Bimetallic reagents of silicon: One-pot synthesis of 2,3,5-trisubstituted tetrahydrofurans by a double Sakurai-Hosomi reaction

Article abstract of DOI:10.1002/anie.200353184

Good yields and high stereoselectivities are found in the cyclization of 1-silylmethyl allylic silanes with aldehydes. The silanes act as bimetallic coupling reagents and effect cyclization to yield a variety of 2,3,5-trisubstituted tetrahydrofurans in a

Full text of DOI:10.1002/anie.200353184

Angewandte
Chemie
Tetrahydrofuran Synthesis
Bimetallic Reagents of Silicon: One-Pot Synthesis
of 2,3,5-Trisubstituted Tetrahydrofurans by a
Double Sakurai–Hosomi Reaction**
Tarun K. Sarkar,* Sk. Anwarul Haque, and
Arindrajit Basak
Unsaturated bimetallic reagents containing Group 14 metals,
such as silicon and tin, have recently emerged as versatile
coupling reagents in organic synthesis.[1,2] These reagents are
particularly appealing when both metals employ a “ferryman”
service to yield metal-free cyclic products in a one-pot
[2]
reaction. However, examples ofsuch reactions are rare.
As part ofour interest in allyl organometallic compounds of
silicon,[3] we envisioned that 1-silylmethyl allylic silanes 1
could be useful synthetic building blocks. In the
presence ofa Lewis acid, a bis-silyl reagent such
as 1 affords a b-silyl carbocation 2 (Scheme 1)
when treated with an aldehyde. This carbocation
may then collapse to give an oxetane 3 (path a)[4]
Scheme 1. Reaction of 1 with an aldehyde. Si=silyl group.
or, more likely, undergo a 1,2-silyl migration to
give the thermodynamically favored bis-b-silyl
that silylmethyl allylic silanes oftype 1 can be channeled
through path c to yield only 2,3,5-trisubstituted tetrahydro-
furans.
As the 1-silylmethyl allylic silanes 1a,b (Scheme 2)
selected for our study are novel, there were no reported
procedures for their synthesis. Smitrovich and Woerpel
carbocation 4,[5] which produces a trisubstituted
tetrahydrofuran 5 (path b).[6] The latter was realized recently
by Peng and Woerpel.[7] Additionally, a third pathway is also
conceivable wherein 2 collapses to give a new allylic silane 6.
The allylic silane 6 should be less reactive than the starting
allylic silane 1 because its p bond is more hindered. Intro-
duction ofa second aldehyde should yield a silicon-rfee
tetrahydrofuran 8 by trapping ofthe oxocarbenium ion
7
(path c). While our studies were in progress Smitrovich and
Woerpel[8] reported a single experiment on the synthesis ofa
tetrahydrofuran lacking silyl groups by reaction of a sub-
stituted silylmethyl allylic silane and an aldehyde; however, a
pure product could not be obtained and a furan structure was
only tentatively assigned. Herein, we provide the first report
[*] Prof. T. K. Sarkar, S. A. Haque
Department of Chemistry
Indian Institute of Technology
Kharagpur-721302 (India)
Fax: (+91)3222-282252-755303
E-mail: tksr@chem.iitkgp.ernet.in
Dr. A. Basak
Scheme 2. Reagents and conditions: a) Me3CCN (8 mol%), Pd(OAc)2
(2 mol%), toluene, reflux, 80% yield; b) MeLi, Et2O, 75% yield;
c) PhLi, Et2O, 80% yield; d) 1. o-NO2C6H4SeCN, nBu3P, THF; 2. H2O2
(30%), pyridine, CH2Cl2, 60–68% yield over two steps.
Department of Chemistry
The University of Chicago
Chicago, IL 60637 (USA)
[**] This work has been financially supported by the DST, Govt. of India.
We are grateful to Prof. H. Yamamoto (Chicago, IL, USA),
Dr. C. Fehr, Dr. R. Brauchli, Dr. L. Wunsche (Firmenich, Geneva,
Switzerland), and Prof. S. Uemura (Kyoto, Japan) for help with NMR
spectroscopic and mass spectrometric analysis. Prof. M. Suginome
(Kyoto, Japan) and Dr. S. Roy (IITKGP) are thanked for their helpful
discussions. S.A.H. thanks the CSIR (India) for a research fellow-
ship.
experienced great difficulties in preparing similar compounds
but eventually succeeded by employing allylic displacement,[9]
which unfortunately could not be applied in our case since
only carbamates from secondary allyl alcohols are amenable
to it. Thus, a convenient synthetic methodology to this family
ofallylic silanes was our initial task. Our successufl route
(Scheme 2) involved heating disilanyl ether 9 (prepared in
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2004, 43, 1417 –1419
DOI: 10.1002/anie.200353184
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1417

Communications
Table 1: Isolated yields for the reaction of allylic silane 1b with aldehydes.
77% yield from 3-butene-1-ol and
1-chloro-1,2,2-trimethyl-1,2-diphe-
nyldisilane) in the presence of
Pd(OAc)2 (2 mol%) and tert-butyl-
isocyanide (8 mol%) under reflux
in toluene. Intramolecular bis-sily-
lation occurred readily to give the
1,2-oxasilolane 10 in good yield
(80%).[10] Ring opening of 10 with
either methyl- or phenyllithium in
diethyl ether followed by Grieco
dehydration[11] ofthe resulting
alcohols 11 and 12 gave the desired
allylic silanes 1a,b in 45–55% over-
all yield from 10.
Entry
Aldehyde
Reaction conditions[a]
Products[b,c] (% yield)
1
BF3·OEt2 ꢀ788C!RT (2 h), then RT (3 h)
2
3
SnCl4 ꢀ788C!RT (2 h), then RT (1 h)
TiCl4 ꢀ788C!RT (2 h), then RT (1 h)
(41%)
(0%)
4
5
BF3·OEt2 ꢀ788C!RT (2 h), then RT (7 h)
Our initial idea for investigat-
ing the concept outlined in
Scheme 1 was to treat the allylic
silanes 1a,b with an aldehyde con-
BF3·OEt2 ꢀ788C!RT (2 h), then RT (6 h)
taining
a proximal heteroatom,
6
BF3·OEt2 ꢀ788C!RT (2 h), then RT (10 h)
such as oxygen, to allow the reac-
tion to be performed under both
chelation and nonchelation condi-
tions. The choice ofthe aldehyde
and reaction conditions were
essential for the successful synthe-
sis oftetrahydroufrans devoid of
silicon appendages. Thus, in the
[a] All reactions were performed in CH2Cl2. [b] Yield of isolated pure material. [c] The product in entry 5 is
a mixture of three diastereomers (3:1:1) with unknown configuration at the starred carbon atoms. LA=
Lewis acid, RT=room temperature, Bn=benzyl.
presence ofBF ·OEt2,[12] 1a was treated with (tert-butyldi-
compatibility ofa
b-alkoxy aldehyde to this reaction.
3
phenylsilyloxy)ethanal to give tetrahydrofuran 13 (Scheme 3)
admixed with the readily separable but undesired monop-
rotected glycol 14 (2:3, respectively). The latter was evidently
Remarkably, 17 was obtained from 2-(benzyloxy)propanal
as a mixture ofthree diastereomers (3:1:1, entry 5) rfom
which the major one could be readily isolated by silica gel
chromatography; however, the configuration at the extracy-
clic centers in this product remains undetermined at present.
Incidentally, the reaction failed to give any tetrahydrofuran
when the aldehyde is hindered (entry 6), although neither the
aldehyde nor the starting allylic silane could be recovered.[15]
Finally, in our most dramatic result, three-component
cyclizations were observed. For example, BF3·OEt2 promoted
the coupling of 1b and two different aldehydes (Scheme 4)
with the formation of a single product (19) in 48% yield
uncontaminated by any regio- or stereoisomers. This synthesis
was only possible because of the difference in reactivity
between 1b and the allylic silane formed in situ (cf. 6) and
between that ofthe two aldehydes. [16]
Scheme 3. The reaction of (tert-butyldiphenylsilyloxy)ethanal with allylic
silane 1a. TBDPSO=tert-butyldiphenylsilyloxy.
formed by protodesilylation of the in situ generated allylic
silane (cf. 6). Use ofan additive, namely 2,6-di- tert-butyl-4-
methylpyridine,[13] to remove traces ofacid that would cause
unwanted protodesilylation was ineffective. However, forma-
tion of 14 could be prevented by use ofBF ·OEt2 freshly
3
distilled over calcium hydride to give only 13 in 50% yield.
Additionally, 1b was found to give 13 in a slightly higher yield
(ca. 10%) under these conditions than did 1a. Tetrahydro-
furan 13 was formed as a single diastereomer, and the
stereochemical outcome ofthe reaction is in accord with
observations reported by Mohr.[14]
Encouraged by these findings, we investigated the scope
and limitations ofthe one-pot tetrahydrofuran synthesis—the
results ofwhich are given in Table 1. BF ·OEt2 was found to
3
be the most effective Lewis acid (entry 1) and not SnCl4
(entry 2), while TiCl4 (entry 3) did not yield any traces of
tetrahydrofuran 15. The example in entry 4 shows the
Scheme 4. Three-component reaction involving allylic silane 1b with
two different aldehydes. TBDPSO=tert-butyldiphenylsilyloxy, Bn=
benzyl.
1418
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 1417 –1419

Angewandte
Chemie
In summary, 1-silylmethyl allylic silanes have been
developed as new bimetallic reagents for rapid access to a
variety of2,3,5-trisubstituted tetrahydrofurans in good yields
and with high stereoselectivities. The synthesis allows two
different groups to be introduced at the 2 and 5 positions of
the furan ring, which should enhance the utility of tetrahy-
drofuran synthesis and help promote its application in organic
synthesis.
Received: October 29, 2003
Revised: December 8, 2003 [Z53184]
Keywords: cyclization · domino reactions · oxygen heterocycles ·
.
silanes · silicon
[1] For some recent work, see: a) G. E. Keck, J. A. Covel, T. Schiff,
T. Yu, Org. Lett. 2002, 4, 1189; b) S. V. Mortlock, E. J. Thomas,
Tetrahedron 1998, 54, 4663; c) M. Lautens, R. N. Ben, P. H. M.
Delanghe, Tetrahedron 1996, 52, 7221, and references therein;
d) M. Lautens, P. H. M. Delanghe, Angew. Chem. 1994, 106,
2557; Angew. Chem. Int. Ed. Engl. 1994, 33, 2448.
[2] a) M. M. Patel, J. R. Green, Chem. Commun. 1999, 509; b) K. C.
Nicolaou, T. K. Chakraborty, A. D. Piscopio, N. Minowa, P.
Bertinato, J. Am. Chem. Soc. 1993, 115, 4419; c) A. Degl'Inno-
centi, P. Dembech, A. Mordini, A. Ricci, G. Seconi, Synthesis
1991, 267.
[3] For a review ofmethods ofr the synthesis ofallylsilanes, see:
a) “Product subclass 40: allylsilanes”: T. K. Sarkar, Sci. Synth.
2002, 4, 837; b) T. K. Sarkar, Synthesis 1990, 1101; c) T. K.
Sarkar, Synthesis 1990, 969.
[4] T. Akiyama, M. Kirini, Chem. Lett. 1995, 723.
[5] I. Fleming, S. K. Ghosh, J. Chem. Soc. Perkin Trans. 1 1998, 2711.
[6] For pioneering work on the synthesis oftetrahydroufrans
through
a [3+2] cycloaddition ofallylsilanes, see: a) J. S.
Panek, M. Yang, J. Am. Chem. Soc. 1991, 113, 9868; b) J. S.
Panek, R. Beresis, J. Org. Chem. 1993, 58, 809; c) for a review on
cycloaddition reactions ofallylsilanes, see: H.-J. Knꢀlker, J.
Prakt. Chem. 1997, 339, 304.
[7] Z.-H. Peng, K. A. Woerpel, Org. Lett. 2000, 2, 1379.
[8] J. H. Smitrovich, K. A. Woerpel, Synthesis 2002, 2778.
[9] J. H. Smitrovich, K. A. Woerpel, J. Org. Chem. 2000, 65, 1601.
[10] a) M. Suginome, Y. Ito, Chem. Rev. 2000, 100, 3221; b) M.
Murakami, M. Suginome, K. Fujimoto, H. Nakamura, P. G.
Andersson, Y. Ito, J. Am. Chem. Soc. 1993, 115, 6487.
[11] I. Fleming, D. Waterson, J. Chem. Soc. Perkin Trans. 1 1984,
1809.
[12] One or both oxygen atoms ofthe aldehyde may be complexed
with BF3·OEt2;[6b] another possibility is the formation of a
chelated species, see: K. Maruyama, Y. Ishihara, Y. Yamamoto,
Tetrahedron Lett. 1981, 22, 4235.
[13] S. R. Angle, N. A. El-Said, J. Am. Chem. Soc. 1999, 121, 10211.
[14] P. Mohr, Tetrahedron Lett. 1993, 34, 6251.
[15] Similarly, in our hands the use ofisobutyraldehyde or benzalde-
hyde did not yield any characterizable products.
[16] The aldehyde group of( tert-butyldiphenylsilyloxy)ethanal is
likely to be more electrophilic because of the inductive effect of
the neighbouring silyloxy substituent and, therefore, more
reactive than 3-(benzyloxy)propanal; see, also: J. S. Panek, M.
Yang, J. Am. Chem. Soc. 1991, 113, 9868.
Angew. Chem. Int. Ed. 2004, 43, 1417 –1419
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1419