Law et al.
SCHEME 1
bis-allylated species. Similar mixtures of 2b and 3b were
obtained irrespective of the change in the reaction temperature,
the relative ratios of In to allyl bromide and imine, or the kind
of IL used (entries 2-7). In addition to [bmim][BF4], we have
used [emim][BF4] (emim = N-ethyl-N′-methylimidazolium),
[moemim][BF4] (moemim = N-(methoxyethyl)-N′-methylimi-
dazolium), and [bpy][BF4] (bpy ) N-butylpyridinium) as the
reaction media. While there were variations in the yield, the
product was invariably a mixture of 2b and 3b. In the case of
[bpy][BF4], we have also carried out the reaction with the
addition of ZnF2 or trimethylsilyl chloride ((TMS)Cl).17 The
reactions gave more or less the same mixture with similar yields
(entries 8 and 9).
In the allylation reaction in [bpy][BF4], we noticed that the
reaction was highly irreproducible, giving at times little or no
formation of the allylated amine 2 when different batches of
ionic liquids were used. We finally realized that the reaction
depended critically on the batch and purity of [bpy][BF4] as
we had previously observed for alkylzinc reactions in [bpy]-
[BF4].16c These observations prompted us to examine the ionic
liquid precursor [bpy][Br] as the additive for the reaction. The
results are summarized in Table 2. Indeed, the allylation reaction
of 1a gave no or low yield of homoallylic amine 2a in pure
[bpy][BF4] (entry 1) or with only about 4% of [bpy][Br] added
(entry 2). The yield of 2a was substantially increased when about
8% of [bpy][Br] was added and the reaction temperature was
increased (entries 4 and 5). With a lesser amount of imine 1a
relative to that of In and allyl bromide, the yield of 2a was
essentially quantitative with >8.5% bromide ion in the reaction
media (entries 5-7). An observable effect of the addition of
bromide ion (in the form of [bpy][Br]) was that the indium metal
powder did not form a cake together during the reaction when
about 8.5% of bromide ion was added.
such as sulfonimines,7 hydrazones,8 and oxime ethers9 have been
employed in aqueous indium-mediated allylation. In addition
to water,10 supercritical carbon dioxide,11 room-temperature ionic
liquids (ILs),12 and, of course, solvent-free conditions13 have
also been explored as reaction media to replace organic solvents.
Recently, it has been reported that imines 1 (Scheme 1) can
undergo allylation mediated by indium under solvent-free
conditions to give a mixture of the monoallylated species 2 and
the bis-allylated species 3.14 Indium-mediated allylation of
imines in ionic liquids has not been reported, but a similar
reaction in a nonvolatile solvent, poly(propylene glycol), has
been reported with a limited range of substrates.15 As part of
our program on the study of organometallic reactions in ionic
liquids,16 we report here our results on the indium-mediated
allylation of imines.
Results and Discussion
(1) Preliminary Studies. In the initial stage, we examined
the allylation reaction of benzaldehyde imines 1 in the ionic
liquid [bmim][BF4] (bmim = N-butyl-N′-methylimidazolium).
As we can see in Table 1, when the imine was derived from
aniline (1a, R ) Ph, entry 1), the homoallylic amine 2a was
obtained in moderate yield. However, the equivalent reaction
with imine derived from benzylamine (1b, R ) PhCH2, entry
2) gave a mixture of products 2 and 3, the mono- and
Addition of [bpy][Br] was also found to have an equally
remarkable effect on the allylation of 1b. No allylation occurred
in its absence (entry 8). When about 4.3% of bromide ion was
added, a mixture of products 2b and 3b resulted (entries 9 and
10). To our delight, in the presence of 8.5% bromide ion, the
allylation of 1b gave 2b only in excellent yield without the
formation of 3b (entries 11 and 12). The use of larger excess
of [bpy][Br] gave a similar result (entries 13 and 14). Finally,
if excess allyl bromide (2.5 equiv to In) was added to the
reaction mixture after the overnight stirring, the eventual product
was exclusively 3b (entry 16), suggesting that 3b was derived
from the allylation of 2b or more likely its precursor before
hydrolytic workup.
(5) (a) Paquette, L. A. In Green Chemistry, Frontiers in Chemically
Benign Synthesis and Processes; Anastas, P. J., Williamson, T. C., Eds.;
Oxford University Press: Oxford, U.K., New York, NY, 1998; p 261 and
footnote 49. (b) Vilaivan, T.; Winotapan, C.; Shinada, T.; Ohfune, Y.
Tetrahedron Lett. 2001, 42, 9073.
(6) Lalyanam, N.; Rao, G. V. Tetrahedron Lett. 1993, 34, 1647.
(7) (a) Chan, T. H.; Lu, W. Tetrahedron Lett. 1998, 39, 8605. (b) Lu,
W.; Chan, T. H. J. Org. Chem. 2000, 65, 8589. (c) Lu, W.; Chan, T. H. J.
Org. Chem. 2001, 66, 3467.
(8) (a) Kumar, H. M. S.; Anjaneyulu, S.; Reddy, E. J.; Yadav, J. S.
Tetrahedron Lett. 2000, 41, 9311. (b) Miyabe, H.; Ueda, M.; Nishimura,
A.; Naito, T. Org. Lett. 2002, 4, 131. (c) Miyabe, H.; Yamaoka, Y.; Naito,
T.; Takemoto, Y. J. Org. Chem. 2004, 69, 1415. (d) Cook, G. R.; Maity,
B. C.; Kargbo, R. Org. Lett. 2004, 6, 1741. (e) Cook, G. R.; Kargbo, R.;
Maity, B. Org. Lett. 2005, 7, 2767.
(9) (a) Miyabe, H.; Nishimura, A.; Ueda, M.; Naito, T. Chem. Commun.
(Cambridge) 2002, 1454. (b) Bernardi, L.; Cere, V.; Femoni, C.; Pollicino,
S.; Ricci, A. J. Org. Chem. 2003, 68, 3348.
(10) Li, C. J.; Chan, T. H. Organic Reactions in Aqueous Media; John
Wiley & Sons: New York, 1997.
(11) Jessop, P. Chemical Synthesis Using Supercritical Fluids; Wiley-
VCH: Weinheim, Germany, 1999.
(12) Rogers, R. D.; Seddon, K. R. Ionic Liquids: Industrial Applications
to Green Chemistry; Oxford University Press: Washington, DC, 2002.
(13) (a) Tanaka, K.; Toda, F. Chem. ReV. 2000, 100, 1025. (b) Varma,
R. S. Green Chem. 1999, 43.
(14) Andrews, P. C.; Peatt, A. C.; Raston, C. L. Tetrahedron Lett. 2004,
45, 243. A likely drawback associated with solvent-free reaction is that
many solid aldehydes did not give a satisfactory result as the mixture tended
to solidify readily and gave incomplete reaction.
(2) Mechanistic Investigations. (a) The Nature of the
Allylindium Intermediates in Ionic Liquids. We have also
studied the reaction mechanism of the indium-mediated ally-
lation reaction in ionic liquids. Our first task was to identify
the reactive allylindium intermediates formed. It is known from
our previous study of the reaction of indium metal with allyl
bromide in water that a transient reactive species, allylindium-
(I) (4), is formed.3b Compound 4 has a characteristic methylene
proton NMR signal at δ 1.7 ppm (d, J ) 8 Hz) in D2O (Scheme
2). In contrast, when indium metal reacts with allyl bromide in
1
an organic solvent such as dimethylformamide (DMF), the H
(15) Andrews, P. C.; Peatt, A. C.; Raston, C. L. Green Chem. 2004, 6,
119. At present, this method is applicable only to sulfonimines and aldimines
derived from aromatic anilines and aldehydes.
(16) (a) Law, M. C.; Wong, K. Y.; Chan, T. H. Green Chem. 2002, 4,
328. (b) Law, M. C.; Wong, K. Y.; Chan, T. H. Green Chem. 2004, 6, 241.
(c) Law, M. C.; Wong, K. Y.; Chan, T. H. J. Org. Chem. 2005, 70, 10434.
(d) Law, M. C.; Wong, K. Y.; Chan, T. H. Chem. Commun. (Cambridge)
2006, 2457.
NMR has two sets of allylic methylene signals at δ 1.75 (d, J
) 8 Hz) and δ 2.02 (d, J ) 8 Hz) in DMF-d7.2a,3a Previously,
the allylindium species in DMF was assigned to have the indium
sesquihalide structure 5 to account for the two sets of methylene
(17) Wang, D.; Dai, L.-N.; Hou, X.-L. Tetrahedron Lett. 1995, 36, 8649.
924 J. Org. Chem., Vol. 72, No. 3, 2007