Indium-mediated asymmetric Barbier-type allylations: Additions to aldehydes and ketones and mechanistic investigation of the organoindium reagents

Article abstract of DOI:10.1021/jo902173j

(Chemical Equation Presented) We report a simple, efficient, and general method for the indium-mediated enantioselective allylation of aromatic and aliphatic aldehydes and ketones under Barbier-type conditions in a one-pot synthesis affording the corresponding chiral alcohol products in very good yield (up to 99%) and enantiomeric excess (up to 93%). Our method is able to tolerate various functional groups, such as esters, nitriles, and phenols. Additionally, more substituted allyl bromides, such as crotyl and cinnamyl bromide, can be used providing moderate enantioselectivity (72% and 56%, respectively) and excellent diastereoselectivity when employing cinnamyl bromide (<95/5 anti/syn). However, the distereoselectivity when using crotyl bromide was poor and other functionalized allyl bromides under our method afforded low enantioselectivities for the alcohol products. In these types of indium-mediated additions, solvent plays a major role in determining the nature of the organoindium intermediate and we observed the susceptibility of some allylindium intermediates to hydrolysis in protic solvents. Under our reaction conditions using a polar aprotic solvent, we suggest that an allylindium(III) species is the active allylating intermediate. In addition, we have observed the presence of a shiny, indium(0) nugget throughout the reaction, irrespective of the stoichiometry, indicating disproportionation of indium halide byproduct formed during the reaction.

Full text of DOI:10.1021/jo902173j

pubs.acs.org/joc
Indium-Mediated Asymmetric Barbier-Type Allylations: Additions to
Aldehydes and Ketones and Mechanistic Investigation of the
Organoindium Reagents
Terra D. Haddad, Lacie C. Hirayama, and Bakthan Singaram*
Department of Chemistry and Biochemistry, University of California Santa Cruz, 1156 High Street,
Santa Cruz, California 95064
singaram@chemistry.ucsc.edu
Received October 12, 2009
We report a simple, efficient, and general method for the indium-mediated enantioselective allylation
of aromatic and aliphatic aldehydes and ketones under Barbier-type conditions in a one-pot synthesis
affording the corresponding chiral alcohol products in very good yield (up to 99%) and enantiomeric
excess (up to 93%). Our method is able to tolerate various functional groups, such as esters, nitriles,
and phenols. Additionally, more substituted allyl bromides, such as crotyl and cinnamyl bromide,
can be used providing moderate enantioselectivity (72% and 56%, respectively) and excellent
diastereoselectivity when employing cinnamyl bromide (>95/5 anti/syn). However, the distereo-
selectivity when using crotyl bromide was poor and other functionalized allyl bromides under our
method afforded low enantioselectivities for the alcohol products. In these types of indium-mediated
additions, solvent plays a major role in determining the nature of the organoindium intermediate and
we observed the susceptibility of some allylindium intermediates to hydrolysis in protic solvents.
Under our reaction conditions using a polar aprotic solvent, we suggest that an allylindium(III)
species is the active allylating intermediate. In addition, we have observed the presence of a shiny,
indium(0) nugget throughout the reaction, irrespective of the stoichiometry, indicating dispropor-
tionation of indium halide byproduct formed during the reaction.
Introduction
is a cornerstone in organic synthesis.1 The metal-mediated
asymmetric allylation of various carbonyls continues to be a
valuable tool in the formation of the chiral alcohol function-
ality.2 In particular, indium has been shown to be an effective
metal in both allylations and propargylations.3b,d,e,j,k In the
past few decades, indium has become a popular metal due to
the ability of organoindium intermediates to tolerate func-
tionality and ambient conditions along with indium being
nontoxic.3
The first indium-mediated allylation reaction was per-
formed by Butsugan and co-workers,4 in which they used
In° (1.5 equiv), allyl iodide (1.5 equiv), and benzaldehyde
(1 equiv) in N,N-dimethylformamide (DMF) providing the
racemic homoallylic alcohol in 87% yield. Later Loh and co-
workers executed the first asymmetric indium-mediated
allylation of benzaldehyde using cinchonidine as the chiral
director.5 They were able to afford the (R)-enantiomer in
modest enantiomeric excess of 75% using an excess of allyl
The generation of stereogenic centers is a vital part of the
synthesis of natural products and potential drug candidates,
and the methodology developed for the asymmetric synthesis
(1) (a) Williams, D. R.; Walsh, M. J.; Claeboe, C. D.; Zorn, N. Pure Appl.
Chem. 2009, 81, 181–194. (b) Montalvo-Gonzalez, R.; Chavez, D.; Aguirre,
G.; Parra-Hake, M.; Somanathan, R. Synth. Commun. 2009, 39, 2737–2746.
(c) Luderer, M. R.; Bailey, W. F.; Luderer, M. R.; Fair, J. D.; Dancer, R. J.;
Sommer, M. B. Tetrahedron: Asymmetry 2009, 20, 981–998. (d) Vishnumaya,
M. R.; Singh, V. K. J. Org. Chem. 2009, 74, 4289–4297. (e) Chemler, S. R.;
Roush, W. R. Recent Applications of the Allylation Reaction to the Synthesis of
Natural Products. In Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley-VCH:
Weinheim, Germany, 2000; pp 403-490.
(2) (a) Yamamoto, H.; Wadamoto, M. Chem. Asian J. 2007, 2, 692–698.
(b) Garcia, C.; Victor Martin, V. S. Curr. Org. Chem. 2006, 10, 1849–1889.
(c) Nakajima, M.; Kotani, S.; Ishizuka, T.; Hashimoto, S. Tetrahedron Lett.
2005, 46, 157–159. (d) Yanagisawa, A.; Nakamura, Y.; Arai, T. Tetrahedron:
Asymmetry 2004, 15, 1909–1913. (e) Denmark, S. E.; Fu, J. Chem. Rev. 2003,
103, 2763–2793. (f) Yamamoto, Y.; Asao, N. Chem. Rev. 1993, 93, 2207–
2293.
642 J. Org. Chem. 2010, 75, 642–649
Published on Web 12/22/2009
DOI: 10.1021/jo902173j
r
2009 American Chemical Society

Haddad et al.
JOCArticle
bromide (6 equiv). Not only has indium been used to mediate
the allylation of aldehydes but it also has been shown to be
effective in the allylation of ketones.6 The latter additions are
desirable due to the ability to synthesize chiral tertiary
homoallylic alcohols, yet they are less extensively explored
compared to the corresponding asymmetric addition to
aldehydes.7 Loh et al. reported that the asymmetric allyla-
tion of acetophenone using indium, cinchonidine, and
6 equiv of allyl bromide gave nearly racemic alcohol pro-
duct whereas the allylation of trifluoroacetophenone affor-
ded the homoallylic alcohol in 70% ee.6b Additionally,
indium has been used in the addition of substituted allyl
bromides to aldehydes,8 specifically Bustugan and co-work-
ers achieved a 34:66 syn/anti ratio using indium, crotyl
bromide, and benzaldehyde.4
Our own efforts in organometallic addition reactions
began with screening various limonene-based amino alco-
hols as effective chiral directors in the diethylzinc addition to
aldehydes.9 This led us to study the effectiveness of other
terpene-derived amino alcohols as chiral directors in the
indium-mediated Barbier-type asymmetric allylation of al-
dehydes. Accordingly, we examined various limonene and
pinene-based amino alcohols synthesized in our labora-
tory,10 along with commercially available ones in the in-
dium-mediated Barbier-type allylation of aldehydes and
ketones.11,12 These studies revealed an efficient and simple
one-pot procedure using a commercially available amino
alcohol for the formation of chiral homoallylic alcohols in
moderate to excellent enantioselectivity. Herein, we describe
the full report on this work in our study of the asymmetric
indium-mediated allylation of aldehydes and ketones using
simple and substituted allyl and crotyl bromides. We will also
give an account of our investigation of the allylindium
intermediates formed using indium metal (In°) and allyl
bromides under different solvent conditions.
Results and Discussion
Indium-Mediated Asymmetric Barbier-Type Allylations of
Aldehydes. Following the stoichiometry that was reported
in the literature for these types of indium-mediation allya-
tions, we began our investigation by mixing benzaldehyde
(1 equiv), a chiral ligand (2 equiv), In° (2 equiv), and allyl
bromide (6 equiv) at -78 °C for 1.5 h. Under these condi-
tions, we obtained the corresponding homoallylic alcohol in
high yield but low enantioselectivity of 40% using limonene-
derived (1S,2S,5R)-2-amino-2-methyl-5-(prop-1-en-2-yl)cy-
clohexanol13 as the chiral director. Next, we screened
various commercially available chiral amino alcohols in this
reaction. We were pleased to find that (þ)-(1S,2R)-2-amino-
1,2-diphenylethanol (þ)-1 proved to be an effective chiral
director under our indium-mediated Barbier-type reaction
conditions and gave the product alcohol in 60% ee.
We speculated that the enantioselectivity could be further
improved by adjusting the stoichiometry of allyl bromide.
We reasoned that the excess of allyl bromide could either be
interacting with the chiral auxiliary or affecting the nature of
the allylindium intermediates formed during the reaction.
Hence, we rationalized that reducing the equivalents of allyl
bromide would minimize the interference with the ligand
and/or control the allylindium species formed, thereby in-
creasing both conversion and enantioselectivity of the reac-
tion. In addition, a stoichiometric amount of a Lewis base,
such as pyridine, was added with the aim of scavenging the
intermediate indium salts and thereby minimizing their
influence on enantioselectivity. We decided to use hetero-
aromatic Lewis bases that contained a structural motif
similar to cinchonidine, an effective chiral auxiliary for allyl
indium reactions.5,6b Using quinine (2 equiv) as the Lewis
base, In° (2 equiv), (þ)-1 (2 equiv), and allyl bromide
(6 equiv), we were able to afford the homoallylic alcohol 2a
in good conversion of 84% but low enantioselectivity of
30% ee (Table 1, entry 1). Changing the Lewis base to 2,2-
bipyridine showed only a slight improvement in the enantio-
meric excess (Table 1, entry 2). Employing pyridine and
increasing the equivalents of allyl bromide afforded 2a in
good enantiomeric excess of 76% (Table 1, entry 3). By
simply decreasing the equivalents of allyl bromide from six to
two, we significantly improved the enantioselectivity to 93%
for 2a (Table 1, entry 4). We were pleased to find that the
addition of pyridine did indeed improve enantioselectivity
and conversion. Enantiomeric excess dropped to 60% and
the conversion to 65% in the absence of pyridine (Table 1,
entry 5). We found that reducing the amount of either In°,
allyl bromide, or the chiral ligand resulted in lower yield and
enantioselectivity 2a (Table 1, entries 6-8). It should be
pointed out that we observed the presence of shiny indium
(3) (a) Alcaide, B.; Almendros, P.; del Campo, T. M. Eur. J. Org. Chem.
2008, 2628–2634. (b) Xu, B.; Hammond, G. B. Chem.;Eur. J. 2008, 14,
10029–10035. (c) Kargbo, R. B.; Cook, G. R. Curr. Org. Chem. 2007, 11,
1287–1309. (d) Kargbo, R.; Takahashi, Y.; Bhor, S.; Cook, G. R.; Lloyd-
Jones, G. C; Shepperson, I. R. J. Am. Chem. Soc. 2007, 129, 3846–3847.
(e) Tan, K. L.; Jacobsen, E. N. Angew. Chem. 2007, 119, 1315–1317. (f) Lee,
P. H. Bull. Korean Chem. Soc. 2007, 28, 17–28. (g) Yadav, J. S.; Subba Reddy,
B. V.; Vishnumurthy, P.; Biswas, S. K. Tetrahedron Lett. 2007, 48, 6641–
6643. (h) Loh, T.-P.; Chau, G.-L. Chem. Commun. 2006, 26, 2739–2749.
€
€
€€
(i) Kallstrom, S.; Jagt, R. B. C.; Sillanpaa, R.; Feringa, B. L.; Minnaard, A.
J.; Leino, R. Eur. J. Org. Chem. 2006, 3826–3833. (j) Cook, G. R.; Kargbo,
R.; Bikash Maity, B. Org. Lett. 2005, 7, 2767–2770. (k) Foubelo, F.; Yus, M.
Tetrahedron: Asymmetry 2004, 15, 3823–3825. (l) Ritson, D. J.; Cox, R. J.;
Berge, J. Org. Biomol. Chem. 2004, 2, 1921–1933. (m) Paquette, L. A.
Synthesis 2003, 765–774. (n) Podlech, J.; Maier, T. C. Synthesis 2003, 5,
633–655. (o) Lee, W.; Kim, K.-H.; Surman, M. D.; Miller, M. J. Org. Chem.
2003, 68, 139–149. (p) Lu, W.; Chan, T. K. J. Org. Chem. 2000, 65, 8589–
8594. (q) Chauhan, K. K.; Frost, C. G. J. Chem. Soc., Perkin Trans. 1 2000,
18, 3015–3019. (r) Li, C.-J.; Chan, T.-H. Tetrahedron 1999, 55, 11149–11176.
(4) Araki, S.; Ito, H.; Butsugan, Y. J. Org. Chem. 1988, 53, 1831–1833.
(5) Loh, T.-P.; Zhou, J.-R.; Yin, Z. Org. Lett. 1999, 1, 1855–1857.
(6) (a) Bao, Z. J.; Lu, J.; Ji, S. J. Chin. Chem. Lett. 2007, 18, 1061–1063.
(b) Loh, T.-P.; Zhou, J.-R.; Li, X.-R. Tetrahedron Lett. 1999, 40, 9333–9336.
(7) (a) Teo, Y.-C.; Goh, J.-D.; Loh, T.-P. Org. Lett. 2005, 7, 2743–2745.
(b) Lu, J.; Ji, S.-J.; Teo, Y.-C.; Loh, T.-P. Tetrahedron Lett. 2005, 46, 7435–
7437. (c) Wu, T. R.; Shen, L.; Chong, J. M. Org. Lett. 2004, 6, 2701–2704.
(d) Cunningham, A.; Woodward, S. Synlett 2002, 1, 43–44. (e) Yasuda, M.;
Kitahara, N.; Fujibayashi, T.; Baba, A. Chem. Lett. 1998, 27, 743–744.
(f) Yanagisawa, A.; Nakashima, H.; Ishiba, A.; Yamamoto, H. J. Am. Chem.
Soc. 1996, 118, 4723–4724. (g) Soai, K.; Ishizaki, M.; Yokoyama, S. Chem.
Lett. 1987, 2, 341–344.
(8) (a) Tan, K. T.; Chng, S. S.; Cheng, H. S.; Loh, T. P. J. Am. Chem. Soc.
2003, 125, 2958–2963. (b) Loh, T. P.; Ken Lee, C. L.; Tan, K. T. Org. Lett.
2002, 4, 2985–2987. (c) Lloyd-Jones, G. C.; Russell, T. Synlett 1998, 903–905.
(9) (a) Steiner, D.; Sethofer, S. G.; Goralski, C. T.; Singaram, B. Tetra-
hedron: Asymmetry 2002, 13, 1477–1483. (b) Steiner, D.; Ivison, L.; Goralski,
C. T.; Appell, R. B.; Gojkovic, J. R.; Singaram, B. Tetrahedron: Asymmetry
2002, 13, 2359–2363.
(10) Chrisman, W.; Camara, J. N.; Marcellini, K.; Singaram, B.;
Goralski, C. T.; Hasha, D. L.; Rudolf, P. R.; Nicholson, L. W.; Borodychuk,
K. K. Tetrahedron Lett. 2001, 42, 5805–5807.
(11) Hirayama, L. C.; Gamsey, S.; Knueppel, D.; DeLaTorre, K.; Steiner,
D.; Singaram, B. Tetrahedron Lett. 2005, 46, 2315–2318.
(12) Haddad, T. D.; Hirayama, L. C.; Tanyton, P.; Singaram, B.
Tetrahedron Lett. 2008, 49, 508–511.
(13) (1S,2S,5R)-2-Amino-2-methyl-5-(prop-1-en-2-yl)cyclohexanol was
synthesized from the cis-limonene epoxide and ammonia (unpublished
results).
J. Org. Chem. Vol. 75, No. 3, 2010 643

Haddad et al.
JOCArticle
TABLE 1. Optimization of Enantioselective Allylation of Benzaldehydea
TABLE 2. The Asymmetric Indium-Mediated Barbier-Type Allylation
of Aldehydesa
allyl
%
%
entry In° (þ)-1 Lewis base (equiv) bromide convb eeb (S)c
1
2
3
4
5
6
7
8
2
2
2
2
2
1
2
1
2
2
2
2
2
2
2
1
quinine (2)
2,2-bipyridine (2)
py (2)
py (2)
py (0)
py (2)
py (2)
py (1)
6
2
6
2
2
2
1
1
84
85
99
99
65
50
55
50
30
36
76
93
60
66
70
79
aTable values refer to number of equivalents. Reactions run with
benzaldehyde (0.5 mmol). bDetermined by chiral GC analysis. cAbsolute
configuration determined by comparison of the optical rotation with
literature value.6b,15
nuggets throughout the duration of this reaction, indicating
the disproportionation of indium bromide salts.14
For the rest of our investigations we used the opti-
mized stoichiometry of 2 equiv each of In°, allyl bromide,
(þ)-1, and pyridine with respect to aldehyde in THF:n-
hexane (7:1) (Table 2).11 We found that our method provided
the homoallylic alcohol in excellent yields (90-99%) and
enantioselectivities (76-93% ee) (Table 2, entries 1 to 7).
This method demonstrates the chemoselectivity of the ally-
lindium intermediates in the presence of other functional
groups such as an ester providing only allylation of the
aldehyde yielding 2e (Table 2, entry 5). Under our reaction
conditions, aldehyde allylation occurs in the presence of a
nitrile affording 2f in high enantiomeric excess of 80%
(Table 2, entry 6). Additionally, these results show that the
para-substituted aldehydes which are electron-donating
through resonance (Table 2, entries 2 and 8), give a higher
enantiomeric excess than those with an electron-withdraw-
ing group in the para position (Table 2, entries 9 and 10).
Possibly the electron-withdrawing groups increase the reac-
tivity of the aldehydes, therefore decreasing the enantio-
selectivity for those substrates under these reaction condi-
tions. We also demonstrated the allylation of an aliphatic
aldehyde, specifically cyclohexanecarbaldehyde, yielding 2g
in high yield and enantiomeric excess (Table 2, entry 7).
Indium-Mediated Asymmetric Barbier-Type Allylations of
Aldehydes with Substituted Allyl Bromides. We next turned
our attention to see whether this indium-mediated allylation
could be extended to substituted allyl bromides. Using our
optimized reaction conditions, we began our study by simply
replacing allyl bromide with crotyl bromide (Scheme 1).
Although the enantioselectivity for the reaction was good
(72% ee for both syn- and anti-product alcohols 3), the
diastereoselectivity was poor (43:57 syn/anti). To improve
the diastereoselectivity, we examined the effects of reaction
aReactions run with In° (2.0 mmol), 1a (2.0 mmol), pyridine
(2.0 mmol), allyl bromide (2.0 mmol), and aldehyde (1.0 mmol) in
THF/n-hexane at -78 °C for 1.5 h. bIsolated yield of analytically pure
product. cDetermined by chiral GC analysis. dAbsolute configura-
tion determined by comparison of the optical rotation with literature
value,6b,15 all others were assigned by analogy.
SCHEME 1. Allylation of Benzaldehyde with Crotyl Bromide
and (þ)-1
(14) (a) Kloo, L.; Rosdahl, J.; Taylor, M. J. Polyhedron 2002, 21, 519–524.
(b) Tuck, D. G. Chem. Soc. Rev. 1993, 22, 269–276.
(15) For an asymmetric Reformatsky reaction with indium and cincho-
nidine see: Johar, P. S.; Araki, S.; Butsugan, Y. J. Chem. Soc., Perkin Trans.
1 1992, 711–713.
644 J. Org. Chem. Vol. 75, No. 3, 2010

Haddad et al.
JOCArticle
TABLE 3. The Asymmetric Indium-Mediated Barbier-Type Allylation of Benzaldehyde with Functionalized Allyl Bromides
1
aSyn/anti ratio determined by H NMR. bDetermined by chiral GC analysis. cThe reaction was conducted with acetophenone and the optimized
ketone conditions, in THF at 25 °C for 24 h. dDetermined by chiral HPLC analysis.
temperature. However, increasing the reaction temperature
from -78 to 25 °C resulted in a similar ratio (45:55 syn/anti).
Increasing the temperature further to 65 °C yielded a mixture
of 3 with a slight bias favoring the syn diasteromer (58:42 syn/
anti). Additionally, heating crotyl bromide and In° under
reflux for 30 min prior to the addition of ligand and pyridine
did not enhance the diastereoselectivity any further. During
the course of these reactions, the presence of shiny indium
nuggets was observed as previously seen in our other indium-
mediated allylation reactions.
We also examined this reaction using other substituted
allylic bromides (Table 3). Using the optimized reaction
conditions, methallyl bromide reacted with benzaldehyde
to give 4a in 70% yield and 45% ee (Table 3, entry 2). The
methallylation of acetophenone took 24 h at 25 °C and
afforded 4b in a moderate yield of 55% and very low
enantiomeric excess of 16% (Table 3, entry 3). The allylation
of benzaldehyde with prenyl bromide gave 5 in 54% yield
and 56% ee (Table 3, entry 4), which was lower than the
published results.5 Under Barbier-type conditions, cinnamyl
bromide reacted with benzaldehyde and furnished 6 in a very
high diastereomeric ratio (>95:5 anti) and moderate enan-
tiomeric excess of 56% (Table 3, entry 5). The high diaster-
eoselectivity realized in this reaction can be attributed to the
reversibility of this reaction in THF. The low conversion is
more than likely due to the increased steric bulk of the phenyl
substituent. These results demonstrate that sterically de-
manding substituted allyl bromides decrease the enantios-
electivity with use of our method and negatively influence the
efficiency of our chiral director.
under Barbier-type reaction conditions warranted the study
of allylation of ketones under the optimized allylation con-
ditions used for aldehydes. Unfortunately, the reaction of
allyl bromide with acetophenone under the Barbier condi-
tions gave abysmal conversion of less than 5% to the
corresponding homoallylic alcohol. We suspected that since
ketones are generally less reactive, higher temperatures
would be needed. We were pleased to find that running the
reaction for 24 h at 25 °C provided the corresponding
homoallylic alcohol in moderate yield and enantiomeric
excess (Table 4, entry 1).11 Using an excess of allyl bromide
(3- to 6-fold) decreased the yield of the product while
maintaining the same level of asymmetric induction. This
result confirmed that 2 equiv of allyl bromide is optimal and
supported our hypothesis that an excess of allyl bromide
affects the efficiency of the reaction. Even in the allylation of
ketones, we observed the transformation of gray indium
powder into shiny indium flakes during the reaction.
This method is fairly general and provided the correspond-
ing tertiary homoallylic alcohols for both aliphatic and
aromatic ketones in moderate to high yield and enantios-
electivity (Table 4).11 In addition, either enantiomer of the
commercially available amino alcohol, (þ)-1 or (-)-1, af-
forded the antipodal homoallylic alcohol products (Table 4,
entries 7 and 8). The ligand can also be recycled and reused
as demonstrated in the allylation of triflouroacetophenone
affording (R)-7g in excellent yield and enantioselectivity
(Table 4, entry 8). This method can be applied to hetero-
aromatic compounds as shown by the allylation of 2-acet-
ylfuran, which provided an excellent yield of >99% but only
moderate enantiomeric excess (Table 4, entry 9). Also docu-
mented by others, the allylindium reagent was able to
Indium-Mediated Asymmetric Barbier-Type Allylations of
Ketones. Our successful asymmetric allylation of aldehydes
J. Org. Chem. Vol. 75, No. 3, 2010 645

Haddad et al.
JOCArticle
TABLE 4. The Asymmetric Indium-Mediated Barbier-Type Allylation of Ketonesa
aReactions run with In° (1.0 mmol), 2-amino-1,2-diphenylethanol (1.0 mmol), pyridine (1.0 mmol), allyl bromide (1.0 mmol), and ketone (0.5 mmol)
in THF at room temperature for 24 h. bIsolated yield of analytically pure product; all products greater than 90% by 1H NMR. cBased on unreacted
starting material. dDetermined by chiral GC analysis. eAbsolute configuration determined by comparison of the optical rotation with literature value,17
all others were assigned by analogy. fDetermined by chiral RP-HPLC analysis. gReaction run with recovered and purified ligand.
tolerate functionality, such as phenol, ester, and nitrile,
showing the versatility of the indium-mediated allylations
compared to allylations mediated by Gringard reagents
(Table 4, entries 2 to 4). Under our reaction conditions,
aliphatic ketones showed only moderate enantiomeric in-
duction (Table 4, entries 5 and 6). On the other hand,
trifluoroacetophenone afforded (S)-7g in 80% ee (Table 4,
entry 7). In the indium-mediated allylation of ketones (þ)-1
is a better chiral director than cinchonidine.6b This metho-
dology not only expands the scope of indium-mediated
additions but also provides a simple synthesis of tertiary,
chiral alcohol centers, a desirable functionality.16
Mechanistic Studies. Although indium has gained appeal
in Barbier-type asymmetric allylation reactions, little is
known about the mechanism. Perusal of the literature
showed that despite intensive research efforts in this area,
the nature of the active species remains elusive due the
fleeting nature of the indium intermediates.3n,18,4 Reactions
of allylhalides with indium metal can produce several orga-
noindium species such as allylindium(I), allylindium(III)
dihalide, diallylindium(III) halide, triallylindium(III), and
allylindium sesquihalide depending on solvent and reaction
conditions.
In 1974, Gynane and Worral suggested alkylindium
sesquihalides, R3In2X3, as intermediates in the reaction of
(16) (a) Riant, O.; Hannedouche, J. Org. Biomol. Chem. 2007, 5, 873–888.
(b) Nicolaou, K. C.; Kim, D. W.; Baati, R. Angew. Chem., Int. Ed. 2002, 41,
3701–3704.
(17) (a) Canales, E.; Prasad, K. G.; Soderquist, J. A. J. Am. Chem. Soc.
2005, 127, 11572–11573. (b) Wu, T. R.; Shen, L.; Chong, J. M. Org. Lett.
2004, 6, 2701–2704.
(18) (a) Babu, S. A.; Yasuda, M.; Shibata, I.; Akio Baba, A. J. Org. Chem.
2005, 70, 10408–10419. (b) Capps, S. M.; Lloyd-Jones, G. C.; Murray, M.;
Peakman, T. M.; Walsh, K. E. Tetrahedron Lett. 1998, 39, 2853–2856. (c)
Tussa, L.; Lebreton, C.; Mosset, P. Chem.;Eur. J. 1997, 3, 1064–1070.
646 J. Org. Chem. Vol. 75, No. 3, 2010

Haddad et al.
JOCArticle
SCHEME 2. Formation of a Possible Indium(III) Center Intermediate
In° with various alkyl iodides based on mass and vibrational
spectral analyses.19 In analogy, Araki et al. proposed an
allylindium sesquihalide as the reactive intermediate in
anhydrous DMF as evidenced by the observation of 1H
NMR signals at 1.7 and 2.1 ppm with relative integration
of 2:1.4 Later, Chan et al. prepared an allylindium reagent
from the reaction of In° and InI with diallylmercury in a
1
mixture of DMF and D2O and attributed the observed H
NMR signal at 1.7 ppm to allylindium(I).20 Araki et al.
reported that in the reaction of indium(I) iodide with ally-
liodide in THF-d8, an allylindium(III) diiodide species is
FIGURE 1. Proposed transition state complex containing an
allylindium(III) intermediate.
1
formed and it gives a signal at 2.0 ppm in the H NMR
1
DMF-d7 and analyzed the reaction mixture by H NMR
spectrum.21 More recently, Pretie and co-workers reported
similar results in DMF-d7 where two allylic signals, 1.75 and
2.02 ppm, were observed upon reacting In° with allyl bro-
mide and the ratio of these signals varied with time.22 We felt
that observing the organoindium intermediates that are
formed during the reaction of allyl bromides and substituted
allyl bromides with indium metal in various aprotic, protic,
and aqueous solvents by 1H NMR spectroscopy would
provide insight into the mechanism of these reactions.
Mechanistic Studies of Indium with Allyl Bromides. We
began our investigation by mixing allyl bromide and indium
in various ratios in DMF-d7 and followed the reactions by 1H
NMR spectroscopy. Regardless of the stoichiometry, we
observed the aforementioned peaks at 1.7 and 2.1 ppm.
However, we found them to be present in roughly a 1:1 ratio.
This led us to concur that two distinct allyl indium species are
formed when indium is reacted with allyl bromide in anhy-
drous solvents. We also observed a single 1H NMR signal at
1.7 ppm in the reaction of allyl bromide with indium in D2O.
Moreover, addition of D2O to the allylindium solution in
DMF-d7 selectively hydrolyzed the allylindium resonating at
spectroscopy. In this instance, we observed only one allyl
indium signal at 2.1 ppm. This result suggested that, under
our reaction conditions, the allyl indium responsible for the
signal at 2.1 ppm could be the active allylating species and is
formed exclusively in the presence of pyridine. Additionally,
when benzaldehyde was added to a mixture of allylindiums
in DMF-d7, the signal at 2.1 ppm disappeared while the
signal at 1.7 ppm persisted, indicating again that the allyl
indium species responsible for the signal at 2.1 ppm is likely
the active allylating intermediate. It should be pointed out
that shiny indium flakes were observed throughout these
NMR tube reactions indicating possible disproportionation
of allylindium intermediates.
We have observed repeatedly and reproducibly the pre-
sence of metallic indium throughout the duration of these
reactions. Conventionally it is accepted that the reduction of
In(III) to In(0) proceeds through several one-electron reduc-
tions where In(II) and In(I) are fleeting intermediates due to
the reactivity of the 5s2 configuration (eq 1).23 Hence, it is
plausible that these indium species are in equilibrium and the
disproportion of various allylindium intermediates results in
the regeneration of In° (eq 2). In fact, a recent paper repo-
rted such an equilibrating allylindium species when allyl
bromide and indium(0) were reacted in an ionic liquid (IL)
and observed the interconversion of an In(I) species into
In(III) species.24 Apparently, organoindium intermediates
are not formed in discrete amounts, but rather are in
equilibrium with each other and their concentration depends
on reaction conditions and solvents.
1
2.1 ppm leaving the H NMR signal at 1.7 ppm. Addition-
ally, when air was bubbled through a separate allylindium
solution in DMF-d7, the peak at 2.1 ppm remained un-
changed, but the peak at 1.7 ppm disappeared. It is reason-
able to infer that the allylindium responsible for the signal at
1.7 ppm is oxidatively unstable whereas the allylindium
responsible for the signal at 2.1 ppm is hydrolytically un-
stable.
As previously stated, the maximum enantioselectivity and
quantitative conversion to the homoallylic alcohol product
was achieved with use of a 2:2:1 ratio of In° allyl bromide,
and carbonyl, respectively. We also found that the addition
of pyridine had a beneficial effect on the yield and enantio-
meric purity of the alcohol product. To investigate the role of
pyridine, we mixed allyl bromide, indium, and pyridine in
InðIIIÞ f InðIIÞ f InðIÞ f Inð0Þ
3InþðsolnÞ f 2In°ðsÞ þ In3þðsolnÞ
ð1Þ
ð2Þ
Consequently, the formation of a single allylindium inter-
mediate in D2O was puzzling. To address this question, we
studied the reaction of allyl bromide with In° in D2O in
an NMR tube. We noticed gas evolution along with the
(19) Gynane, M. J. S.; Worrall, I. J. J. Organomet. Chem. 1974, 81, 329–
334.
(20) Chan, T. H.; Yang, Y. J. Am. Chem. Soc. 1999, 121, 3228–3229.
(21) Araki, S.; Ito, H.; Katsumura, N.; Butsugan, Y. J. Organomet. Chem.
1989, 369, 291–296.
(23) Pardoe, J. A. J.; Downs, A. J. Chem. Rev. 2007, 107, 2–45.
(24) Law, M. C.; Cheung, T. W.; Wong, K. Y.; Chan, T. H. J. Org. Chem.
2007, 72, 923–929.
ꢀ
(22) Pretie, M; Perez-Carvajal, A. Synlett 2006, 19, 3337–3339.
J. Org. Chem. Vol. 75, No. 3, 2010 647

Haddad et al.
JOCArticle
TABLE 5. 1H NMR Spectroscopy Study of Allyl Bromides with Indium(0)
entry
allyl bromide
solvent
allylindium
allylic protons (δ, ppm)a
product alcohol
1
2
3
4
5
6
crotyl bromide
crotyl bromide
cinnamyl bromide
cinnamyl bromide
prenyl bromide
methallyl bromide
MeOH-d4
D2O
MeOH-d4
D2O
MeOH-d4
MeOH-d4
crotylindiumb
crotylindiumb
cinnamylindium
cinnamylindium
prenylindium
1.83-1.91, m
3
3
6
6
1.63-1.70, m
2.11, d J = 8.75 Hz
1.93, d J = 8.75 Hz
1.83, d J = 9 Hz
5
3-deuterio-2-methylprop-1-ene
am = multiplet, d = doublet. bE- and Z-crotyl indium.
formation of a white precipitate. In an attempt to dissolve the
white solid in a deuterated solvent, we added CDCl3 to the
NMR tube and monitored the chloroform layer by 1H NMR
spectroscopy. We could clearly see the absence of any
allylindium intermediates in the CDCl3 layer and identified
the presence of deuterated propene. The subsequent addition
of benzaldehyde to the NMR tube, however, showed the
formation of homoallylic alcohol indicating incomplete
hydrolysis of the allylindium intermediates and their insolu-
bility in chloroform. We therefore suggest that a mixture of
allylindium intermediates are formed even in aqueous solu-
tions and the hydrolytically unstable allylindum intermedi-
ate is converted to deuterated propene in D2O. The white
precipitate was isolated by centrifugation, dissolved in DCl
We next focused on the nature of the organoindium
intermediates that are formed from the substituted allyl
bromides under aqueous or protic conditions. Crotyl bro-
mide and indium react in both D2O and methanol-d4 to give a
mixture of crotylindium intermediates, presumably cis- and
trans-crotylindium, which then reacts with benzaldehyde to
form 3 (Table 5, entries 1 and 2). However, cinnamyl bro-
mide gives essentially a single organoindium species in both
D2O and methanol-d4 and the addition of benzaldehyde
yielded the anti-homoallylic alcohol 6 (Table 5, entries 3
and 4). Prenyl bromide reacts with In° in methanol-d4 to
from an allylindium intermediate that produces 5 in the
presence of benzaldehyde (Table 5, entry 5). In contrast,
methallylindium intermediate, formed by the reaction of
methallyl bromide and In°, is hydrolyzed rapidly to deuter-
ated isobutene in methanol-d4 and no homoallylic alcohol
product is observed upon the addition of benzaldehyde.
Since we conducted these reactions in aqueous or protic
solvents we did not observe any downfield allylindium signal
by 1H NMR spectroscopy.26
1
and analyzed by H NMR spectroscopy. Since no organic
component could be detected in this solution the white
precipitate presumably is indium deuteroxide.
We were still intrigued by the nature of the active allylating
species under our reaction conditions and decided to further
examine the asymmetric indium-mediated allylation of ben-
zaldehyde using (þ)-1. We synthesized the lithium salt of
chiral ligand (þ)-1 then added indium tribromide as a
complexing Lewis acid followed by benzaldehyde. After
stirring for 1 h at -78 °C, we then transferred this solution
to a mixture of In° and allyl bromide in THF at 25 °C. This
reaction afforded 2a in essentially quantitative yield and
60% ee, a value identical with that obtained with 1:1:1
stoichiometry of In°:benzaldehyde:allyl bromide (vide
supra) (Scheme 2). We tentatively suggest that the identity
of the In(III) complex may be similar to that shown in
Scheme 2. The simplest transition state that incorporates
all the experimental data is provisionally proposed borrow-
ing the analogy from Et2Zn addition to carbonyl compounds
and oxazaborolindine (Figure 1).25 This would result in a
si-face attack of the aldehyde, which is in agreement with the
fact the S-homoallylic alcohol is obtained when using (þ)-1
as the chiral director. The phenyl rings should aid in blocking
the top face of the chiral In(III) complex therefore forcing
coordination from the bottom face, which is similar to the
rationale that was proposed to explain the enantioselectivity
exhibited by the chiral reagent oxazaborolidine in asym-
metric additions to aldehydes. Additionally in D2O, racemic
homoallylic alcohol product is obtained due to the fact that
our chiral director is insoluble in this solvent and therefore
cannot form the chiral In(III) complex.
Conclusions
We have developed methodology for the asymmetric
indium-mediated allylation of aldehydes and ketones achiev-
ing excellent yield and enantioselectivity to the correspond-
ing chiral alcohol products using a recyclable and reusable
commercially available amino alcohol. This methodology
was extended to the addition of substituted allyl bromides
including crotyl, cinnamyl, prenyl, and methallyl bromide
to benzaldehyde. Unfortunately, these substituted allyl
bromides did not afford high enantioselectivity and/or dis-
tereoselectivity, aside from cinnamyl bromide, which pro-
duced a very high diastereomeric ratio (>95:5 anti) but a
moderate enantiomeric excess of 56% ee. We were also able
to demonstrate the chemoselectivity of allylindium reagents
using our methodology. In addition, this provides a simple
way to synthesize secondary or tertiary chiral alcohols under
Barbier-type conditions.
We have shown that solvent plays a major role in the
formation of organoindium intermediates. It is plausible that
aqueous solvents such as D2O induce the formation of a
more hydrolytically stable organoindium species while polar
aprotic solvents such as DMF-d7 produce organoindium
species of two different oxidation states as evidenced by 1H
NMR spectroscopy. Our mechanistic studies show that
(25) Corey, E. J.; Cimprich, K. A. J. Am. Chem. Soc. 1994, 116, 3151–
(26) Yasuda, M.; Haga, M.; Baba, A. Organometallics 2009, 28, 1998–
3152.
2000.
648 J. Org. Chem. Vol. 75, No. 3, 2010

Haddad et al.
JOCArticle
contrary to what is conventionally thought, allylindium
compounds are not entirely hydrolytically stable. The ally-
lindium intermediate formed with 3-bromo-2-methylbutene
was rather unstable in methanol and water and was hydro-
lyzed almost immediately. On the other hand, some allylin-
dium intermediates are more stable and react with
benzaldehyde in both water and methanol to provide the
corresponding alcohol product. However, in our studies, the
addition of pyridine allows the sole formation of the allyl
indium(þ3) species by possible coordination. We have also
proposed a transition state model containing an In(III)
center where the allylindium species is coordinated from
the bottom, which is in agreement with the fact that we
obtain the (S)-homoallylic alcohol product. Therefore, we
believe that an In(III) intermediate forms under our reaction
conditions and then reacts with the electrophile, either an
aldehyde or ketone, to provide the corresponding homo-
allylic alcohol product.
bromide (1 mmol). The flask was purged with argon (4 vacuum-
backfill cycles) and the mixture was stirred vigorously at 25 °C
for 30 min, at which time dry n-hexane (1 mL) was added, the
solution was cooled to -78 °C (dry ice/acetone bath), and
freshly distilled benzaldehyde (0.5 mmol) was added dropwise.
After 1.5 h the reaction was quenched with saturated ammo-
nium chloride (1 mL) and allowed to warm to 25 °C. The organic
layer was diluted with n-hexane (5 mL), decanted, dried with
anhydrous magnesium sulfate, filtered through a small plug of
silica gel ,and evaporated under reduced pressure.
2-Methyl-1-phenylbut-3-en-1-ol, 3 (Table 3, entry 1). Follow-
ing the general procedure above, 3 was obtained as a clear,
yellow oil (0.080 g, 99% yield). 1H NMR (250 MHz, CDCl3): δ
0.88 (d, J = 6.75 Hz, 3H, anti), 1.02 (d, J = 6.75 Hz, 3H, syn),
2.51 (sextet, J = 7.75 Hz, 2H, anti), 2.61 (sextet, J = 7.75 Hz,
2H, syn), 4.36 (d, J = 8 Hz, 1H, anti), 4.61 (d, J = 5.5 Hz, 1H,
syn), 5.02-5.09 (m, 2H, syn), 5.17-5.25 (m, 2H, anti), 5.69-5.89
(m, 2H, both diast.), 7.24-7.37 (m, 10H, both diast.). 13C NMR
(62.9 MHz, CDCl3): syn δ 14.2, 44.8, 77.5, 115.7, 126.7, 127.5,
128.2, 140.5, 142.8; anti δ 16.7, 46.5, 78.1, 117.0, 127.0, 127.8,
128.4, 140.8, 141.5, 142.6. Enantiomeric excess was determined
to be 72% by chiral HPLC analysis. HPLC conditions: 99:1
hexanes/iPrOH, 1.0 mL/min, tR for all diastereomers = 17.67,
19.22, 21.23, 22.27.
General Procedure for the Allylation of Ketones 4b and 7a-h.
To a 25 mL round-bottomed flask charged with stir bar were
added cooled under Argon, indium (1 mmol), 2-amino-1,2-
diphenylethanol (1 mmol), and THF (7 mL). Pyridine (1 mmol)
and allyl bromide (1 mmol) were added and the entire mixture
was allowed to mix for 30 min at which point the ketone
(0.5 mmol) was added dropwise. After 24 h, the reaction was
quenched with saturated ammonium chloride (NH4Cl, 6 mL)
and transferred to a separatory funnel with either hexanes
(5 mL) or ethyl acetate (EtOAc, 5 mL). The aqueous phase
was removed and the organic phase was washed with dilute
hydrochloric acid (HCl, 2 ꢀ 8 mL) and brine (1 ꢀ 8 mL), dried
with magnesium sulfate (MgSO4), filtered through a silica plug,
and evaporated in vacuo to yield the alcohol product.
Experimental Section
General Procedure for the Allylation of Aldehydes 2a-k. An
oven-dried 50 mL round-bottomed flask was cooled under
argon and charged with (1S,2R)-(þ)-2-amino-1,2-diphenyletha-
nol (2 mmol), indium powder (2 mmol), anhydrous THF
(14 mL), anhydrous pyridine (2 mmol), and allyl bromide
(2 mmol). The mixture was stirred vigorously at 25 °C for
30 min, at which time n-hexane (2 mL) was added, the solution
was cooled to -78 °C (dry ice/acetone bath), and freshly distilled
aldehyde (1 mmol) was added dropwise. After 2 h the reaction
was quenched with saturated ammonium chloride (2 mL) and
allowed to warm to 25 °C. Deionized water was added and the
mixture was extracted with n-hexane (2 ꢀ 5 mL), then the
combined organic layers were dried with anhydrous magnesium
sulfate, filtered through a small plug of silica gel, and evaporated
to give the alcohol product as an oil.
4-Chloro-r-(2-propenyl)benzenemethanol, 2h (Table 2,
entry 8). Following the general procedure above, 2h was ob-
tained as a clear, colorless oil (0.168 g, 92% yield). 1H NMR (250
MHz, CDCl3): δ 2.44-2.50 (m, 2H), 2.77 (s, 1H, OH),
4.66-4.71 (t, J = 6.3 Hz, 1H), 5.12-5.18 (m, 2H), 5.69-5.86
(m, 1H), 7.18-7.33 (m, 4H). 13C NMR (62.9 MHz, CDCl3): δ
43.8, 72.5, 118.8, 127.2, 128.5, 132.9, 133.9, 142.2. Enantiomeric
excess was determined to be 93% by chiral GC analysis. GC
conditions: 151 °C isothermal; tR for the (R)-alcohol = 37.45
min, tR for the (S)-alcohol = 38.32 min.
General Procedure for the Allylation of Benzaldehyde with
Substituted Allyl Bromides to Provide Products 3, 4a, 5, and 6. An
oven-dried 25 mL round-bottomed flask equipped with a stirbar
was cooled under argon and charged with indium powder
(1 mmol), (1S,2R)-2-amino-1,2-diphenylethanol (1 mmol), an-
hydrous THF (7 mL), anhydrous pyridine (1 mmol), and allyl
2-Phenyl-2-triflouromethyl-3-buten-2-ol, (S)-7g (Table 4,
Entry 8). Following the general procedure above, (S)-7g was
obtained as a yellow oil (0.102 g, 94% yield). H NMR (500
1
MHz, CDCl3): δ 2.66 (s, OH), 2.86 (dd, J = 15, 10 MHz, 1H),
2.99 (dd, J = 15, 10 MHz, 1H), 5.24-5.29 (m, 2H), 5.54-5.62
(m, 1H), 7.36-7.43 (m, 3H), 7.59 (d, J = 7.5 MHz). 13C NMR
(125 MHz, CDCl3): δ 40.2, 94.8, 122.0, 126.5, 128.4, 130.4,
136.9, 145.5, 151.4. Enantiomeric excess was determined to be
80% by chiral GC analysis. GC conditions: 121 °C isothermal,
tR for the (R)-alcohol = 20.93 min, and tR for the (S)-alcohol =
21.32 min.
Supporting Information Available: 1H and 13C NMR spec-
tra of the products. This material is available free of charge via
the Internet at http://pubs.acs.org.
J. Org. Chem. Vol. 75, No. 3, 2010 649