Catalytic activation of pinacolyl allylboronate with indium(I): Development of a general catalytic allylboration of ketones

Article abstract of DOI:10.1002/anie.200700899

(Chemical Equation Presented) Have to B In it to win it! The unprecedented catalytic activation of a Group 13 metal reagent (boron) with a Group 13 metal catalyst in its low oxidation state (indium) was employed for the development of a general catalytic

Full text of DOI:10.1002/anie.200700899

Angewandte
Chemie
DOI: 10.1002/anie.200700899
Indium(I) Catalysis
Catalytic Activation of Pinacolyl Allylboronate with Indium(I):
Development of a General Catalytic Allylboration of Ketones**
Uwe Schneider and Sh u¯ Kobayashi*
[
17]
The discovery and development of new catalytic methods for
boron derivatives. On the basis of this concept involving
two Group 13 elements, we hypothesized that indium(I) as a
Lewis base catalyst might activate a Lewis acidic allylboro-
nate through formation of a metal–metal bond, and that the
bimetallic allylborate so generated might enhance nucleophi-
licity towards electrophiles such as ketones. Herein we report
the unprecedented catalytic activation of pinacolyl allylbor-
onate with indium(I) and its application to the general
catalytic allylboration of ketones.
efficient CÀC bond formations is one of the most important
[
1]
tasks in synthetic organic chemistry. In this context, the
[
2]
allylation of ketones is among the most useful and
challenging transformations in organic synthesis, as the
resulting tertiary homoallylic alcohols have proved to be
[
3]
highly versatile intermediates and synthetic building blocks.
Typical protocols for the allylation of ketones involve the use
of allylindium reagents generated in situ under Barbier-type
conditions from allyl halides and a stoichiometric amount of
Our initial experiments were conducted with the reaction
of acetophenone (1a) with pinacolyl allylboronate (2;
1.5 equiv) as model substrates in dry THF at 408C (Table 1).
[
4]
[5]
indium(0) or indium(I). Catalytic methods that have been
reported include various catalytic enantioselective allylstan-
[
6]
nations of ketones employing chiral indium(III) Lewis acids
As the indium source, we employed a stoichiometric amount
[
7]
[19]
and others. However, the substrate scope for ketones is
generally not satisfactory and furthermore these methods
require the use of more than one equivalent of toxic
stannanes, which is undesirable with respect to safety and
environmental considerations. In recent years, significant
advances in ketone allylation have been achieved through the
development of catalytic (asymmetric) allylsilylations by
of commercially available indium(I) iodide,
which was
selected for its higher stability compared with other indium(I)
halides.
Table 1: Examination of various indium catalysts in the allylation of
acetophenone (1a) with pinacolyl allylboronate (2).
[
8]
[9]
[10]
using (chiral) copper(I) or silver(I) Lewis acids. Most
recently, very elegant, catalytic asymmetric allylborations
[
11]
[12]
catalyzed by chiral Lewis or Brønsted acids have also
[
13]
been reported. However, with only a handful of exceptions,
the substrate generality is limited and therefore a truly
general catalytic allylation method for ketones remains to be
developed.
The chemistry of indium in its low oxidation state (I) is
an underexplored area, and only sporadic examples of its use
[a]
Entry
Indium cat. (x [mol%])
Yield [%]
1
2
3
4
5
6
7
8
9
–
trace
quant.
quant.
99
85
88
70 (99)
quant.
4
InI (100)
InI (50)
InI (20)
InI (10)
InI (5)
InI (1)
In (20)+InI (10)
In (20)
[
14]
[
15]
as stoichiometric reagents have been reported. To the best
of our knowledge, a catalytic synthetic method involving the
use of a “catalytic” amount of indium(I) is to date unknown.
Nevertheless, it has been shown that, depending on the nature
of the ligand by which it is coordinated, indium(I) can act as a
[
b]
0
3
0
1
1
12
13
1
0
1
InI (10)
3
3
InCl (20)
InBr (20)
InOTf (20)
65
87
39
16
[
16]
[17]
Lewis acid or Lewis base owing to the presence of both
[
18]
free p orbitals and a lone pair of electrons. Indeed, a recent
report revealed that indium(I) as a s donor is able to form
donor–acceptor complexes with electron acceptors such as
4
In(OTf) (20)
3
[a] Yields after preparative thin-layer or flash chromatography (silica gel).
B(pin): pinacolyl boronate; OTf: trifluoromethanesulfonate. [b] Concen-
tration of 1m in THF under otherwise identical conditions.
[
*] Dr. U. Schneider, Prof. Dr. S. Kobayashi
Graduate School of Pharmaceutical Sciences
The University of Tokyo
The HFRE Division, ERATO, JST
Hongo, Bunkyo-ku, Tokyo 113-0033 (Japan)
Fax: (+81)3-5684-0634
As shown, the noncatalyzed reaction essentially did not
proceed, with only a trace amount of product being formed
even after 24 h (Table 1, entry 1). Gratifyingly, however, the
indium(I)-mediated transformation cleanly afforded the
desired tertiary homoallylic alcohol 3a in quantitative yield
(Table 1, entry 2). Next, we reduced the amount of catalyst to
E-mail: skobayas@mol.f.u-tokyo.ac.jp
[
**] Financial support was provided by ERATO, Japan Science and
Technology Agency (JST). Dr. Masaharu Ueno is acknowledged for
stimulating discussions.
5
0 mol% and then to 20 mol%, which provided, to our
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
delight, the desired product in quantitative and 99% yields,
Angew. Chem. Int. Ed. 2007, 46, 5909 –5912
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5909

Communications
[
20]
respectively (Table 1, entries 3 and 4).
catalytic” results are remarkable because, in general,
stoichiometric amounts of indium(I) reagents are required
These excellent
Table 2: Substrate generality for the indium(I) iodide catalyzed allylation
of ketones 1 with allylboronate 2.
“
[
5]
for indium-mediated Barbier-type and Reformatsky-type
reactions, palladium-to-indium transmetallations, or rad-
ical-generating reactions. Further examination of catalyst
loadings (Table 1, entries 5–7) revealed that use of as little as
mol% of indium(I) gave the desired addition product in up
[
21]
[22]
[
23]
[a]
Entry
Ketone 1
Yield [%]
1
2
3
4
5
6
7
1a: X=H
1b: X=3,5-di-OH
1c: X=2-Br
1d: X=4-NH2
1e: X=4-NO2
1 f: 1-naphthyl
1g: 2-naphthyl
88
quant.
97
99
91
1
to 70% isolated yield after 24 h (use of more concentrated
[
24,25]
conditions gave 99% yield; entry 7).
Various other
solvents such as dichloromethane, N,N-dimethylformamide,
acetonitrile, ethanol, and water were examined but proved to
be significantly less effective than tetrahydrofuran. Next,
several control experiments were carried out (Table 1,
entries 8–10). Indium(I) is known to be rather prone to
redox disproportionation thereby generating indium(0) and
indium(III) in a molar ratio of 2:1. If this is the case in the
current system, indium metal might activate 1a and/or 2
through single electron transfer, whereas indium(III) as a
Lewis acid might activate 1a and/or 2. The combined use of
indium(0) and indium(III) iodide in a molar ratio of 2:1
provided product 3a in quantitative yield (Table 1, entry 8).
On the other hand, use of either indium catalyst independ-
ently proved to be ineffective (Table 1, entries 9 and 10).
These experiments strongly support the idea that indium(I) is
the active catalyst species here and suggest redox synpropor-
tionation (in situ formation of indium(I) from indium(0) and
indium(III)) as the cause of the excellent result in entry 8
quant.
98
8
9
0
1
1h
quant.
1i: R=Et, X=OH
1j: R=Et, X=Cl
1k: R=nPr, X=Cl
98
99
90
1
1
12
1l
99
1
1
1
1
3
4
5
6
1m: n=1, X=5-OMe
1n: n=1, X=6-Me
1o: n=2, X=6-OMe
1p: X=H
81
85
70
90
[
b]
17
1q: X=Ph
72
1
8
1r
1s
1t
70
81
55
(
Table 1). The best catalyst among the indium(I) halides was
shown to be indium(I) iodide, a trend which can be ascribed to
its higher thermodynamic stability compared with indium(I)
chloride and bromide (Table 1, entry 4 vs entries 11 and 12).
As indium(I) iodide has a very low solubility in THF, the more
19
2
0
[
26]
soluble indium(I) triflate was used as a catalyst; however, in
this case, product 3a was obtained in moderate yield (39%;
Table 1, entry 13). Interestingly, this result was still superior to
that observed with indium(III) triflate (16%; Table 1,
entry 14), which is generally accepted to be a better Lewis
acid.
21
22
1u: X=O, 2-furyl
1v: X=S, 2-thienyl
1w: X=S, 3-thienyl
92
98
quant.
23
24
1x
80
Next, we investigated the substrate generality for the
allylation of ketones 1 with allylboronate 2 using 5 mol% of
indium(I) iodide (Table 2). Gratifyingly, it was found that the
reaction displayed good generality and various substituted
aryl methyl ketones including an a,b-unsaturated substrate
were transformed into the corresponding tertiary homoallylic
alcohols in excellent yields (Table 2, entries 1–8). Moreover,
aryl ethyl, aryl propyl, and diaryl ketones proved to be very
good substrates (Table 2, entries 9–12). In addition, various
cyclic aromatic and aliphatic as well as acyclic aliphatic
ketones were shown to undergo smooth allylation (Table 2,
entries 13–19). Note that both a,b-unsaturated substrates
tested underwent exclusive 1,2-addition (Table 2, entries 8
and 18). Finally, a range of heterocyclic ketones were
converted into the corresponding tertiary homoallylic alco-
hols with high yields (Table 2, entries 20–24). Importantly,
several functionalities such as hydroxy, methoxy, amino,
amide, chloro, bromo, and nitro groups were compatible
with the mild conditions of this operationally simple
indium(I)-catalyzed CÀC bond formation.
[
a] Yields after preparative thin-layer or flash chromatography (silicagel).
[b] Mixture of two diastereoisomers (ratio 85:15).
Our attention then turned to the mechanism of this
reaction. As mentioned in our working hypothesis above, one
possible pathway involves catalytic activation of allylboronate
2 with indium(I) as a Lewis base, thereby generating
bimetallic allylborate A with enhanced nucleophilicity
(Figure 1). Alternatively, the iodide-induced formation of an
allylborate of type B might be imagined. In this context, note
that allyltrifluoroborate 4 was found to be significantly more
reactive in the noncatalyzed allylation of ketone 1a than
allylboronate 2 (39% yield vs trace amount, Figure 1; cf.
[
27]
Table 1, entry 1).
To examine the possibility of Lewis base activation of 2,
we used tetrabutylammonium difluorotriphenylsilicate
(TBAT; 20 mol%) or tetrabutylammonium fluoride (TBAF;
100 mol%) as fluoride anion sources in the allylation of 1a
5
910
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5909 –5912

Angewandte
Chemie
concept to other (asymmetric) CÀC bond formations are
ongoing in our laboratories.
Received: February 28, 2007
Published online: June 19, 2007
Keywords: allylation · boron · CÀC coupling · indium · ketones
.
[
1] a) Transition Metal Reagents and Catalysts: Innovations in
Figure 1. Possible intermediates A–D in the indium(I) iodide catalyzed
allylation of ketones 1 with 2.
Organic Synthesis (Ed.: J. Tsuji), Wiley, Chichester, 2000;
b) Catalytic Asymmetric Synthesis (Ed.: I. Ojima), Wiley-VCH,
Weinheim, 2000.
[
2] In contrast to ketones, many allylation methods have been
reported for aldehydes. For recent reviews on the allylation of
aldehydes, see: a) A. Yanagisawa in Comprehensive Asymmetric
Catalysis, Vol. 2 (Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamoto),
Springer, Berlin, 1999, pp. 965 – 979; b) S. E. Denmark, J. Fu,
Chem. Rev. 2003, 103, 2763 – 2794.
with 2. However, these metal-free Lewis base reagents proved
to be only moderately effective (10% and 59% yields,
[
28]
respectively). These results indicate that simple Lewis base
activation of 2 might not be sufficient, and that indium(I)
might additionally act as a Lewis acid activator of 1a (dual-
activation mechanism). On the other hand, simple activation
[3] For selected examples of the use of tertiary homoallylic alcohols,
see: a) J. Barluenga, A. DiØguez, F. Rodríguez, F. J. Faæanµs,
Angew. Chem. 2005, 117, 128 – 130; Angew. Chem. Int. Ed. 2005,
[
29]
of 1a and/or 2 with indium(I) as a Lewis acid (to form
intermediate C) seems unlikely considering our results with
more Lewis acidic indium(III) reagents (much lower yields;
cf. Table 1, entries 10 and 14). Finally, another mechanism
might involve catalytic activation of allylboronate 2 through
boron-to-indium transmetallation to generate allylindium
44, 126 – 128; b) S. Hayashi, K. Hirano, H. Yorimitsu, K. Oshima,
Org. Lett. 2005, 7, 3577 – 3579; c) Y. Takada, S. Hayashi, K.
Hirano, H. Yorimitsu, K. Oshima, Org. Lett. 2006, 8, 2515 – 2517;
d) S. Hayashi, K. Hirano, H. Yorimitsu, K. Oshima, J. Am.
Chem. Soc. 2006, 128, 2210 – 2211; e) P. Zhao, C. D. Incarvito,
J. F. Hartwig, J. Am. Chem. Soc. 2006, 128, 9642 – 9643; f) T.
Ohmura, H. Furukawa, M. Suginome, J. Am. Chem. Soc. 2006,
[
30]
species of type D. To establish whether in situgenerated
allylindium was responsible for the observed reactivity, we
examined allylindium reagents prepared independently under
Barbier-type conditions from allyl bromide or allyl iodide
using a stoichiometric amount of indium metal or indium(I)
iodide. However, under our typical reaction conditions (use of
128, 13366 – 13367.
[
4] For selected examples, see: a) V. Nair, C. N. Jayan, S. Ros,
Tetrahedron 2001, 57, 9453 – 9459; b) J. G. Lee, K. I. Choi, A. N.
Pae, H. Y. Koh, Y. Kang, Y. S. Cho, J. Chem. Soc. Perkin Trans. 1
2002, 1314 – 1317; c) H. Y. Kim, K. I. Choi, A. N. Pae, H. Y. Koh,
J. H. Choi, Y. S. Cho, Synth. Commun. 2003, 33, 1899 – 1904;
d) V. Nair, S. Ros, C. N. Jayan, S. Viji, Synthesis 2003, 2542 – 2546.
5] S. Araki, H. Ito, N. Katsumura, Y. Butsugan, J. Organomet.
Chem. 1989, 369, 291 – 296.
6] For selected examples, see: a) Y.-C. Teo, J.-D. Goh, T.-P. Loh,
Org. Lett. 2005, 7, 2743 – 2745; b) J. Lu, M.-L. Hong, S.-J. Ji, Y.-C.
Teo, T.-P. Loh, Chem. Commun. 2005, 4217 – 4218; c) J. Lu, S.-J.
Ji, Y.-C. Teo, T.-P. Loh, Tetrahedron Lett. 2005, 46, 7435 – 7437;
d) J. Li, S.-J. Ji, Chin. J. Chem. 2006, 24, 1439 – 1442.
1
a, THF, 408C), the desired product 3a was formed only in
[
[
moderate to good yields (60–85%). Whilst it must be
acknowledged that these results cannot completely rule out
a boron-to-indium transmetallation pathway, they indicate
that in situ generated allylindium reagents are unlikely to play
an important role in the indium–boron system, which is the
1
subject of the current presentation. Preliminary H and
1
1
B NMR experiments in [D ]THF at 408C with allylboronate
[7] For selected examples, see: a) H. Hanawa, S. Kii, K. Maruoka,
Adv. Synth. Catal. 2001, 343, 57 – 60; b) A. Cunningham, S.
Woodward, Synthesis 2002, 43 – 44; c) K. M. Waltz, J. Gavenonis,
P. J. Walsh, Angew. Chem. 2002, 114, 3849 – 3851; Angew. Chem.
Int. Ed. 2002, 41, 3697 – 3699; d) S. Kii, K. Maruoka, Chirality
8
2
and indium(I) iodide in the absence of ketone 1a revealed
slow formation of allylindium(I) and allylindium(III) diio-
dide, whereas no evidence for an allylborate species of type A
or B was observed. Furthermore, monitoring of indium(I)
iodide catalyzed allylation of 1a with 2 by NMR spectroscopy
showed that the initially formed product is the allylborated
ketone (OÀB bond), which was transformed into homoallylic
2003, 15, 68 – 70; e) A. Cunningham, V. Mokal-Parekh, C.
Wilson, S. Woodward, Org. Biomol. Chem. 2004, 2, 741 – 748;
f) A. J. Wooten, J. G. Kim, P. J. Walsh, Org. Lett. 2007, 9, 381 –
384.
[
8] Use of a CuCl/TBAT catalyst and allylsilane (racemic version):
S. Yamasaki, K. Fujii, R. Wada, M. Kanai, M. Shibasaki, J. Am.
Chem. Soc. 2002, 124, 6536 – 6537.
alcohol 3a upon hydrolysis.
In conclusion, we have discovered an unprecedented
catalytic activation of pinacolyl allylboronate with
[
9] Use of a AgF/chiral diphosphine catalyst and allylsilane: M.
Wadamoto, H. Yamamoto, J. Am. Chem. Soc. 2005, 127, 14556 –
14557.
[
31]
indium(I) and have applied this novel method successfully
to a general catalytic allylboration of ketones. To the best of
our knowledge, this report represents not only the first
example of a catalytic synthetic method involving the use of a
[10] For a stoichiometric method for enantioselective ketone allyla-
tion using a chiral strained allylsilane, see: N. Z. Burns, B. M.
Hackman, P. Y. Ng, I. A. Powelson, J. L. Leighton, Angew.
Chem. 2006, 118, 3895 – 3897; Angew. Chem. Int. Ed. 2006, 45,
“
catalytic” amount of indium(I) but also the first catalytic
activation of a Group 13 element (boron) with another
member of Group 13 in its low oxidation state (indium).
Further mechanistic studies on the catalytic activation of
allylboronates with indium(I) and the application of this
3811 – 3813.
[
11] Use of a CuF /chiral diphosphine catalyst and allylboronate: R.
2
Wada, K. Oisaki, M. Kanai, M. Shibasaki, J. Am. Chem. Soc.
2004, 126, 8910 – 8911.
Angew. Chem. Int. Ed. 2007, 46, 5909 –5912
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5911

Communications
[
12] Use of a chiral diol catalyst and allylboronate: S. Lou, P. N.
Moquist, S. E. Schaus, J. Am. Chem. Soc. 2006, 128, 12660 –
2661.
THF) proved to be equally effective (90% and 84% yields,
respectively) compared with conditions of entry 6 (Table 1;
1.5 equiv, 5 mol% InI, 0.2m in THF, 88% yield).
1
[
13] For stoichiometric methods for enantioselective ketone allyla-
tion using a chiral allylboronate or allylborane, see: a) T. R. Wu,
L. Shen, J. M. Chong, Org. Lett. 2004, 6, 2701 – 2704; b) E.
Canales, K. G. Prasad, J. A. Soderquist, J. Am. Chem. Soc. 2005,
[25] It is notable that in all indium(I) iodide catalyzed experiments
(Table 1, entries 2–7) no undesired compounds such as pinacol
coupling type or reduction products were detectable in the crude
reaction mixtures.
1
27, 11572 – 11573.
[26] Indium(I) triflate was prepared from commercially available
indium(I) chloride and triflic acid in toluene at room temper-
ature according to a known procedure: C. L. B. Macdonald,
A. M. Corrente, C. G. Andrews, A. Taylor, B. D. Ellis, Chem.
Commun. 2004, 250 – 251.
[
14] For recent reviews on the lower oxidation states of indium, see:
a) D. G. Tuck, Chem. Soc. Rev. 1993, 22, 269 – 276; b) J. A. J.
Pardoe, A. J. Downs, Chem. Rev. 2007, 107, 2 – 45.
[
[
[
[
15] In contrast, indium(III) derivatives are commonly used in
catalytic quantities as Lewis acid catalysts (see Ref. [6]).
16] C. G. Andrews, C. L. B. Macdonald, Angew. Chem. 2005, 117,
[27] On the other hand, use of 20 mol% indium(I) iodide in
combination with trifluoroborate 4 (1.5 equiv) in the allylation
of ketone 1a provided product 3a in only 41% yield.
[28] Additionally, TBAI (tetrabutylammonium iodide; 100 mol%)
was used as a catalyst in an NMR experiment under our typical
7619 – 7622; Angew. Chem. Int. Ed. 2005, 44, 7453 – 7456.
17] R. J. Wright, A. D. Phillips, N. J. Hardman, P. P. Power, J. Am.
Chem. Soc. 2002, 124, 8538 – 8539.
18] For a recent highlight on this topic, see: S. Aldridge, Angew.
Chem. 2006, 118, 8275 – 8277; Angew. Chem. Int. Ed. 2006, 45,
conditions (use of 2, [D ]THF, 408C); however, conversion of
8
ketone 1a into product 3a was not observed.
8
097 – 8099.
[29] Lewis acid activation of allylboronates of type 2 through
coordination of a boronate alkoxy ligand to the corresponding
Lewis acid has been reported. However, this catalytic method
has been applied only to the use of aldehydes as electrophiles:
a) J. W. J. Kennedy, D. G. Hall, J. Am. Chem. Soc. 2002, 124,
11586 – 11587; b) for a recent minireview on the activation of
boron reagents, see: J. W. J. Kennedy, D. G. Hall, Angew. Chem.
2003, 115, 4880 – 4887; Angew. Chem. Int. Ed. 2003, 42, 4732 –
4739; c) M. Gravel, H. Lachance, X. Lu, D. G. Hall, Synthesis
2004, 1290 – 1302; d) V. Rauniyar, D. G. Hall, J. Am. Chem. Soc.
2004, 126, 4518 – 4519.
[
[
19] Indium(I) iodide is commercialized as 10-mesh beads, but the
powdered form gave identical results.
20] The use of other allylboronates under the employed conditions
proved to be less efficient than the use of pinacolyl allylboronate
(
2).
21] S. A. Babu, M. Yasuda, I. Shibata, A. Baba, Org. Lett. 2004, 6,
475 – 4478.
[
[
4
22] a) S. Araki, T. Kamei, T. Hirashita, H. Yamamura, M. Kawai,
Org. Lett. 2000, 2, 847 – 849; b) I. R. Cooper, R. Grigg, W. S.
MacLachlan, V. Sridharan, M. Thornton-Pett, Tetrahedron Lett.
2
003, 44, 403 – 405; c) H. Miyabe, Y. Yamaoka, T. Naito, Y.
[30] For electrochemical in situformation and regeneration of
allylindium(I), see: G. Hilt, K. I. Smolko, Angew. Chem. 2001,
113, 3514 – 3516; Angew. Chem. Int. Ed. 2001, 40, 3399 – 3402.
[31] Indium is known as a “rare metal”, thus catalytic use of indium is
highly important: The Elements (Ed.: J. Emsley), 3rd ed., Oxford
Press, Oxford, 1998.
Takemoto, J. Org. Chem. 2003, 68, 6745 – 6751; d) H. Miyabe, Y.
Yamaoka, T. Naito, Y. Takemoto, J. Org. Chem. 2004, 69, 1415 –
1
418; e) G. Fontana, A. Lubineau, M.-C. Scherrmann, Org.
Biomol. Chem. 2005, 3, 1375 – 1380.
[
[
23] B. C. Ranu, T. Mandal, Tetrahedron Lett. 2006, 47, 2859 – 2861.
24] The use of a smaller amount of allylboronate 2 (1.1 equiv,
20 mol% InI, 0.2m in THF; or 1.0 equiv, 5 mol% InI, 0.5m in
5
912
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 5909 –5912