Angewandte
Chemie
N-Oxide Arylation
DOI: 10.1002/anie.200602773
Palladium-Catalyzed Cross-Coupling Reactions of
Diazine N-Oxides with Aryl Chlorides, Bromides,
and Iodides**
Jean-Philippe Leclerc and Keith Fagnou*
The use of palladium-catalyzed cross-coupling reactions in
biaryl synthesis[1] is linked to, and limited by, the synthetic and
commercial availability of organometallic reagents such as
aryl boronic acids. Not only are most of these compounds
expensive, important classes of aryl organometallic are very
challenging prepare and/or use in cross-coupling reactions,
including electron-deficient nitrogen heterocycles. The impor-
tance of these building blocks in medicinal chemistry and
materials sciences[2] has prompted continued methodological
efforts, and two very recent reports highlight the importance
of this goal.[3,4] Absent from these and the predominance of
reports are some of the most problematic organometallic
reagents—azines bearing the metal adjacent to a nitrogen
atom. The problem is even more severe with diazines
(Scheme 1). These organometallic compounds are unstable,
can rarely be isolated, and commonly decompose under cross-
coupling reaction conditions. While some are commercially
available, the price reflects both their value and the challenge
associated with their preparation.[5]
Recently, the potential of direct arylation as an efficient
alternative to standard cross-couplings is becoming recog-
nized,[6] and we have been studying the use of N-oxides as
replacements for unstable/unreactive organometallics.[7] In
the context of this strategy, diazine N-oxides are more
challenging than simple pyridine N-oxides since they possess
a free nitrogen atom that could bind and poison the catalyst.[8]
They are also more p-electron-deficient and less nucleophilic
than pyridine N-oxides.
Scheme 1. Direct arylation in aryl diazine synthesis. Pyrazine, pyrimi-
dine, and pyridazine boronic acids are unstable, difficult to prepare,
and unsuited for cross-coupling reactions. Their N-oxides are readily
prepared, bench-stable replacements for the organometallic reagent in
biaryl synthesis.
and we demonstrate that the N-oxide functionality can be
removed easily after cross-coupling or transformed into a
wide range of other functional groups. These new reactions
can be performed with aryl iodides, bromides, and chlorides
and include the first examples of N-oxide arylation with
equimolar ratios of the two coupling partners. Furthermore,
the relative reactivities and regioselectivities point to C–H
acidity as a critical factor in reactivity, encouraging consid-
eration of this property in the design of other novel direct
arylation processes.
Initial reaction screens with N-oxides 1–3 under previ-
ously described conditions led to disappointing results. Noting
that the N-oxides were only sparingly soluble in toluene, we
reinvestigated the reaction conditions. We found that chang-
ing the solvent from toluene to dioxane provides superior
conversions with N-oxides 1 and 3, giving the cross-coupled
products 4 and 5 in yields of 75% and 72%, respectively
(Table 1, entries 1 and 2). These two substrates are actually
more reactive than pyridine N-oxide, as demonstrated by a
competition experiment between 1 and pyridine N-oxide,
which resulted in exclusive arylation of 1 (Table 1, entry 4). In
contrast to the excellent results obtained with 1 and 3, the
reaction of pyrimidine N-oxide (2) proceeds in very low yield
(Table 1, entry 3).
Herein we describe the establishment of reaction con-
ditions that enable the use of readily available,[9] bench-
stable[10] diazine N-oxides as cheap, high-yielding reagents in
palladium-catalyzed cross-coupling reactions (Scheme 1). To
overcome catalyst poisoning associated with some N-oxide
substrates, a benefical effect of copper(I) salts was uncovered,
Further investigations revealed that the poor outcomes
associated with 2 do not result from low reactivity alone. For
example, the addition of pyrimidine N-oxide (2) to a reaction
with pyrazine N-oxide (1) results in the exclusive formation of
4 but in a significantly lower yield (9%) than when the
reaction was performed in the absence of 2 (75% yield;
Table 1, entries 5 and 1). The reason for catalyst inhibition
with 2 and not with 1 or 3 is a focus of ongoing study.[11,12] We
note that neither pyridine N-oxide nor pyridine poison the
reaction of 1 (Table 1, entries 4 and 6), indicating that these
deleterious effects are specific to pyrimidine N-oxide.[12]
To overcome catalyst inhibition, a variety of additives
were investigated including phosphines, halides, and metals.
[*] J.-P. Leclerc, Prof. Dr. K. Fagnou
Center for Catalysis Research and Innovation
Department of Chemistry, University of Ottawa
10 Marie Curie, Ottawa, ON K1N6N5 (Canada)
Fax: (+1)613-562-5170
E-mail: keith.fagnou@science.uottawa.ca
[**] We thank NSERC, CFI, OIT the University of Ottawa, the Ontario
government (PREA, K.F.), and the Research Corporation (Cottrell
Scholar Award, K.F.) for support of this work. Additional research
support was provided by Boehringer Ingelheim (Laval) and Merck
Frosst Canada. We thank Dr. Greg Hughes (Merck Frosst) for
obtaining and providing differential scanning calorimetry data.
Supporting information for this article is available on the WWW
Angew. Chem. Int. Ed. 2006, 45, 7781 –7786
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7781