Electrochemical Homocoupling of 2-Bromopyridines
SCHEME 2
SCHEME 3
preparing arylzincs as a good alternative to routes
involving Mg-Zn or Li-Zn metal exchange.10 The nickel-
zinc exchange is not favored, however, on the basis of
thermodynamics, contrary to the reverse process. Figure
1a, however, shows that this Ni-Zn exchange is efficient
in these reaction conditions. We have been able previ-
ously to show that, indeed, the Ni-Zn exchange does not
occur unless there is an excess (3 to 4 equiv) of the ligand
bipyridine versus nickel.11 The reaction investigated here
produces bipyridine, which therefore increasingly favors
the formation of the arylzinc intermediate as the reaction
proceeds. The excess bipyridine may bridge zinc and
nickel in a bimetallic intermediate, thus allowing the
random distribution of each metal in the vicinity of the
aryl group to give one of the two pyridilmetals.
Table 1 shows some other results obtained for the
electrochemical homocoupling of the 2-bromo-5-meth-
ylpyridine in the presence of a Zn sacrificial anode and
of different types and concentrations of supporting elec-
trolytes, while the current intensity and the charge are
similar.
We clearly see (Table 1, entries 1 and 2) that a high
concentration of an ammonium salt prevents the dimer-
ization and allows for an efficient formation of the
arylzinc intermediate based on the GC measurement of
the iodo derivative. Also, this is true whatever the nature
of the anion (Table 1, entries 3 and 4). On the contrary,
with NaI (Table 1, entry 5), even at high 0.2 M concen-
tration, the major product is the expected dimer, while
in this case the reduction product, 5-picoline, is also
formed in significant amount.
It thus comes out that NaI is more suitable as the
electrolyte to conduct the homocoupling of bipyridine
when a zinc anode is used. Indeed, with NaI, the shape
of the reaction profile looks different (Figure 1b) from the
one with TBAI (Figure 1a). Actually, the two first stages
look roughly similar and notably regarding the formation
of the arylzinc intermediate. This species is not, however,
so stable in the presence of NaI, and it can be converted
into the dimer if the current is still applied.
mopicoline was always high. To improve these homocou-
plings, we investigated several reaction parameters and
notably found that yields could be higher for all models
with iron instead of zinc as the anode6 (Scheme 2).
It should be mentioned that the use of a sacrificial
anode involves the production of salts in stoichiometric
amounts, and these salts may play a key role in the
overall process and in some cases even a synergistic effect
along with the catalyst, as already illustrated in previous
articles.7 There is also a parameter that is often concealed
(i.e., the nature of the supporting electrolyte).
In this article we present a reinvestigation of the
process to understand why the reaction cannot be ef-
ficient with a zinc anode and, notably, if there is an effect
of the nature of the supporting electrolyte. Nickel and
zinc are indeed classically found in couplings of aromat-
ics. Kumada8 and Iyoda9 have thus reported an efficient
homocoupling process of aryl halides, including pyridyl
halides,9 using [NiII(PPh3)]X2 (X ) Cl,8 Br9), along with
Zn as reducing agent, DMF8 or THF9 as solvent, and in
the presence of iodide ions (KI8 or Et4NI9). They reported
the yield increase of substituted biaryles in the presence
of iodide ion as compared to its absence and explained
the role of iodide as a bridging ligand between nickel
complexes and zinc ions.
Results and Discussion
The dimerization of 2-bromopyridine was studied as
the model reaction. A typical course of the reaction is
shown in Figure 1a, TBAI being used as the supporting
electrolyte.
There are clearly three stages in this process. During
the first 20 min, the reaction goes slowly for both the
consumption of the starting halopyridine and the forma-
tion of bipyridine. Next, the consumption of the starting
reagent follows the coulometry, while the formation of
bipyridine remains slow. Actually, another product, io-
dopyridine, can be detected by GC as the major compo-
nent if the crude reaction mixture is reacted with iodine
and which likely forms by metal-iodine exchange (Scheme
3). Finally, even after holding the current over a long
time, the yield in bipyridine remains at around 40%, thus
indicating the stability of the reaction intermediate, very
likely pyridylzinc halide.
The consumption of the organozinc intermediate by
transmetalation (Scheme 4) is now thermodynamically
favored to regenerate the organonickel intermediate.
The experiments described in Figure 2 show another
interesting feature of this process. Theoretically, the
electric current is only required for the full consumption
of the starting halopyridine, and it appears that it is
indeed consumed in about 60 min, which is not far from
the theory (40 min at I ) 0.1 A). If we then set off the
current, the consumption of the organozinc continues
because of the reversibility of the transmetalation pro-
cess,12,13 and the formation of the homocoupling product
(2,2′-bipyridine) occurs. However, the process is very
slow, and the arylzinc is only fully consumed after nearly
The formation of arylzinc intermediates has already
been observed in related reaction conditions involving the
use of Zn sacrificial anode. More interestingly, this
transformation has been optimized into a method for
(6) De Franc¸a, K. W. R.; Navarro, M.; Leonel, E.; Durandetti, M.;
Ne´dele´c, J.-Y. J. Org. Chem. 2002, 67, 1838.
(7) (a) Gosmini, C.; Rollin, Y.; Ne´de´lec, J.-Y.; Pe´richon, J. J. Org.
Chem. 2000, 65, 6024. (b) Conan, A.; Sibille, S.; Pe´richon, J. J. Org.
Chem. 1991, 56, 2018. (c) Mellah, M.; Labbe´, E.; Ne´de´lec, J.-Y.;
Pe´richon, J. New J. Chem. 2001, 25, 318.
(8) Zembayashi, M.; Tamao, K.; Yoshida, J.; Kumada, M. Tetrahe-
dron Lett. 1977, 4089.
(9) Iyoda, M.; Otsuka, H.; Sato, K.; Nisato, N.; Oda, M. Bull. Chem.
Soc. Jpn. 1990, 63, 80.
(10) Gosmini, C.; Lasry, S.; Ne´de´lec, J.-Y.; Pe´richon, J. Tetrahedron
1998, 54, 1289.
(11) Devaud, M.; Troupel, M.; Pe´richon, J. (LECSO), 1995. Unpub-
lished results.
J. Org. Chem, Vol. 70, No. 26, 2005 10779