Indium-promoted chemo- and diastereoselective allylation of α,β-epoxy ketones with potassium allyltrifluoroborate

Article abstract of DOI:10.1021/ol1023757

A practical method for the chemo- and diastereoselective allylation of α,β-epoxy ketones has been developed by using the convenient air and moisture stable reagent potassium allyltrifluoroborate. Indium metal was found to promote addition in stoichiometric or catalytic amounts, to afford α,β-epoxyhomoallylic tertiary alcohols in high yields and diastereoselectivities, without competing ring-scission of the epoxide.

Full text of DOI:10.1021/ol1023757

ORGANIC
LETTERS
2010
Vol. 12, No. 23
5490-5493
Indium-Promoted Chemo- and
Diastereoselective Allylation of
r,ꢀ-Epoxy Ketones with Potassium
Allyltrifluoroborate
Farhad Nowrouzi, John Janetzko, and Robert A. Batey*
DaVenport Research Laboratories, Department of Chemistry, UniVersity of Toronto, 80
St. George Street, Toronto, Ontario, Canada, M5S 3H6
rbatey@chem.utoronto.ca
Received October 1, 2010
ABSTRACT
A practical method for the chemo- and diastereoselective allylation of r,ꢀ-epoxy ketones has been developed by using the convenient air and
moisture stable reagent potassium allyltrifluoroborate. Indium metal was found to promote addition in stoichiometric or catalytic amounts, to
afford r,ꢀ-epoxyhomoallylic tertiary alcohols in high yields and diastereoselectivities, without competing ring-scission of the epoxide.
The formation of C-C bonds via the allylation of carbonyl
compounds is an important method for the synthesis of
homoallylic alcohols1 and has been widely applied in the
synthesis of natural products and other complex targets.2
Classical allylation methods include the in situ reactions
of allylic halides with stoichiometric metals (e.g., Mg, Cr,
Zn, In) under Barbier-type3 conditions (type III1d re-
agents), or the use of allyl or crotylmetalloids.4,5 The
addition of type II1d allyl or crotylmetalloids (e.g., Sn, Si)
to carbonyl groups requires Lewis acid activation, while type
I1d reagents, exemplified by allylboranes or boronates,
undergo direct addition. Generally, additions to ketones are
slower than for aldehydes thus requiring more vigorous
conditions.
Given the well-known synthetic versatility of epoxy
alcohols,6 there has been surprisingly little attention devoted
to the addition of organometallic or metalloid reagents to
epoxy ketones.7 Nucleophilic additions to epoxy ketones
(1) For reviews on allylations, see: (a) Roush, W. R. In ComprehensiVe
Organic Synthesis; Trost, B. M., Fleming, I., Heathcock, C. H., Eds.;
Pergamon: Oxford, UK, 1991; Vol. 2, pp 1-53. (b) Chemler, S. R.; Roush,
W. R. In Modern Carbonyl Chemistry; Otera, J., Ed.; Wiley-VCH:
Weinheim, Germany, 2000; Chapter 10. (c) Kennedy, J. W. J.; Hall, D. G.
Angew. Chem., Int. Ed. 2003, 42, 4732–4739. (d) Denmark, S. E.; Fu, J. P.
Chem. ReV. 2003, 103, 2763–2794. (e) Lachance, H.; Hall, D. G.; Organic
Reactions; Denmark, S. E., Ed.; John Wiley & Sons Inc.: New York, 2008;
Vol. 73 and references cited therein.
(4) For selected examples using allylsilanes, see: (a) Wadamoto, M.;
Yamamoto, H. J. Am. Chem. Soc. 2005, 127, 14556–14557. Using
allylstannanes, see: (b) Waltz, K. M.; Gavenonis, J.; Walsh, P. J. Angew.
Chem., Int. Ed. 2002, 41, 3697–3699. (c) Lu, J.; Hong, M.-L.; Ji, S.-J.;
Teo, Y.-C.; Loh, T. P. Chem. Commun. 2005, 4217–4218. (d) Gu, Y.;
Ogawa, C.; Kobayashi, J.; Mori, Y.; Kobayashi, S. Angew. Chem., Int. Ed.
2006, 45, 7217–7220. (e) Wooten, A. J.; Kim, J. G.; Walsh, P. J. Org. Lett.
2007, 9, 381–384. Using allylboronates, see: (f) Lou, S.; Moquist, P. N.;
Schaus, S. E. J. Am. Chem. Soc. 2006, 128, 12660–12661.
(2) See for example: (a) Nicolaou, K. C.; Kim, D. W.; Baati, R. Angew.
Chem., Int. Ed. 2002, 41, 3701–3704. (b) Felpin, F.-X.; Lebreton, J. J. Org.
Chem. 2002, 67, 9192–9199. (c) Hornberger, K. R.; Hamblett, C. L.;
Leighton, J. L. J. Am. Chem. Soc. 2000, 122, 12894–12895.
(5) See for example: (a) Rauniyar, V.; Zhai, H.; Hall, D. G. J. Am. Chem.
Soc. 2008, 130, 8481–8490. (b) Wu, T. R.; Shen, L.; Chong, J. M. Org.
Lett. 2004, 6, 2701–2704. (c) Lachance, H.; Lu, X.; Gravel, M.; Hall, D. G.
J. Am. Chem. Soc. 2003, 125, 10160–10161.
(3) For selected recent examples of allylations of ketones under Barbier-
like conditions, see: (a) Zhao, L.; Burnell, D. J. Tetrahedron Lett. 2006,
47, 3291–3294. (b) Miller, J. J.; Sigman, M. S. J. Am. Chem. Soc. 2007,
129, 2752–2753.
(6) Riera, A.; Moreno, M. Molecules 2010, 15, 1041–1073.
10.1021/ol1023757  2010 American Chemical Society
Published on Web 11/11/2010

constitute a problem for many of the aforementioned
allylation conditions, since epoxide ring-opening occurs in
the presence of more nucleophilic reagents or Lewis acids,8,9
and deoxygenation in the presence of reducing metals.
Reaction of R,ꢀ-epoxy ketones has been achieved diaste-
reoselectively with allylstannanes using BF3·OEt210a or PbI2
activation,10b or with Grignard reagents.10c-e Unfortunately,
these methods utilize toxic reagents, show limited substrate
scope, and occur with diminished yields for cyclic sub-
strates.10 An alternative one-pot sequential asymmetric
allylation/directed epoxidation affords syn-epoxy homoallylic
alcohols; however, it only works well for cyclic enones, with
R-substitution necessary to obtain high selectivity.11 We now
report a general and selective organoboron-based method12
for the allylation of R,ꢀ-epoxy ketones using indium metal
as a catalyst,13 and demonstrate how the stereoselectivity of
addition depends upon substrate class.
Table 1. Allylation of R,ꢀ-Epoxycyclohexanone (1a) with
Potassium Allyltrifluoroborate
entry
additive (equiv)
solvent (mL)
yielda (dr)b
1
2
3
4
5
6
7
Mont. K10 (0.1 g)d CH2Cl2/H2O (1.4:0.1)
62 (98:2)
64 (98:2)
23 (82:18)
48 (91:9)
43 (98:2)
72 (76:24)
77 (93:7)
BF3·OEt2 (0.1)e
In(OTf)3 (0.1)
Cu(OTf)2 (0.1)
B(OiPr)3 (0.1)
B(OiPr)3 (1.0)
In (1.0)
CH2Cl2 (1.5)
CH2Cl2 (1.5)
CH2Cl2 (1.5)
CH2Cl2 (1.5)
CH2Cl2 (1.5)
CH2Cl2/H2O (1.4:0.1)
8
9
In (1.0)
In (1.0)
In (1.0)
In (1.0)
In (1.0)
In (0.1)
In (1.0)
CH2Cl2/H2O (1.45:0.05) 83 (95:5)
CH2Cl2 (1.5) N.R.
CH2Cl2/MeOH (1.4:0.1) N.R.
10
11
12
13
14
Organotrifluoroborate salts14 are well established as air
and moisture stable reagents, acting as synthetic equivalents
to boronic acids.15 Potassium allyl and crotyltrifluoroborate
salts undergo addition reactions to carbonyl derivatives under
Lewis acidic, phase-transfer catalyzed, or Montmorillonite
K10 promoted conditions.16 Initial attempts to apply previ-
ously developed protocols for in situ activation of potassium
allyltrifluoroborate using the model epoxy ketone 1a gave
only moderate yields of epoxy alcohol 2a due to competing
side reactions of 1a (Table 1, entries 1 and 2). Alternative
conditions using a variety of Lewis acids, fluorophiles, and
MeCN/H2O (1.45:0.05)
THF/H2O (1.45:0.05)
33 (91:9)
76 (94:6)
CH2Cl2/H2O (1.45:0.05) 41 (95:5)
CH2Cl2/H2O (1.45:0.05) 73c (92:8)
a Yield of product isolated after silica gel chromatography. b dr was
determined by the integration of the R-epoxide C-H signal in the 1H NMR
of the crude reaction mixture. c Allylboronic acid pinacol ester (2.0 equiv)
was used. d Mont. K10 is Montmorillonite K10. e Use of 1.0 equiv of
BF3·OEt2 led to decomposition.
solid additives were examined.17 Several Lewis acids were
capable of promoting addition to 1a, but product yields were
disappointing (Table 1, entries 3-5). While the use of 0.1
equiv of B(OiPr)3 led to poor product yields, the use of a
stoichiometric quantity led to an erosion of diastereoselec-
tivity of 2a (Table 1, entries 5 and 6). Intriguingly, metallic
indium (1.0 equiv) was an effective promoter for the
reaction,18,19 giving a good yield and dr of 2a without leading
to side-product formation (Table 1, entry 7). Further opti-
mization revealed that reducing the amount of water im-
proves yields of 2a, but that poor reaction conversions were
obtained in the absence of water (ca. 5%) (Table 1, entries
8 and 9). Reaction of 1a with MeOH as a protic additive
was unsuccessful, while the use of other solvents such as
acetonitrile or THF resulted in lower yields (Table 1, entries
10-12). It is noteworthy that while potassium allyltrifluo-
roborate is stable in water, showing less than 5% decomposi-
tion over 24 h in the absence of the ketone, the presence of
indium leads to formation of propene gas,20 necessitating
the use of 2.0 equiv to achieve full conversion of 1a. The
use of In in catalytic quantities was successful although a
lower yield of 2a was obtained with a similar reaction time
(Table 1, entry 13). Reaction in the absence of In using water/
CH2Cl2 is ineffective. Stoichiometric quantities of In were
therefore used for subsequent studies. Use of the pinacol ester
(7) Lauret, C. Tetrahedron: Asymmetry 2001, 12, 2359–2383.
(8) (a) Invernize, P. R.; da Silva Filho, L. C.; da Silva, G. V. J.; Ju´nior,
V. L.; Constantino, M. G. Synth. Commun. 2007, 37, 3529–3539. (b) Suzuki,
T.; Saimoto, H.; Tomioka, H.; Oshima, K.; Nozaki, H. Tetrahedron Lett.
1982, 23, 3597–3600. (c) Chong, J. M.; Sharpless, K. B. J. Org. Chem.
1985, 50, 1560–1563
.
(9) For nucleophilic ring opening of epoxides see: (a) Schneider, C.
Synthesis 2006, 3919–3944. (b) Pineschi, M. Eur. J. Org. Chem. 2006,
4979–4988. (c) Nielsen, L. P. C.; Jacobsen, E. N. Aziridines and Epoxides
in Organic Synthesis; Yudin, A. K. ; Wiley-VCH Verlag GmbH & Co.:
Weinheim, Germany, 2006; pp 229-269 and references cited therein.
(10) (a) Maruyama, K.; Naruta, Y. Chem. Lett. 1978, 431–432. (b)
Shibata, I.; Fukuoka, S.; Yoshimura, N.; Matsuda, H.; Baba, A. J. Org.
Chem. 1997, 62, 3790–3791. (c) Runie, K. A.; Taylor, R. J. K. Org. Lett.
2001, 3, 3237–3239. (d) Dechoux, L.; Agami, C.; Doris, E.; Mioskowski,
C. Tetrahedron 2003, 59, 9701–9706. (e) Tanaka, T.; Hiramatsu, K.;
Kobayashi, Y.; Ohno, H. Tetrahedron 2005, 61, 6726–6742.
(11) Kim, J. G.; Waltz, K. M.; Garcia, I. F.; Kwiatkowski, D.; Walsh,
P. J. J. Am. Chem. Soc. 2004, 126, 12580–12585.
(12) For additions to R,ꢀ-epoxyaldehydes using chiral allylboron
reagents, see for example: (a) Roush, W. R.; Straub, J. A.; VanNieuwenhze,
M. S. J. Org. Chem. 1991, 56, 1636–1648. (b) Murata, T.; Sano, M.;
Takamura, H.; Kadota, I.; Uemura, D. J. Org. Chem. 2009, 74, 4797–4803.
(13) For reviews of the use of indium metal and its salts in organic
synthesis, see: (a) Yadav, J. S.; Antony, A.; George, J.; Subba Reddy, B. V.
Eur. J. Org. Chem. 2010, 591–605. (b) Auge, J.; Lubin-Germain, N.; Uziel,
J. Synthesis 2007, 1739–1764. (c) Nair, V.; Ros, S.; Jayan, C. N.; Pillai,
B. S. Tetrahedron 2004, 60, 1959–1982.
(14) (a) Molander, G. A.; Ellis, N. Acc. Chem. Res. 2007, 40, 275–286.
(b) Stefani, H. A.; Cella, R.; Vieira, A. S. Tetrahedron 2007, 63, 3623–
3658. (c) Darses, S.; Geneˆt, J.-P. Chem. ReV. 2008, 108, 288–325.
(15) Vedejs, E.; Fields, S. C.; Shrimpf, M. R. J. Am. Chem. Soc. 1993,
117, 11612–11613.
(17) Attempts to activate potassium allyltrifluoroborate with a variety
of fluorophiles such as TBDPSCl, TBSCl, TIPSOTf, and HMDS afforded
only a complex mixture of products.
(16) (a) Batey, R. A.; Thadani, A. N.; Smil, D. V.; Lough, A. J. Synthesis
2000, 7, 990–998. (b) Thadani, A. N.; Batey, R. A. Org. Lett. 2002, 4,
3827–3830. (c) Thadani, A. N.; Batey, R. A. Tetrahedron Lett. 2003, 44,
8051–8055. (d) Li, S. W.; Batey, R. A. Chem. Commun. 2004, 1382–1383.
(e) Nowrouzi, F.; Thadani, A. N.; Batey, R. A. Org. Lett. 2009, 11, 2631–
2634.
(18) Thadani, A. N., Ph.D. Dissertation, University of Toronto, Toronto,
ON, Canada, 2001
.
(19) The use of metals such as Sn, Zn, Mg, and Ag was unsuccessful.
(20) See the Supporting Information for details.
Org. Lett., Vol. 12, No. 23, 2010
5491

Table 2. Stereoselective Synthesis of Epoxy Alcohols 2
a Isolated yield of the major diastereomer after silica gel chromatography, except for entries 10, 15, and 16 which were isolated as inseparable mixtures
of diastereomers. b dr values were determined by 1H NMR integration of the R-epoxide C-H’s in the crude reaction mixture and/or GC analysis. c Reaction
performed on 6.0 mmol of substrate scale. d 2.5 equiv of potassium allyltrifluoroborate was used to ensure full conversion. e The starting material was used
as a mixture of epoxide diastereomers in a 93:7 ratio favoring the shown compound. The corresponding product was isolated in an identical ratio. f The
diastereomeric product was obtained when Montmorillonite K10 is used instead of metallic indium. g The starting material was used as a mixture of enantiomers
(er ) 93:7, as determined by HPLC). The respective product was isolated with the same optical purity as determined by chiral HPLC.
of allylboronic acid under otherwise identical conditions
failed to achieve full conversion to the desired product (Table
1, entry 14).21 Direct application of the protocol reported
by Kobayashi occurred with diminished yield and selectivity
(54%, crude dr 86:14).20
The optimized conditions were then applied to a range of
R,ꢀ-epoxy ketones to establish the generality of the protocol.
In most cases examined the products 2 were obtained in
excellent yields and good diastereoselectivity (Table 2). The
reaction could be readily scaled with negligible effect on
yield and dr (Table 2, entry 1). R-, ꢀ-, and γ-substitution on
the epoxy ketones 1 was well tolerated (Table 2, entries
2-7). The steric constraints of the R-methyl group in 1c
did not result in a loss in dr,4e,11 but did necessitate a modest
increase in the amount of allyltrifluoroborate salt required
to achieve full conversion within 24 h (Table 2, entry 3).
Chemoselective allylation of the less sterically encumbered
carbonyl group of 1f occurred (Table 2, entry 6). The mild
nature of the protocol is apparent from the good functional
group tolerance achieved, such as with hydroxyl and acid
labile dimethyl ketal groups previously shown to result in
several side products using other organometallic reagents
(Table 2, entries 7 and 8).22 Reaction of seven- and five-
membered ring R,ꢀ-epoxy ketones 1i and 1j occurred in good
yields (Table 2, entries 9 and 10), although the dr of 2j was
modest, favoring the cis-stereoisomer. For additions to cyclic
six-membered ring R,ꢀ-epoxy ketones, the observed diaste-
reoselectivity can be rationalized through attack of the
nucleophile via a pseudoequatorial trajectory syn to the
epoxide ring, matching well with conformational analy-
(21) Schneider, U.; Ueno, M.; Kobayashi, S. J. Am. Chem. Soc. 2008,
130, 13824–13825.
(22) McKillop, A.; Taylor, R. J. K.; Watson, R. J.; Lewis, N. J. Chem.
Soc., Chem. Commun. 1992, 1589–1591.
5492
Org. Lett., Vol. 12, No. 23, 2010

ses,23,24 and steric and electronic factors previously exam-
ined.25 The higher dr values observed for the formation of
2d and 2f are also consistent with this notion as well as
reactions of the more conformationally rigid steroidal system
1g. The reactions of 1h and the cyclopentanone oxide 1j
occur through addition anti to the epoxide ring, presumably
as a result of steric factors.22 Acyclic R,ꢀ-epoxy ketones
could also be allylated by using the In protocol. Variation
of the ꢀ-substituent in 1k-m was found to have little effect
upon the reaction (Table 2, entries 11-13). Interestingly,
while the use of the Montmorillonite K10 promoted
conditions16e with epoxy ketones generally occurred with
lower yields, a reversal of diastereoselectivity was observed
for the reaction of 1m (Table 2, entry 14). This result
suggests that reaction occurs via a different allylation species
under the In promoted conditions than the tricoordinate
allylboron species that is postulated for the Montmorillonite
K10 based protocol. Reaction via an allylindium species26,27
in a chelation-controlled manner may account for the
selectivity difference, since chelation control does not for
the direct addition of tricoordinate allylboron species. An
increase in dr was observed as the steric bulk of the carbonyl
substituent R2 on 1k-q increased (Table 2, entries 11 and
15-18). Reaction of enantiomerically enriched 1k28 occurred
without loss of stereochemical fidelity (Table 2, entry 19).
It is noteworthy that although a stoichiometric amount of
indium was employed, it can be recovered and reused for
further reactions.20
Table 3. Conversion of Compounds 2 to Substituted anti-Allylic
Epoxy Alcohols 3
entry
substrate
product
yield (%)a
1
2
3
4
5
2a
2c
2i
2l
2m
3a
3c
3i
3l
3m
90
89b
85
88
94
a Isolated yield of the major compound after silica gel chromatography.
b The starting material was used as a mixture of epoxide diastereomers in
a 93:7 ratio favoring the shown compound. The corresponding product was
isolated in an identical ratio.
Lewis acidic conditions to diastereoselectively afford the
aldol-like ketone 4 (Scheme 1). Conventional aldol conditions
Scheme 1. Semi-Pinacol Rearrangement of Alcohol 2l
The epoxy alcohols are useful intermediates for further
transformations.6 For example, Payne rearrangement29 of 2
afforded the more substituted epoxides 3 in excellent yields
(Table 3). This also served to verify the anti-selectivity
observed between the hydroxyl and oxirane moiety in both
acyclic and cyclic substrates.30 Overall, the two-step
allylation-Payne rearrangement sequence allows for access
to various structurally diverse anti-allylic epoxy alcohols
which were, prior to this work, only accessible through
limited means.31 Another useful demonstration of the utility
of 2 is the semi-pinacol-type rearrangement32 of 2l under
would not be suitable for the synthesis of 4 due to the base
lability of the ꢀ,γ-unsaturated ketone functionality.
In summary, an operationally straightforward procedure
for the allylation of R,ꢀ-epoxy ketones using the stable
potassium allyltrifluoroborate salt and In metal catalysis has
been developed. The protocol shows good functional group
tolerance, is readily scalable, and avoids the use of toxic
reagents which have been required for the synthesis of 2.10
The mild protocol solves the problems of chemoselective
addition to R,ꢀ-epoxy ketones with reaction occurring in a
highly diastereoselective fashion. Further studies on related
diastereoselective transformations of 1-3 will be reported
in due course.
(23) Cherest, M. Tetrahedron 1980, 36, 1593–1598
(24) Sepu´lveda, J.; Soriano, C.; Roquet-Jalmar, J.; Mestres, R.; Riego,
J. Bull. Soc. Chim. Fr. 1987, 1, 189–192
.
.
(25) Aljarilla, A.; Co´rdoba, R.; Csaky, A. G.; Ferna´ndez, I.; Ortiz, F. L.;
Plumet, J.; Go´mez, G. R. Eur. J. Org. Chem. 2006, 3969–3976.
(26) The reaction does not, however, appear to proceed via a classical
allylindium(III) species, since reaction of 1a with allylindium (prepared in
situ from allyl bromide and In: see the Supporting Information) resulted in
a complex mixture of products. (a) Chan, T. H.; Yang, Y. J. Am. Chem.
Soc. 1999, 121, 3228–3229. (b) Koszinowski, K. J. Am. Chem. Soc. 2010,
Acknowledgment. The Natural Science and Engineering
Research Council (NSERC) of Canada funded this research.
We also thank Dr. Alex Young (University of Toronto) for
MS analysis and Dr. Timothy Burrow (University of Tor-
onto) for NMR assistance.
132, 6032–6040
.
(27) Kobayashi has also proposed the intermediacy of allylindium
species, see ref 21.
(28) See the Supporting Information for the synthesis of (-)-1k.
(29) (a) Hanson, R. M. Organic Reactions; Overman, L. E., Ed.; John
Wiley & Sons Inc.: New York, 2002; Vol. 60 and references cited therein.
(b) Payne, G. B. J. Org. Chem. 1962, 27, 3819–3822.
Supporting Information Available: Full experimental
procedures, analytical data, and spectra for all compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(30) Treatment of the crude mixture of diastereomers 2j, under the same
conditions, led only to Payne rearrangement of the minor isomer, revealing
that 2j was formed preferentially as the syn-diastereoisomer.
(31) Svenda, J.; Myers, A. G. Org. Lett. 2009, 11, 2437–2440.
(32) Snape, T. J. Chem. Soc. ReV. 2007, 36, 1823–1842.
OL1023757
Org. Lett., Vol. 12, No. 23, 2010
5493