Granular indium Barbier allylation of carbonyl compounds: A more economical protocol

Article abstract of DOI:10.1055/s-2006-951529

A new protocol, amenable to be used in large-scale preparations, using a cheaper form of indium metal and mild warming is reported for the Barbier allylation of aldehydes and ketones with allyl bromide in N,N-dimethylformamide. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2006-951529

LETTER
3337
Granular Indium Barbier Allylation of Carbonyl Compounds:
A More Economical Protocol
G
ranular Indium Barbi
a
er Allylatio
r
n
of Ca
r
c
bonyl Com
e
pounds lo D. Preite,* Andrés Pérez-Carvajal
Departamento de Química Orgánica, Facultad de Química, Pontificia Universidad Católica de Chile, Vicuña Mackenna 4860, Casilla 306,
Correo-22, 6904411 Santiago, Chile
Fax +56(2)3544744; E-mail: mpreite@uc.cl
Received 12 July 2006
We wish to dedicate this paper to Prof. Jaime Valderrama on the occasion of his retirement.
several different conditions, we learned that indium metal
Abstract: A new protocol, amenable to be used in large-scale prep-
in granular form works exactly as well as the powder, in
arations, using a cheaper form of indium metal and mild warming is
reported for the Barbier allylation of aldehydes and ketones with al-
terms both of yield and product purity (Table 1).5 The
carbonyl compounds tested were both aldehydes (Table 1,
entries 1–10; Figure 1) and ketones (Table 1, entries 11–
20; Figure 1), either aliphatic and aromatic, including
several bifunctional compounds and two dicarbonyl com-
pounds (entries 9 and 13). In all cases good to excellent
isolated yields were obtained, ranging from 42% to 100%,
as long as a small excess of indium was used. In a couple
of cases (entries 3 and 4) a rather low yield was observed,
which we attributed mainly to decomposition and prob-
lems during the workup and chromatographic purifica-
tion. In one case (entry 9) dialdehyde 1i gave a cyclic
hemiacetal instead of the usual homoallylic alcohol, prod-
uct of intramolecular attack of the homoallylic alcohol ox-
ygen on the remaining carbonyl group, and also a rather
low (56%) yield was obtained in that case.
lyl bromide in N,N-dimethylformamide.
Key words: indium, carbonyl complexes, allylation, homoallylic
alcohols, addition reactions
The indium-mediated allylation of aldehydes and ketones
is a well established method for C–C bond formation and
its chemistry has recently been extensively reviewed.1 In-
dium chemistry has received considerable attention from
the organic chemists, especially after the discovery that
the indium-mediated Barbier addition reactions of allylic
bromides (or iodides) on aldehydes and ketones can be
performed in aqueous media (Scheme 1).2 Despite the rel-
ative abundance of references on the subject almost all ex-
perimental data involve simple unfunctionalized carbonyl
compounds containing only one reactive aldehyde or ke-
tone, and perhaps an aromatic ring. For our research on
the general applicability of a new reaction of lead
tetraacetate3 we needed an easy access to functionalized
homoallylic alcohols and, after trying several conditions
and reagents, we decided to try several indium allylations
of aldehydes and ketones containing diverse chemical
functions.
At room temperature the Barbier allylations using granu-
lar indium turned out to be sluggish, taking several days
to complete. Fortunately, we found that a slight warming
to 40–50 °C, together with vigorous magnetic stirring sig-
nificantly shortened our reaction times to less than two
hours, giving better yields and purer products. The use of
indium in small pieces had an extra bonus, as it allowed us
to observe obvious color changes on the metal surface that
would have gone unnoticed when using the powdered
metal.
O
HO
R1
In
R2
We observed that for the reaction to start the originally
clean and shiny metal surface had to lose its sheen first
and it acquired a grayness, which sometimes turned to a
deep-purple or dark-brown color. Simultaneously, the ini-
tially clear or slightly colored N,N-dimethylformamide
solution became somewhat cloudy and moss-green in col-
or.6 Both changes occurred between five minutes and 30
minutes at 40–50 °C, and sometimes even earlier. Next,
the reaction mixture remained in the same state until the
limiting reactant, whichever it was, got exhausted, and the
metal surface turned shiny and clean again, as it had been
initially.
R1
R2
Br
2
1
Scheme 1 Indium-mediated allylation of aldehydes and ketones
Indium powder, the customary form of metal used in such
reaction, was too expensive for our preparative needs.
Nevertheless, we found that other forms of commercial
indium metal were priced up to ten times less than the
powder. We used the most affordable commercial form
that we could locate: metal shot, 99.9% pure.4 The solvent
was a small amount (1 mL/mmol) of technical grade N,N-
dimethylformamide, used as purchased. After trying
If the green reaction mixture was centrifuged, a reddish-
brown solution and a gray solid were obtained. The solu-
tion contained no identifiable organic materials but the in-
soluble solid on treatment with water followed by solvent
extraction gave the product, homoallylic alcohol 2. When
SYNLETT 2006, No. 19, pp 3337–3339
0
1
.1
2
.2
0
0
6
Advanced online publication: 23.11.2006
DOI: 10.1055/s-2006-951529; Art ID: S15206ST
© Georg Thieme Verlag Stuttgart · New York

3338
M. D. Preite, A. Pérez-Carvajal
LETTER
left overnight after the reaction completion, the moss- one or various radical species involved, as previously re-
green mixture also turned reddish-brown and sometimes a ported.9–11 These observations imply that in the absence of
heavy and dark-gray precipitate was observed. Once wa- a proton source the allylation is heterogeneous, giving a
ter or any other proton source was added to quench the re- highly insoluble indium alcoholate as the final product
action and extract the organics, the mixture turned to a that, upon protonation, finally yields alcohol 2.
clear pale-yellow solution.7
Table 1 Indium-Mediated Reaction of Allyl Bromide with Several
Carbonyl Compounds 1
CHO
CHO
CHO
Entry
1
Substrate
1a
Product
2a
2b
2c
Yield (%)a
75
R1
R2
1a
1b: R1 = R2 = H
1c: R1 = H; R2 = Cl
1d: R1 = H; R2 = Br
1e
2
1b
1c
100
57
3
OMe
O
CHO
4
1d
1e
2d
2e
42
CHO
CHO
R3
5
87
MeO
OMe
6
1f
2f
79
H
1f
1g
1h: R3 = Me
7
1g
2g
2h
2i
90
1i: R3 = CHO
8
1h
1i
97
CHO
O
O
R4
9
56b
86
10
11
12
13
14
15
16
17
18
19
20
1j
2j
O
1j
1k: R4 = H
1l: R4 = Me
1m
1k
1l
2k
2l
91
O
O
O
89
89b
88
R5
1m
1n
1o
2m
2n
2o
2p
2q
2r
O
1p
MeO
1n: R5 = Me
1o: R5 = Ph
1q
94
OH
O
O
O
1p
1q
1r
94
73
O
1r
86
1s
1t
1s
2s
78
Figure 1 Carbonyl compounds tested
1t
2t
100
We tried to correlate the disappearance of signals belong- a All yields correspond to products isolated after column chromatog-
raphy.
ing to the starting materials to the appearance of colored
b Corresponds to attack on a single carbonyl group.
products by performing a series of 1H NMR experiments.
Unfortunately, we could not observe the advent of any
new signals corresponding to a reaction intermediate or
product, but instead we observed that the moss-green col-
or intensified as the signals corresponding to the starting
allyl bromide and carbonyl compound disappeared. No
precipitate was evident, but rather an apparent increase in
the sample viscosity (and the consequent loss in field ho-
mogeneity) occurred. When an aliquot of that reaction
mixture was allowed to react with water the green color
disappeared and the signals corresponding to alcohol 2
were detected in the spectra. As indium is a main group
metal, the green color should arise due to electronic tran-
sitions within ligands orbitals.8 So we reasoned that the
metal Lewis acid could act as an anchor allowing a ligand
coupling of the carbonyl and allyl ligands to occur, with
We also reacted allyl bromide and indium metal in N,N-
dimethylformamide (Grignard conditions) and in this case
no green color was observed. The appearance of two sets
of signals in the 1H NMR spectrum, at d = 2.02 ppm and
d = 1.75 ppm, was assigned to the methylene group direct-
ly bonded to indium metal, in obvious agreement with
those previously reported for the N,N-dimethylformamide
and tetrahydrofuran solutions of allylindium organome-
tallics.12,13 Previously,12 both signals were assigned to a
triallyldiindium trihalide (In2Br3R3), formed by reaction
between indium and allyl bromide, on the basis of their
relative intensities. However, we believe that they belong
to two different organometallic compounds as signal at
d = 2.02 ppm quenches immediately on addition of water
Synlett 2006, No. 19, 3337–3339 © Thieme Stuttgart · New York

LETTER
Granular Indium Barbier Allylation of Carbonyl Compounds
3339
to leave the one at d = 1.75 ppm alone, which disappears
at a much slower rate, and a N,N-dimethylformamide
solution left standing overnight shows a reduction in the
intensity of the signal at d = 1.75 ppm and an enhancement
of the signal at d = 2.02 ppm.13,14 Chan et al.,13 previously
postulated that the signal at d = 1.75 ppm belongs to an
allylindium(I) species. We believe that the other signal at
d = 2.02 ppm, corresponds to another, less reactive
organometallic species, in slow dynamic equilibrium with
the former (Grignard conditions). Under Barbier condi-
tions like those used in the present syntheses, the actual
concentration of the active organoindium species tends to
be zero at all times, as it would react with the carbonyl
species immediately after its formation, which explains
the complete absence of signals at d = 2.02 ppm and d =
1.75 ppm in the NMR spectra.14–16
References and Notes
(1) (a) Nair, V.; Ros, S.; Jayan, C. N.; Pillai, D. S. Tetrahedron
2004, 60, 1959. (b) Poddlech, J.; Maier, T. C. Synthesis
2003, 633. (c) Cintas, P. Synlett 1995, 1087.
(2) Lin, C. J. Tetrahedron 1996, 52, 5643.
(3) Preite, M. D.; Cuellar, M. A. Chem. Commun. 2004, 1970.
(4) Purchased from Alfa Aesar Chemical Co. A somewhat more
expensive indium wire, 99.99% pure, purchased from
Indium Corporation of America, also gave identical results.
(5) Indium is so soft that an ingot of metal can be easily flattened
by pressing with a spatula (or even strong fingernails)
between two pieces of clean filter paper, and then cut into
small pieces of a few square millimeters with scissors.
(6) It was previously reported that when a THF suspension of In
powder and allyl bromide was stirred for 90 min, and then
allowed to react with a ketone, a yellow-green color
developed: Capps, S. M.; Clarke, T. P.; Charmant, J. P. H.;
Höppe, H. A. F.; Lloyd-Jones, G. C.; Murray, M.; Peakman,
T. M.; Stentiford, R. A.; Walsh, K. E.; Worthington, P. A.
Eur. J. Org. Chem. 2000, 963.
(7) If the reaction is run in water instead of DMF, only a slightly
greenish yellow color is observed in the reaction suspension
that turns to an orange-ocher color when the reaction is
complete. In this case, a better indication of reaction
progress is to observe the metal surface, dark at the
beginning, and shiny at the end.
(8) Das, S.; Hung, C. H.; Goswani, S. Inorg. Chem. 2003, 42,
5153; and references cited therein.
(9) (a) Finet, J. P. Ligand Coupling Reactions with
Heteroatomic Compounds; Pergamon Press: Oxford, 1998.
(b) Donnelly, D. M. X.; Finet, J. P.; Guiry, P. J.; Nesbitt, K.
Tetrahedron 2001, 57, 413.
Details of the reaction mechanism and new applications of
the above-described protocol aiming to develop a method
usable at a larger scale are currently under investigation in
our laboratories.
Typical Procedure
The carbonyl compound 1 (1 mmol) and allyl bromide (484 mg, 4
mmol) were dissolved in DMF (1 mL), contained in a thick-walled
reaction tube with threaded Teflon cap. Granular indium (230 mg,
2 mmol) was added at once, the reaction tube was tightly closed and
magnetically stirred into a hot oil bath with the temperature careful-
ly set at 40–50 °C, and controlled by an immersion thermometer.
The progress of the reaction was followed by TLC, until total con-
sumption of 1 (usually 2 h) was observed. The resulting mixture
was diluted with EtOAc, poured into water, and washed thrice with
water. The organic extracts were dried over Na2SO4 and evaporated
in vacuo to yield crude homoallylic alcohol 2, which was purified
by flash chromatography. Compounds 2a,17 2b,18 2c,19 2d,20 2g,19
2j,19 2k19 and 2n19 have been previously described and character-
ized in detail.
(10) Radical reactions initiated by In: (a) Miyabe, H.; Naito, T.
Org. Biomol. Chem. 2004, 2, 1267. (b) Miyabe, H.; Ueda,
M.; Nishimura, A.; Naito, T. Org. Lett. 2002, 4, 131; and
references cited therein.
(11) For a study on the direct binding of carbonyl compounds to
the In surface, see: (a) Tachikawa, H.; Kawabata, H. J.
Mater. Chem. 2003, 13, 1293. (b) Tachikawa, H.;
Kawabata, H. J. Organomet. Chem. 2003, 678, 56.
(c) Tachikawa, H.; Kawabata, H. Phys. Chem. Chem. Phys.
2003, 5, 3587.
(12) Araki, S.; Ito, H.; Butsugan, Y. J. Org. Chem. 1988, 53,
1831.
(13) Chan, T. H.; Yang, Y. J. Am. Chem. Soc. 1999, 121, 3228.
(14) For a similar study on the indium-mediated Reformatsky
reaction, see: Babu, S. A.; Yasuda, M.; Shibata, I.; Baba, A.
Org. Lett. 2004, 6, 4475.
(15) For a similar NMR analysis of Ga and Al allyl derivatives,
see: Takai, K.; Ikawa, Y. Org. Lett. 2002, 4, 1727.
(16) It is also possible that no actual organometallic compounds
form under Barbier conditions, but rather a series of single
electron-transfer processes occur at the metal surface, as
previously proposed for other metals: (a) Molle, G.; Bauer,
P. J. Am. Chem. Soc. 1982, 130, 3481; and references cited
herein. (b) Li, C. J.; Zhang, W. C. J. Am. Chem. Soc. 1998,
120, 9102.
Acknowledgment
The authors are grateful to ‘Facultad de Química, Pontificia Univer-
sidad Católica de Chile’, and FONDECYT (Fondo Nacional de
Desarrollo Científico y Tecnológico, Grant No 1060595) for finan-
cial support. Partial Support from MECESUP (Grants PUC-0004
and RED QUIMICA UCH-0116) is also acknowledged. We also
wish to thank Prof. Dr. Juan M. Manríquez for a generous gift of in-
dium wire, Prof. Dr. Ricardo Tapia for providing a sample of com-
pound 1r, and Mr. Sergio E. Alegría for his help in obtaining the
NMR spectra.
(17) Chen, W.; Liu, Y.; Chen, Z. Eur. J. Org. Chem. 2005, 1665.
(18) Solin, N.; Kjellgren, J.; Szabò, K. J. J. Am. Chem. Soc. 2004,
126, 7026.
(19) Huang, Y.; Liao, Y. J. Org. Chem. 1991, 56, 1381.
(20) Doucet, H.; Santelli, M. Tetrahedron: Asymmetry 2000, 11,
4163.
Synlett 2006, No. 19, 3337–3339 © Thieme Stuttgart · New York