Pd-Catalyzed Intermolecular Amidation of Aryl Halides
A R T I C L E S
source and Xantphos as the ligand, the amidation of unactivated
aryl halides could be efficiently carried out (entries 8-15).
Electronically neutral or slightly electron-rich aryl halides or
triflates reacted efficiently with both aromatic and aliphatic
primary amides, N-methylformamide (entries 12), and lactams
(entries 12-15) at 100 °C. It is worth mentioning that couplings
between bromobenzene and four- to seven-membered-ring
lactams could generally be realized with as little as 1 mol % of
Pd at 100 °C. Notably, under these basic conditions, none of
the product resulting from ring opening was observed.
When less reactive aryl halides were used, small amounts
(2-8%) of N-phenylated amides were detected as the byproducts
in crude reaction mixtures (GC and GC-MS analysis); removal
of these side products was straightforward with flash chroma-
tography. Increasing the Xantphos:Pd ratio usually resulted in
increased amounts of the N-phenylamide. These byproducts are
likely formed via exchange between the aryl group bound to
Pd(II) and the phenyl group of the phosphine ligand;15 Hartwig
and co-workers have observed similar aryl group exchange
processes in Pd-catalyzed aminations.15a
During the course of these studies, we found an unusual
dependence of catalyst loading on the reaction efficiency. For
example, using 5 mol % of Pd (as Pd2dba3)/7.5 mol % of
Xantphos in the coupling of 4-tert-butylbromobenzene and
benzamide, a product:ArBr ratio of 1:5 was obtained as
compared to 1.1:1 with 2 mol % of Pd, implicating the
possibility of an unknown catalyst decomposition pathway
occurring at higher catalyst concentrations. Notably, higher
amounts of N-phenylated side products were observed at higher
catalyst loadings as well (∼1 M in ArBr).
methodology to more challenging substrate combinations (Table
2). Thus, with 4 mol % of Pd and a concentration of 0.5 M,
2-bromotoluene reacted with benzamide in 98% yield (entry
1). The lactam 2-pyrrolidinone was also coupled with sterically
hindered 2-bromo-p-xylene at 0.25 M (88% yield, entry 2).
Electron-rich aryl halides such as 2-bromoanisole and 2-iodo-
anisole both underwent amidations in good yields (entries 3 and
4). Previously extremely inactive amides such as N-methyl-
acetamide (an acyclic secondary amide), as well as a primary
or secondary sulfonamide, also reacted with unactivated aryl
bromides under these more controlled conditions (entries 5-7).
It is interesting to note that although dioxane has been the
solvent of choice for most amidations, toluene proved to be the
preferred solvent for the coupling of the secondary sulfonamide
N-methyl p-tolylsulfonamide (entry 6).
We also examined the C-N bond-forming reactions using
other amide analogues such as cyclic carbamates7-9 and ureas.12
With K3PO4 as base, 2-oxazolidone reacted with 4-chloro-
bromobenzene in 87% yield (entry 8). Both five- and six-
membered cyclic ureas were doubly arylated with 3-bromo-
anisole with as little as 1 mol % of the palladium catalyst (entries
9 and 10). In the latter two cases, increasing the loading of the
palladium catalyst and Xantphos ligand resulted in an increase
of the aryl group exchange byproducts.
The Pd-catalyzed amidations of aryl halides described above
have shown very good substrate scope and functional group
compatibility. Aryl halides bearing electron-withdrawing groups
at ortho, meta, or para positions underwent C-N bond formation
reactions generally with primary and secondary amides, car-
bamate, and sulfonamides. Unactivated or deactivated aryl
halides also coupled with various amides under more carefully
controlled conditions. However, a few classes of substrate
combinations still remain challenging for this Pd-catalyst
system. Amidations of aryl halides bearing a ketone functional
group suffered from a competitive ketone arylation processes.
Additionally, amidation reactions involving aryl halides that
possess a para-electron-donating group or acyclic secondary
amides were usually sluggish and yielded large amounts (>20%)
of aryl group exchange products (N-phenyl amides).
The fact that only Xantphos worked efficiently for amidation
reactions prompted us to study the mechanism of the reaction
in greater detail. To this end, we first tried to isolate the oxidative
addition product from 4-bromobenzonitrile, Pd2(dba)3, and
Xantphos. (Bisphoshine)Pd(Ar)X complexes have been prepared
from their monophosphine complex dimers [(P(o-Tol)3)Pd-
(Ar)X]216 (I) or [(PPh3)Pd(Ar)X]2.17 Following a procedure used
to prepare (BINAP)Pd(4-cyanophenyl)(Br) in our laboratories,16
complex I underwent ligand exchange when treated with
Xantphos in methylene chloride at room temperature to afford
the corresponding Xantphos complex, (Xantphos)Pd(4-cyano-
phenyl)(Br) (II), which is air-stable and can be stored in a vial
on the bench for over a year without noticeable decomposition,
as judged by NMR (Scheme 1). We subsequently found that
complex II could be prepared in 80% yield by simply stirring
4-bromobenzonitrile, Pd2(dba)3, and Xantphos in benzene at
room temperature for 22 h. To our surprise, complex II showed
only one singlet in 31P NMR at +9.3 ppm, compared to a pair
At this point in our investigations, the amidations of activated
aryl halides were fairly general, but the scope of the amidation
involving unactivated aryl halides was still limited. For example,
acyclic secondary amides as well as sulfonamides were un-
reactive, and sterically hindered or electron-rich unactivated aryl
halides were not viable substrates under the optimized conditions
shown above. Additionally, aryl group exchange became more
problematic in these challenging couplings, and increasing
amounts of the phenylated byproduct was observed. Generally,
higher catalyst loadings may be used to accelerate a reaction;
however, for these amidation processes, additional catalyst often
led to an increase in aryl group exchange product and faster
decomposition of the catalyst. Unfortunately, replacement of
the phenyl groups on Xantphos with o-tolyl groups to suppress
aryl group exchange yielded an inefficient amidation catalyst.
A detailed examination of the reaction parameters led to the
discovery that careful control of catalyst loading and the reaction
concentration was key to the successful amidation of less
reactive substrates.
A substrate concentration of 1 M was suitable for most cases
reported in Table 1; however, lower concentrations (0.125 to
0.5 M) and higher palladium catalyst loadings were required
for these more challenging reactions to proceed efficiently.
Lower concentrations may decrease the rate of catalyst decom-
position, particularly when more catalyst was used.
With a better understanding of the reaction variables important
to the success of these couplings, we sought to extend the
(16) (a) Widenhoefer, R. A.; Zhong, H. A.; Buchwald, S. L. J. Am. Chem. Soc.
1997, 119, 6787-6795. (b) Widenhoefer, R. A.; Buchwald, S. L. J. Am.
Chem. Soc. 1998, 120, 6504-6511.
(15) (a) Hamann, B. C.; Hartwig, J. F. J. Am. Chem. Soc. 1998, 120, 3694-
3703. For a mechanistic study, see: (b) Goodson, F. E.; Wallow, T. I.;
Novak, B. M. J. Am.Chem. Soc. 1997, 119, 12441-12453.
(17) Driver, M. S.; Hartwig, J. F. J. Am. Chem. Soc. 1997, 119, 8232-8245.
9
J. AM. CHEM. SOC. VOL. 124, NO. 21, 2002 6045