide moiety or undesired aryl–aryl homocoupling was observed.
electron-donating e.g. 4-Me, 4-i-Pr, 4-OMe and electron-withdrawing Pentafluorobenzyl bromide also gave the desired product in a
substituents e.g. 4-CO2Me, 4-NO2 were tolerated (entries 1–9). satisfactory yield (entry 14). 3,4-Methylenedioxybenzyl bromide
Sterically congested benzyl halides bearing bulky substituents (entry 15) reacted in quantitative yield and could be used as a
(3,5-t-Bu) could participate as well in the homocoupling reaction building block in synthesis.2b,c Naphthalene based derivative
(entry 10). In addition ortho- and meta-substituted benzyl halides (entry 16) could also be isolated in good yield. Furthermore,
bearing chemically susceptible functional groups (entries 11–13) products that have substituents in the benzylic position were
also afforded the dimerized product in good to excellent yields obtained in moderate to good yields (entries 17 and 18).
(65–92%). In these cases, it is worth noting that no reduction of the Pleasingly we could also apply our protocol to the synthesis of
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016

View Article Online
NJC
Letter
pyridine containing compounds (entry 19) that could find use access to reaction intermediates is achievable. We first examined
as ligands in organometallic chemistry or as pharmaceutical the reaction of 10 mol% ClCo(PMe3)3 in the absence of Me2Zn.
scaffold.14 Finally benzyl chlorides have shown to undergo homo- A yield of 10% of the desired homocoupling product was
coupling albeit with slightly higher catalytic loading and longer observed, suggesting ClCo(PMe3)3 is capable of promoting the
reaction time (entries 20 and 21).
dimerization of 1 equivalent of benzyl halide (entry 5). With the
We then focused on the reaction mechanism and tried to addition of Me2Zn to the reaction mixture we were able to
shed light both on the role of dimethylzinc and on the activation reactivate the catalytic species leading to complete conversion
mode of the benzyl halide. Toward this end, we initially carried to the desired product (entry 6). Finally we exposed Cl2Co(PMe3)3
out a series of blank experiments (Table 2). In the absence of to benzyl bromide under standard reaction conditions in the
cobalt no homocoupling product was observed rather quantita- absence of Me2Zn. No conversion to the homocoupling product
tive conversion of 1a to ethylbenzene 3a through substitution of was observed but as before we could regain catalytic activity
the benzyl bromide by Me2Zn (entry 1). The use of 10 or 20 mol% through the addition of Me2Zn (entries 7 and 8).16
Co(PMe3)4 without Me2Zn furnished the dimerised product in
To further explore the presence of a radical pathway in our
respectively 20 and 40% yield (entries 2 and 3). These results reaction we examined the reductive Heck type coupling previously
suggested that the role of Me2Zn is purely regeneration of the reported by Oshima.13d,17 Using cyclohexyl bromide with styrene in
catalyst and that Co(PMe3)4 can promote the dimerisation of our standard reaction conditions we observed the Heck type product
2 equivalents of benzyl halide. Homocoupling reactions mediated 5 and the dimer 2v as already described by Oshima. This result
by a stoichiometric amount of ClCo(PPh3)3 have been reported argues in favour of a radical mechanism in our system (Scheme 2).18
to proceed through single electron transfer (SET) followed by
Putting together our control and radical trapping experiments
dimerization of the produced benzylic radical.9 Involvement of in combination with our study on the precise role of zinc we can
a radical pathway in our system was suspected. Indeed, using propose the following catalytic cycle (Scheme 3). A first single
1 equivalent of TEMPO as radical scavenger, 4a was isolated as electron transfer would generate the cobalt(I) intermediate (B)
the sole product (entry 4).15 The suggestion of a radical process and a benzylic radical that can undergo dimerization.14 Inter-
was also reinforced by the absence of any b-hydride elimination mediate (B) also undergoes SET to furnish the cobalt(II) (C)
type product in the case of substituted starting material species and a benzylic radical that again participates in dimerization
(Table 1, entry 17).
reaction. Intermediate (C) would then undergo transmetallation with
We thus hypothesised that the formation of X2Co(PMe3)3 dimethylzinc to lead to the formation of dimethyl cobalt(II) species
(X = Cl or Br) would be the result of two SET reactions. To further (D). This intermediate would evolve rapidly through a reductive
probe this hypothesis, we synthesised the postulated reaction elimination of ethane (note: bubbling is observed during
intermediates i.e. ClCo(PMe3)3 and Cl2Co(PMe3)3. Indeed, one of
the many advantages of working with well-defined catalytic
species is that mechanistic studies are greatly simplified and
Table 2 Mechanistic investigation of the reaction
Scheme 2 Radical Heck reaction.
Entry Cat
x mol% Me2Zn (equiv.) Additive Yielda (%)
1
2
3
4
5
6
7
8
—
—
10
20
10
1.2
—
—
0.6
—
0.6
—
—
—
—
99 (3a)
20b (2a)
40b (2a)
Co(PMe3)4
Co(PMe3)4
Co(PMe3)4
ClCo(PMe3)3 10
ClCo(PMe3)3 10
C2Co(PMe3)3 10
Cl2Co(PMe3)3 10
TEMPO 99c (4a)
—
—
—
—
10b (2a)
90 (2a)
—
0.6
88b (2a)
a
Yield determined by GC using decane as internal standard. b Consumption
of starting material are respectively 20, 40 and 10% for entries 1, 2
c
and 6. Reaction ran using 1.0 equiv. of TEMPO (2,2,6,6-tetramethyl-
piperidin-1-yl)oxyl.
Scheme 3 Proposed catalytic cycle.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016
New J. Chem.

View Article Online
Letter
NJC
the reaction)19 leading to the regeneration of the catalytically
active cobalt(0) species Co(PMe3)3 (A0).18
D. Alonso-Dıaz and H. Waldmann, Angew. Chem., Int. Ed.,
1998, 37, 688–749; (c) J. D. Klemm, S. L. Schreiber and
G. R. Crabtree, Annu. Rev. Immunol., 1998, 16, 569–592;
(d) C.-H. Heldin, Cell, 1995, 80, 213–224. For a review on
molecules that promote the dimerization of proteins see:
(e) P. A. Clemons, Curr. Opin. Chem. Biol., 1999, 3, 112–115.
4 (a) A. C. Frisch and M. Beller, Angew. Chem., Int. Ed., 2005,
44, 674–688; (b) B. L. H. Taylor and E. R. Jarvo, Synthesis,
2011, 2761–2765; (c) R. Jana, T. P. Pathak and M. S. Sigman,
Chem. Rev., 2011, 111, 1417–1492; (d) A. Rudolph and
M. Lautens, Angew. Chem., Int. Ed., 2009, 48, 2656–2670.
5 (a) I. Hashimoto, N. Tsuruta, M. Ryang and S. J. Tsutsumi, J. Org.
Chem., 1970, 35, 3748–3752; (b) S.-I. Inaba, H. Matsumoto and
R. D. J. Rieke, J. Org. Chem., 1984, 49, 2093–2098; (c) M. R.
Prinsell, D. A. Everson and D. J. Weix, Chem. Commun., 2010, 46,
5743–5745; (d) C. D. Mboyi, S. Gaillard, M. D. Mabaye,
N. Pannetier and J.-L. Renaud, Tetrahedron, 2013, 69,
4875–4882; (e) M. Iyoda, M. Sakaitani, H. Otsuka and M. Oda,
Chem. Lett., 1985, 127–130; ( f ) T. Chen, L. Yang, L. Li and
K.-W. Huang, Tetrahedron, 2012, 68, 6152–6157.
In summary, to the best of our knowledge we disclose the
first catalytic version of a cobalt-catalysed reductive homocoupling of
benzyl halides. Chemical susceptible motifs such as bromo-, chloro-,
iodo- and nitro-substituted benzyl halides were all smoothly
converted under our reaction conditions. In addition the use
of well-defined low valent cobalt complexes has allowed us to
gain a greater mechanistic insight than previously reported. We
also report a novel approach to the regeneration of the cobalt(0)
species through a combination with Me2Zn.
Experimental
Commercial reagents were used without any purification.
Tetrahydrofuran was purified by mean of distillation under dry
nitrogen atmosphere on benzophenone/sodium and degassed
by freeze–pump–thaw technique.
General procedure for the cobalt catalyzed homocoupling of
benzyl halides
6 K. Sato, Y. Inoue, T. Mori, A. Sakaue, A. Tarui, M. Omote,
I. Kumadaki and A. Ando, Org. Lett., 2014, 16, 3756–3759.
7 (a) A. Lei and X. Zhang, Org. Lett., 2002, 4, 2285–2288;
To a solution of Co(PMe3)4 (0.5–5 mol%) in degassed THF
(0.5 mL) was added Me2Zn, 1.0 M in hexane (0.3 mL, 0.3 mmol)
followed by benzyl halide 1 (0.5 mmol) at ambient temperature.
The reaction was then transferred to an oil bath and heated to
66 1C for between 10–60 min. The resulting mixture was
quenched with 10% HCl, and extracted with AcOEt. The AcOEt
layer was washed with brine and dried over MgSO4. The solvent
was removed in vacuo and the residue was purified by column
chromatography on silica to give the desired product.
´
(b) B. Liegault, J.-L. Renaud and C. Bruneau, Chem. Soc.
Rev., 2008, 37, 290–299.
8 (a) S.-Y. Chen, J. Zhang, Y.-H. Li, J. Wen, S.-Q. Bian and X.-Q. Yu,
Tetrahedron Lett., 2009, 50, 6795–6797; (b) M. Mayer,
W. Maximilian and A. J. von Wangelin, Synlett, 2009, 2919–2923.
9 (a) Y. Yamada and D.-I. Momose, Chem. Lett., 1981, 1277–1281;
(b) D.-I. Momose, K. Iguchi, T. Sugiyama and Y. Yamada,
Tetrahedron Lett., 1983, 24, 921–924; (c) S. Goswami and
A. K. Mahapatra, Tetrahedron Lett., 1998, 39, 1981–1984;
(d) S. Goswami, A. K. Mahapatra and R. J. Mukherjee,
J. Chem. Soc., Perkin Trans. 1, 2001, 2717–2726; (e) N. Oka,
Y. S. Sanghvi and E. A. Theodorakis, Synlett, 2004, 823–826.
10 The use of zinc metal to generate ClCo(PPh3)3 from Cl2Co(PPh3)2
is known see: S. L. Baysdon and L. S. Liebeskind, Organometallics,
1982, 1, 771–775. However, we did not consider this approach in
order to avoid benzylzinc halide formation by zinc insertion.
11 (a) S. Ventre, E. Derat, M. Amatore, C. Aubert and M. Petit,
Adv. Synth. Catal., 2013, 355, 2584–2590; (b) B. J. Fallon,
E. Derat, M. Amatore, C. Aubert, F. Chemla, F. Ferreira,
A. Perez-Luna and M. Petit, J. Am. Chem. Soc., 2015, 137,
2448–2451; (c) B. J. Fallon, J. B. Garsi, E. Derat, M. Amatore,
C. Aubert and M. Petit, ACS Catal., 2015, 5, 7493–7497;
(d) B. J. Fallon, E. Derat, M. Amatore, C. Aubert, F. Chemla,
F. Ferreira, A. Perez-Luna and M. Petit, Org. Lett., 2016, 18,
Acknowledgements
This work was supported by CNRS, MRES, UPMC and ANR
(ANR-12-BS07-0031-01COCACOLIGHT), which we gratefully
acknowledge.
Notes and references
1 (a) D. J. Weix, Acc. Chem. Res., 2015, 48, 1767–1775; (b) D. A.
Everson and D. J. Weix, J. Org. Chem., 2014, 79, 4793–4798;
(c) C. C. C. Johansson Seechurn, M. O. Kitching, T. J. Colacot
and V. Snieckus, Angew. Chem., Int. Ed., 2012, 51, 5062–5085;
(d) A. D. Meijere and F. Diederich, Metal-Catalyzed Cross-
Coupling Reactions, Wiley-VCH, Weinheim, 2nd edn, 2004,
pp. 1–938.
´
2292–2295; (e) A. Rivera-Hernandez, B. J. Fallon, S. Ventre,
2 (a) E. Iwasa, Y. Hamashima, S. Fujishiro, D. Hashizume and
M. Sodeoka, Tetrahedron, 2011, 67, 6587–6599; (b) T. N. Barrett,
D. C. Braddock, A. Monta, M. R. Webb and A. J. P. White, J. Nat.
Prod., 2011, 74, 1980–1984; (c) X.-J. Duan, X.-M. Li and B.-G.
Wang, J. Nat. Prod., 2007, 70, 1210–1213.
3 For a review of polyvalent interactions in biology see:
(a) M. Mammen, S.-K. Choi and G. M. Whitesides, Angew.
Chem., Int. Ed., 1998, 37, 2754–2794. For the importance
of dimers in signal transduction see: (b) K. Hinterding,
C. Simon, M.-H. Tremblay, G. Gontard, E. Derat, M. Amatore,
C. Aubert and M. Petit, Org. Lett., 2016, 18, 4242–4245.
12 (a) S. Pal, S. Chowdhury, E. Rozwadowski, A. Auffrant and
C. Gosmini, Adv. Synth. Catal., 2016, 358, 2431–2435; (b) Y. Cai,
X. Qian and C. Gosmini, Adv. Synth. Catal., 2016, 358,
2427–2430; (c) Y. Suh, J.-S. Lee, S.-H. Kim and R. D. Rieke,
J. Organomet. Chem., 2003, 684, 20–36; (d) W. S. Lindsay,
P. Stokes, L. G. Humber and V. Boekelheide, J. Am. Chem. Soc.,
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016

View Article Online
NJC
1961, 83, 943–949; (e) X. Xu, D. Cheng and W. Pei, J. Org. Chem.,
Letter
L.-M. Chamoreau, E. Derat, J.-P. Goddard, C. Ollivier and
2006, 71, 6637–6639.
L. Fensterbank, Angew. Chem., Int. Ed., 2015, 54,
´ ˆ
´
13 For selected reviews see: (a) C. Gosmini and A. Moncomble,
Isr. J. Chem., 2010, 50, 568–576; (b) W. Hess, J. Treutwein
and G. Hilt, Synthesis, 2008, 3537–3562; (c) G. Cahiez
11414–11418; (c) C. Leveque, L. Chenneberg, V. Corce,
C. Ollivier and L. Fensterbank, Chem. Commun., 2016, 52,
9877–9880.
and A. Moyeux, Chem. Rev., 2010, 110, 1435–1462; 16 For further discussion on the role of zinc see the ESI†.
(d) H. Yorimitsu and K. Oshima, Pure Appl. Chem., 17 Y. Ikeda, T. Nakamura, H. Yorimitsu and K. Oshima, J. Am.
2006, 78, 441–449; (e) J. Iqbal, M. Mukhopadyay and
Chem. Soc., 2002, 124, 6514–6515.
A. K. Mandal, Synlett, 1997, 876–886; ( f ) H. Shinokubo 18 The formation of product 5 indicates that the formation of a
and K. Oshima, Eur. J. Org. Chem., 2004, 2081–2091.
14 N. V. El’chishcheva, Y. V. Shklyaev, Z. A. Vnutskikh,
T. F. Odegova, Y. S. Chekryshkin and S. S. Dubrovina,
Pharm. Chem. J., 2010, 44, 251–253.
benzyl organocobalt species might be a competitive path-
way. However, no evidence to support this path was
obtained.
19 Intermediate D is known to be unstable above ꢀ20 1C, see:
H.-F. Klein and H. H. Karsch, Chem. Ber., 1976, 109,
1453–1464.
15 (a) Y. Jiang, O. Blacque, T. Fox, C. M. Frech and H. Berke,
´
– Eur. J., 2010, 16, 2240–2249; (b) V. Corce,
Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2016
New J. Chem.