I. Conesa Lerma et al. / Journal of Organometallic Chemistry 690 (2005) 5841–5848
5843
ture protocols for amination of aryl halides. Whilst a
number of groups have highlighted the use of palla-
dium–NHC systems for amination of aryl chlorides
+Ar
+Ar
Ar
N
Ar
N
Pri
Pri
N
N
Cl-
Cl-
Pd
Ar =
N
Ar
N
Ar
N
Ar
N
Ar
[
6], we have initiated investigations on the development
of a practical protocol based on palladium/imidazo-
lium salts for amination of aryl halides. We have spe-
cifically focused on bromides and iodides in the first
phases of this study. This is, in part because of the
lack of general studies on these apparently simpler
classes of substrate. We believe that the development
of an imidazolium salt based protocol for bromides
and iodides will be a valuable addition to the arsenal
of methods currently available. In this work we de-
scribe our initial studies on this class of transformation
and report on a new and practical protocol. The key
finding has been the use of lithium hexamethyldisilaz-
ide (LHMDS) as a base to aid reproducibility in these
reactions. This base has been used for palladium/phos-
phine-catalysed reactions by Urgaonkar and Verkade
3
4
5
Additionally, we focused on the nature of the base
employed, as a variety have been used previously in pal-
ladium-catalysed aminations and the nature of the base
can have a profound impact on reactivity. Initial exper-
iments using potassium tert-butoxide usually required
elevated temperatures, but a brief assessment of a vari-
ety of bases indicated that LHMDS was preferred [18],
with other inorganic bases (cesium carbonate, sodium
tetraborate and potassium phosphate) and organic bases
(2,6-lutidine and DBU) being less effective. Commer-
cially available LHMDS was used as a 1 M solution in
THF, as this gave faster reactions than the alternative
1 M solution in toluene.
[
[
18], Louie and Hartwig [19] and Buchwald et al.
20]. LHMDS has also been employed as an ammonia
The use of this base saw improved yields and short-
ened reaction times. In the case of aryl iodides, reactions
that only worked at 100 °C with KO Bu, now worked at
equivalent in the palladium/phosphine-catalysed reac-
tion with aryl halides to form primary anilines
t
[
21,22]. However, to our knowledge, LHMDS has
room temperature. Trials with other hexamethyldisilaz-
ides (i.e. NaHMDS and KHMDS) showed that they
could also be used as the base for the amination, but
that LHMDS gave the best results. The results of the
optimised amination protocols for bromides and iodides
are presented in Tables 1 and 2.
not been used as the base for the Pd–NHC mediated
amination of aryl halides.
2
. Results and discussion
As expected, our best results were obtained in reac-
tions with cyclic secondary amines, most notably piper-
idine and morpholine. Good yields were obtained for
N-methylaniline and secondary amines containing aro-
matic rings (e.g. N-methylbenzylamine) or short alkyl
chains (e.g diethylamine). Acyclic secondary amines
(e.g. di-n-butylamine) did work at room temperature,
but were somewhat disappointing by comparison to
their cyclic counterparts. Slower reactions could how-
ever be accelerated with higher loadings of palladium
and imidazolium ligand.
Elevated temperatures were required for the reactions
of primary amines with aryl bromides. Whilst these reac-
tions were usually complete within 24 h, they were
generally low-yielding. For example, in the case of
n-hexylamine (Scheme 4), a bisarylated species was the
major product observed. Here, an excess of amine
(3.6 eq.) was used to hinder this reaction, which was fa-
voured due to the secondary aniline product being more
reactive than the primary alkyl amine. This also required
an increase in the quantity of catalyst used (2 mol%
We evaluated a variety of aryl bromides and iodides in
aminations catalysed by palladium/imidazolium salts.
Dimeric tris(dibenzylideneacetone)dipalladium(0), Pd2-
(
dba) , was used as the Pd(0) source in this reaction, as
3
the monomeric bis(dibenzylideneacetone)palladium(0)
was found to be ineffective. The saturated ligand, 1,3-
bis(2,6-diisopropylphenyl)-imidazolinium chloride (Si-
Pr Æ Cl) 3 was used for the reactions with aryl bromides,
as previous research has shown it to be the most suitable
for this reaction [6]. However, preliminary results indi-
cated that the unsaturated 1,3-bis(2,6-diisopropylphe-
nyl)-imidazolium chloride (IPr Æ Cl) 4 was slightly
better for reactions of aryl iodides. The imidazolium
chloride ligands were prepared using literature proce-
dures [23]. Despite previous findings of this group and
others, a 1:1 Pd/ligand ratio did not appear to be the
optimum ratio in this reaction. Instead, for reasons that
are not yet understood, a 1:2 Pd/ligand ratio was
preferred, with modest improvements in conversion.
The use of the pre-formed two-coordinate Pd–NHC
complex of bis(1,3-bis(2,6-diisopropylphenyl)imidazo-
Pd
(dba)
and 8% SIPr Æ Cl).
3
2
lin-2-ylidene) palladium(0) (Pd(IPr) ) 5 was also investi-
It would appear from the data that the reactions of
2
gated for the amination of aryl iodides, but was found to
be slightly less effective than the corresponding in situ
palladium/imidazolium chloride protocol, consistent
with results we have observed elsewhere [15].
aryl iodides are more facile and higher-yielding than
the reactions of aryl bromides. For example, iodobenzene
reacted with adamantylamine at room temperature in a
yield of 70%; conversely, bromobenzene failed to react