LETTER
445
Practical Protocol for the Palladium-Catalyzed Synthesis of Arylphosphonates
from Bromoarenes and Diethyl Phosphite
1
P
L
alladium-Catal
u
yze
d
Synthe
k
sis of
A
rylph
a
osphonate ss J. Gooßen,* Mohammad K. Dezfuli
Max-Planck Institut für Kohlenforschung, Kaiser-Wilhelm-Platz 1, 45470 Mülheim an der Ruhr, Germany
Fax +49(241)8093663; E-mail: goossen@oc.rwth-aachen.de
Received 23 November 2004
Since, in our opinion, these protocols represented the most
convenient synthetic entry, we attempted to apply them to
the preparation of various functionalized aryl phospho-
Abstract: A greatly improved, reliable protocol for the palladium-
catalyzed cross-coupling of dialkyl phosphites with aryl bromides
has been developed. The use of an alcoholic solvent was the key to
high yields in the synthesis of a broad variety of arylphosphonates, nates as starting materials for an ongoing synthetic
with Pd(OAc) /PPh3 as the catalyst and a sterically demanding project. Unfortunately, we were confronted with strongly
2
tertiary amine as the base.
varying and often unsatisfactory yields. In order to eluci-
Key words: palladium, catalysis, arylphosphonates, cross-cou- date the reason for this and potentially come to a more
pling, aryl halides
reliable reaction protocol, we chose the reaction of 4-bro-
moanisole (1a) with diethyl phosphite (2) as a model sys-
tem to screen various catalysts and reaction conditions.
Selected results are summarized in Table 1.
Aryl- and vinylphosphonic acid moieties are key func-
2
tionalities in many biologically active compounds, flame
In our hands, this model reaction did not go to completion
3
4
retardants, or polymer additives. Traditional synthetic
entries into this important substrate class include Friedel–
Crafts reactions of arenes with phosphoric acid deriva-
10
when applying literature conditions (entries 1 and 2).
Further studies revealed that among all reaction para-
meters studied, the choice of the solvent had the strongest
influence on the reaction outcome: The addition of polar
solvents proved to be beneficial, and alcoholic solvents
such as ethanol led to excellent yields even at lower tem-
peratures (entries 2–6). Further experiments revealed that
even small amounts of ethanol, which may arise from
partial hydrolysis of the starting material diethyl phos-
phite, strongly facilitate the reaction. This may explain
why our yields strongly depended on the quality of the
starting materials and the water content of the solvents
when following the literature protocols.
5
tives, Cu-catalyzed reactions of diazonium salts with
6
PCl , nucleophilic substitution reactions of activated aryl
3
7
halides with sodium dialkylphosphites, Ni- or Cu-medi-
ated couplings between aryl halides and trialkyl phosphi-
8
tes, and reaction of arylmetal derivatives and trialkyl
9
phosphites. However, due to the aggressive reagents or
harsh conditions required, these transformations tend to
be incompatible with sensitive functionalities, impeding
applications in combinatorial chemistry and in the synthe-
sis of complex, functionalized molecules.
A potentially more general access to dialkyl arylphospho-
nates based on the palladium-catalyzed reaction of aryl
halides with dialkyl phosphite has been disclosed by
We then examined several Pd-sources and found
Pd(OAc) to be the most effective precatalyst (entries 6–
2
9
). Using ethanol as the solvent, inorganic bases did not
1
0,11
Hirao et al. (Scheme 1).
Pd(PPh ) as the catalyst and triethylamine as the base, the
In the initial protocol using
give satisfactory yields as they facilitated side reactions
such as dehalogenation of the bromoanisole (entries 10–
3
4
best yields were obtained in the absence of a solvent.
Beletskaya et al. reported that the reaction can also be per-
formed under phase-transfer conditions using potassium
1
2). Tertiary amines with a low nucleophilicity led to
much higher selectivities (entries 9, 13–16), and the best
results were obtained using sterically crowded derivatives
such as Hünig’s base or dicyclohexylmethylamine.
1
2
carbonate as the base.
The choice of the phosphine was also important (entries
1
7–22). Interestingly, neither chelating nor electron-rich
alkyl phosphines gave satisfactory results in this transfor-
mation, although these ligands had proved to be highly ef-
fective in other Pd-catalyzed couplings. Instead, the best
results were obtained with simple triaryl phosphines such
as P(p-MeO-Ph) and P(p-Cl-Ph) . However, since the in-
Scheme 1 Pd-catalyzed synthesis of arylphosphonates
3
3
expensive triphenylphosphine was found to be almost as
effective, we considered it to be the best choice.
After having identified convenient and reliable reaction
conditions, we investigated the scope of the new reaction
protocol using a variety of aryl bromides (Scheme 2). Se-
SYNLETT 2005, No. 3, pp 0445–0448
Advanced online publication: 04.02.2005
DOI: 10.1055/s-2005-862372; Art ID: G45204ST
1
6.
0
2.
2
0
0
5
13
lected results are summarized in Table 2. Gratifyingly, it
©
Georg Thieme Verlag Stuttgart · New York