Formation of Bicyclic Ethers from Lewis Acid Promoted Cyclizations of Cyclic Oxonium Ions

Article abstract of DOI:10.1021/ol025570d

(Matrix Presented) Oxacycles carrying a vinyl group and an acetal were extended with a cyclization terminator (vinyl silane or vinyl sulfide) by Suzuki coupling. Treatment with Lewis acid induced cyclization to provide bicyclic ethers in reasonable yields. In the case of the vinyl silane, an ene-like mechanism is preferred, whereas the thioenol ether enters into a Prins-type reaction channel. In one instance, the bicyclic compound was opened to give the ten-membered functionalized enone 24.

Full text of DOI:10.1021/ol025570d

ORGANIC
LETTERS
2002
Vol. 4, No. 8
1271-1274
Formation of Bicyclic Ethers from Lewis
Acid Promoted Cyclizations of Cyclic
Oxonium Ions
Pradip K. Sasmal and Martin E. Maier*
Institut fu¨r Organische Chemie, UniVersita¨t Tu¨bingen, Auf der Morgenstelle 18,
D-72076 Tu¨bingen, Germany
martin.e.maier@uni-tuebingen.de
Received January 16, 2002
ABSTRACT
Oxacycles carrying a vinyl group and an acetal were extended with a cyclization terminator (vinyl silane or vinyl sulfide) by Suzuki coupling.
Treatment with Lewis acid induced cyclization to provide bicyclic ethers in reasonable yields. In the case of the vinyl silane, an ene-like
mechanism is preferred, whereas the thioenol ether enters into a Prins-type reaction channel. In one instance, the bicyclic compound was
opened to give the ten-membered functionalized enone 24.
Oxonium ions are common electrophiles in organic synthesis.
This has to do with the fact that they are highly reactive and
that they react with a range of nucleophiles.1 Since oxonium
ions can be generated from various precursors, they can be
included into a synthetic scheme very easily. Thus, oxonium
ions can be formed from acetals2 or enol ethers3. The most
important method, however, is the generation of oxonium
ions from carbonyl compounds and Lewis acids. The addition
of nucleophiles to oxonium ions can be used for carbon-
carbon bond formation with the simultaneous creation of
stereocenters. Connecting the nucleophile and the oxonium
ion allows for ring-forming reactions, making cyclic ethers
and carbocycles easily accessible. Cyclic oxonium ions are
special in that they are key intermediates in the formation
of glycosidic bonds and spiroacetals.4 An interesting exten-
sion of the chemistry of cyclic oxonium ions would be further
intramolecular reaction with a nucleophilic group, leading
to bicyclic ethers.5,6 Such a setting could be of interest for
mechanistic reasons, that is, to probe the stereoelectronics
of the addition step. Moreover, bicyclic ethers might provide
an interesting scaffold for combinatorial application. Finally,
cleavage at the ether bridge could lead to functionalized
carbocycles.
(1) For reviews, see: (a) Blumenkopf, T.; Overman, L. E. Chem. ReV.
1986, 86, 857-873. (b) Schinzer, D. Nachr. Chem. Technol. Lab. 1989,
37, 498-505.
In this paper, we present our initial results in this area.
The general idea is illustrated in Scheme 1. Thus, a
hydroxyacetal, containing a nucleophilic group, for example,
a vinylsilane moiety, might undergo cyclization to an
oxonium ion that then would be trapped by the nucleophile.
(2) For some illustrative examples, see: (a) Trost, B. M.; Lee, D. C. J.
Am. Chem. Soc. 1988, 110, 6556-6558. (b) Hirst, G. C.; Howard, P. N.;
Overman, L. E. J. Am. Chem. Soc. 1989, 111, 1514-1515. (c) Castan˜eda,
A.; Kucery, D. J.; Overman, L. E. J. Org. Chem. 1989, 54, 5695-5707.
(d) Burke, S. D.; Strickland, S. M. S.; Organ, H. M.; Silks, L. A., III.
Tetrahedron Lett. 1989, 30, 6303-6306. (e) Bratz, M.; Bullock, W. H.;
Overman, L. E.; Takemoto, T. J. Am. Chem. Soc. 1995, 117, 5958-5966.
(f) Berger, D.; Overman, L. E.; Renhowe, P. A. J. Am. Chem. Soc. 1997,
119, 2446-2452. (g) Kusama, H.; Hara, R.; Kawahara, S.; Nishimori, T.;
Kashima, H.; Nakamura, N.; Morihira, K.; Kuwajima, I. J. Am. Chem. Soc.
2000, 122, 3811-3820. (h) Kadota, I.; Ohno, A.; Matsuda, K.; Yamamoto,
Y. J. Am. Chem. Soc. 2001, 123, 6702-6703.
(4) Perron, F.; Albizati, K. F. Chem. ReV. 1989, 89, 1617-1661.
(5) For some isolated examples, see: (a) Sonawane, H. R.; Maji, D. K.;
Jana, G. H.; Pandey, G. J. Chem. Soc., Chem. Commun. 1998, 1773-1774.
(b) Ohmura, H.; Smyth, G. D.; Mikami, K. J. Org. Chem. 1999, 64, 6056-
6059.
(6) For the [3 + 5] annulation route to bicyclic ethers, see: Molander,
G. A.; Eastwood, P. R. J. Org. Chem. 1995, 60, 4559-4565.
(3) (a) Petasis, N. A.; Lu, S.-P. J. Am. Chem. Soc. 1995, 117, 6394-
6395. (b) Kopecky, D. J.; Rychnovsky, S. D. J. Am. Chem. Soc. 2001, 123,
8420-8421.
10.1021/ol025570d CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/19/2002

Scheme 1. Tandem Cyclization of Hydroxyacetals Containing
Scheme 3. Synthesis of the Cyclization Substrates by Suzuki
a Nucleophilic Group
Cross-coupling Reaction
The synthesis of the appropriate substrates is shown in
Scheme 2. Beginning with the acetal aldehydes7 1a-c in
Scheme 2. Synthesis of 2-Alkoxy-vinyloxacycles
alkenes 3a-d with 9-BBN followed by the addition of the
alkenyl bromide 6a [1-bromo-1-trimethylsilylethene] or 6b
[1-bromo-1-(phenylthio)ethene], tetrakistriphenylphosphine-
palladium, aqueous NaOH, and benzene and refluxing of the
mixture provided the cyclization substrates 7-9 (Scheme
3).13 In a similar fashion, compound 3aOEt was subjected
to Suzuki coupling with 6c [1-bromo-2-trimethylsilylethene]
to afford compound 8aOEt.
In the case of 6a, the desired coupling product 7 was
always accompanied with the rearranged product 8 (ratio of
7:8 ≈ 2:1). Most likely, this rearrangement, which has been
observed by other groups,14 takes place via â-hydride
elimination from the alkenylpalladium species E (Figure 1).
Grignard reaction with vinyl- or 2-propenylmagnesium
bromide provided the allylic alcohols 2a-d. These com-
pounds could be cyclized very easily to the cyclic hemiacetals
3a-d as anomeric mixtures.8 As opposed to a report in the
literature,9 compound 2b did not cyclize spontaneously. Since
the procedure for obtaining the aldehyde 1a by ozonolysis
of 1,5-cyclooctadiene10 turned out to be quite unreliable in
our hands, another route to a compound of type 2a was
usually employed. Thus, Grignard addition of the C3 reagent
4 to acrolein gave the allylic alcohol 5,11 which was subjected
to transacetalization with ethanol giving 3aOEt.
Figure 1. Possible stereoelectronic effect during the rearrangement
of E to G.
To install the nucleophilic group, a Suzuki cross-coupling
reaction was employed.2e,12 Thus, hydroboration of the
This elimination might be facilitated by hyperconjugation
of the C-SiMe3 bond to the σ*-orbital of the C-H bond. If
the cross-coupling reaction between 3c and 6a was carried
out at lower temperatures, the rearranged product even
becomes the major product (ratio ≈ 1:2, 61% yield). It should
be noted that in this instance, modified conditions were used
(5% (dppf)2PdCl2, 5% Ph3As, Cs2CO3, DMF/THF/H2O, 23
(7) Claus, R. E.; Schreiber, S. L.; Jeong, N.; Semmelbeck, M. F. Org.
Synth. 1986, 64, 150-156.
(8) Diastereomeric ratios: 3a, 1.4:1; 3b, 2.2:1; 3c, 2.3:1; 3d, 4.7:1;
3aOEt, 1.4:1.
(9) Felkin, H.; Gault, Y.; Roussi, G. Tetrahedron 1970, 26, 3761-3778.
(10) Li, P.; Wang, J.; Zhao, K. J. Org. Chem. 1998, 63, 3151-3152.
(11) Kawashima, M.; Fujisawa, T. Bull. Chem. Soc. Jpn. 1988, 61, 1,
3377-3379.
1272
Org. Lett., Vol. 4, No. 8, 2002

Table 1. Lewis Acid Induced Cyclization to Bicyclic Ethers
entry substratesa promotorb
products
1
2
3
4
5
8a OEt
7a , 8a (2:1) SnCl4
7bc
7bc
7cc
7d c
SnCl4
13b (53%)
13a , 13b (2:1, 48 %)19
14a (12%), 14b (43%)
14b (56%)
SnCl4
TMSOTf
SnCl4
15 (50%)
16 (8:1, 60%)
Figure 2. Diastereoselective hydroboration of the propenyl deriva-
tive 3d.
6
SnCl4
7
8
11, 12 (2:1) SnCl4
11, 12 (2:1) TMSOTf
17a (5%), 17b (33%), 17c (19%)
17b, 17c (2:1, 59%)
9
9a
9b
9c
9d
BF3‚Et2O 19a , 19b (2:1, 69%)
BF3‚Et2O 20 (72%)
BF3‚Et2O 21 (71%)
°C).15 This result is in contrast to the findings described by
Gilchrist et al.
10
11
12
In the case of the propenyl derivative 3d, the hydroboration
was highly diastereoselective (8:1). The relative stereochem-
istry was assigned on the basis of the model of Houk et al.16
According to this model, anti-diols are produced via a
staggered transition state I where the substituents R1 and OR2
occupy the anti and outside positions, respectively (Figure
2). This assignment could be confirmed at a later stage (vide
infra).
To probe the influence of an alkoxy substituent, located
next to the acetal function, on the cyclization reaction, the
substrate 10 was prepared from D-ribose via a sequence of
acetalization, acetonide formation, Swern oxidation, and
Wittig olefination.17 A subsequent hydroboration and Suzuki
coupling with 6a provided the compound 11 together with
12 in good yield (11:12 ) 2:1).
Cyclization Reactions. As has been described by Over-
man et al.,18 alkenylsilane cyclizations onto acetals are best
performed with tin(IV) chloride or trimethylsilyl triflate as
promoters in dichloromethane as the solvent. On the other
hand, cyclization with vinyl sulfide terminators seems to be
more efficient with borontrifluoride etherate (BF3‚Et2O) in
tert-butylmethyl ether. Accordingly, we applied similar
conditions with our substrates. These results are summarized
in Table 1 (for the structures, see Figure 3).
BF3‚Et2O 22a , 22b (8:1, 73%)
a Concentration of the starting material, 0.02 M; reaction scale, 30-60
mg. b TMSOTf (1.5 equiv), SnCl4 (1.5 equiv), and BF3‚Et2O (2.0 equiv)
were used; reactions with TMSOTf and SnCl4 were carried out in CH2Cl2
and reactions with BF3‚Et2O in tBuOMe. c Particular pure fractions
(containing less than 10% of the rearranged isomer 8) were used. In these
small-scale cyclizations, the ∆3,4-bicyclic ether could not be detected.
vinyl sulfide derivatives (2-thiophenylalkenes) proceeds most
likely via an intramolecular Prins cyclization (intermediate
O generating N and/or P). This pathway is supported by the
electronic nature of the thiophenyl group. The subsequent
elimination of a proton from the carbocation can take place
in two directions. The regiochemistry of the elimination
depends on the size of the cyclic oxonium ion and the
stability of the cyclized products. The six-membered oxo-
nium ion gave only the ∆4,5-bicyclic ether, whereas the
seven-membered oxonium ions gave mainly the ∆3,4-bicyclic
ether. In the case of five-membered oxonium ions, a mixture
of the two possible double-bond isomers is formed. The
product ratio is in part reflected in the calculated heat of
formation (HyperChem 6.02).20
The ribose derivative could also be induced to cyclize.
For the cyclization, a 2:1 mixture of the 2- and 1-trimeth-
From these experiments, the following conclusions can
be made (Figure 4). (a) The 2-trimethylsilylalkenes prefer
an ene-type cyclization (leading to N via transition state L
and intermediate M) irrespective of the ring size of the cyclic
oxonium ion. An intramolecular Prins reaction would gener-
ate products with an exocyclic methylene group, but no such
product was observed in any of these cases. Usually, ene
cyclizations are accompanied by protodesilylation, presum-
ably by adventitious HCl generated inside the reaction flask
or during workup (N, R ) H). (b) The cyclization of the
(12) For reviews, see: (a) Miyaura, N.; Suzuki, A. Chem. ReV. 1995,
95, 2457-2483. (b) Chemler, S. R.; Trauner, D.; Danishefsky, S. J. Angew.
Chem. 2001, 113, 4676-4701; 2001, 40, 4544-4568.
(13) Miyaura, N.; Ishiyama, T.; Sasaki, H.; Ishikawa, M.; Satoh, M.;
Suzuki, A. J. Am. Chem. Soc. 1989, 111, 314-321.
(14) Ennis, D. S.; Gilchrist, T. L. Tetrahedron 1990, 46, 2623-2632.
(15) Trost, B. M.; Lee, C. B. J. Am. Chem. Soc. 1998, 120, 6818-6819.
(16) Houk, K. N.; Rondan, N. G.; Wu, Y.-D.; Metz, J. T.; Paddon-Row,
M. N. Tetrahedron 1984, 40, 2257-2274.
(17) Ugarkar, B. G.; DaRe, J. M.; Kopcho, J. J.; Brown, C. E., III;
Schanzer, J. M.; Wiesner, J. B.; Erion, M. D. J. Med. Chem. 2000, 43,
2883-2893.
(18) Blumenkopf, T. A.; Bratz, M.; Castan˜eda, A.; Look, G. C.; Overman,
L. E.; Rodriguez, D.; Thompson, A. S. J. Am. Chem. Soc. 1990, 112, 4386-
4399.
Figure 3. Structures of cyclization products (cf. Table 1).
Org. Lett., Vol. 4, No. 8, 2002
1273

Figure 5. Calculated conformation of 24conf and important
NOESY cross-peaks.
Figure 4. Mechanistic pathway for the reaction of the oxonium
ion K.
NMR data of 24 are in accordance with the conformer
24conf.25 This conformation is characterized by a pseudoaxial
O-acetoxy group. Prominent NOESY cross-peaks were
observed between H-2/H-9, H-2/H-10, H-6/H-8, and H-6/
H-9, supporting the calculated conformation.
To summarize, we were able to show that the cyclization
of cyclic oxonium ions represents an efficient strategy for
obtaining unsaturated bicyclic ethers. Particularly, the at-
tachment of the nucleophilic terminators via a Suzuki cross-
coupling reaction makes the whole sequence very concise.
The general strategy might also be applicable for the
synthesis of bicyclic amines.
ylsilyl alkene 11 and 12 was employed. Use of SnCl4 as a
Lewis acid afforded a mixture of three products with a ∆4,5/
∆3,4 ratio of 2:1 (57% total yield).
∆
It is very likely that the 4,5 isomers are formed via the
ene pathway (boatlike transition state), whereas the ∆3,4
isomers result from the terminal vinylsilane via a Prins
mechanism. Although it was not possible to separate the two
isomers 17b and 17c, they could easily be distinguished in
1
the H and 13C NMR spectra due to their different sym-
metry.21 Hydrogenation of the 17b/17c mixture over pal-
ladium hydroxide in ethyl acetate gave the saturated bicyclic
ether 18.
Acknowledgment. Financial support by the Deutsche
Forschungsgemeinschaft and the Fonds der Chemischen
Industrie is gratefully acknowledged. P. K. Sasmal thanks
the Alexander von Humboldt foundation for a fellowship.
To study the cleavage of the ether bridge, the thioenolether
moiety of 22a/b was hydrolyzed with aqueous HCl to the
bicyclic ketone 23. This ketone was treated with LDA to
give the monocyclic keto alcohol, which was directly
converted to the acetate 24, a functionalized ten-membered
carbocycle. The base-induced ring opening of 23 is very
slow, but longer reaction times gave a better yield.22 In
principle, the vinyl sulfide cyclization products could be
opened to from eight- to ten-membered carbocycles, which
are important structural motifs in several natural products23
or serve as precursors for cyclic peptide mimetics.24
Supporting Information Available: Experimental pro-
cedures for the sequence of Grignard reaction, cyclic acetal
formation, Suzuki coupling, and bicyclic ether formation and
the conversion of 22 to 24, NMR spectra for important
compounds, and the NOESY spectrum of compound 24. This
material is available free of charge via the Internet at
http://pubs.acs.org.
OL025570D
(19) Two of the compounds, 13a and 13b, have been prepared before
for 1,5-cyclooctadiene: Cope, A. C.; McKervey, M. A.; Weinshenker, N.
M. J. Org. Chem. 1969, 34, 2229-2231.
(20) The following values were obtained: 19a -41.19, 19b -26.54, 21
-86.65, 21-∆4,5-isomer -72.22 kJ/mol. For 20, calculations predict the
3,4-isomer to be more stable by 4.66 kJ/mol. In this case, the ene-type
pathway might dominate.
Scheme 4. Base-Induced Ring-opening of the Bicyclic Keto
Ether 22
(21) For example, the Cs-symmetric 17b shows two 13C signals in the
region between 80 and 90 ppm (2 RCHOR carbons). The minor isomer
17c shows the expected four peaks.
(22) For the ring-opening of bicyclic ethers, see: Lautens, M.; Ma, S.;
Chiu, P. J. Am. Chem. Soc. 1997, 119, 6478-6487.
(23) For example, germacranes can be mentioned: Utagawa, A.; Hirota,
H.; Ohno, S.; Takahashi, T. Bull. Chem. Soc. Jpn. 1988, 61, 1207-1211.
(24) Olson, G.; Voss, M. E.; Hill, D. E.; Kahn, M.; Madison, V. S.;
Cook, C. M. J. Am. Chem. Soc. 1990, 112, 323-333.
(25) This conformation turned out to be the global minimum in a
conformational search using MacroModel 7.0 (1000 starting conformations,
50 kJ/mol cutoff, MM+ force field). (a) Saunders, M.; Houk, K. N.; Wu,
Y.-D.; Still, W. C.; Lipton, M.; Chang, G.; Guida, W. C. J. Am. Chem.
Soc. 1990, 112, 1419-1427. (b) Kolossva´ry, I.; Guida, W. C. J. Comput.
Chem. 1993, 14, 691-698.
1274
Org. Lett., Vol. 4, No. 8, 2002