Synthesis of cyclic ethers utilizing a cyclization-fragmentation strategy

Article abstract of DOI:10.1016/S0040-4020(99)01063-7

A variety of cyclic ethers have been prepared via both solution phase and polymer-supported sequences of 3 + 2 cycloaddition of nitrile oxides to alkenes and dienes to give isoxazolines, followed by electrophile-induced cyclization. Library generation by

Full text of DOI:10.1016/S0040-4020(99)01063-7

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 711–717
Synthesis of Cyclic Ethers Utilizing
a Cyclization–Fragmentation Strategy
*
Denise A. Ockey, David R. Lane, Juliette A. Seeley and Neil E. Schore
Department of Chemistry, University of California, Davis, CA 95616, USA
Received 14 October 1999; revised 25 November 1999; accepted 7 December 1999
Abstract—A variety of cyclic ethers have been prepared via both solution phase and polymer-supported sequences of 3ϩ2 cycloaddition of
nitrile oxides to alkenes and dienes to give isoxazolines, followed by electrophile-induced cyclization. Library generation by alkylative
elaboration of isoxazolines was unsuccessful, but both simple and substituted dienes were found suitable for polymer-supported formation of
cyclic ethers of ring sizes five through seven. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
have included hydroxymethyl polystyrene, mercaptomethyl
polystyrene, or oxidized (e.g. aldehyde or carboxylic acid)
derivatives. With these types of linking strategies, part of
the linkage function normally remains present in the product
after cleavage from the solid support. Consequently, there
Solid Phase Organic Synthesis (SPOS) has in principle
many advantages over more traditional synthetic methods.
Excess reagents and soluble side products can be washed
Figure 1. Literature examples of polymer-supported cyclization reactions.
away from the polymer thus making purification easier;
resin bound toxic substances can be handled more easily
and more safely; reactions that exhibit poor chemoselec-
tivity can often be made more efficient by attachment of
one of the components to the solid support. Reaction
substrates may be linked to the polymer via any number
of cleavable functional groups. Common modifications of
the basic Merrifield-type resins for substrate attachment
has been considerable effort devoted to developing the
so-called “traceless linkers” in order to circumvent this
problem. One of the earliest, which utilizes a silicon–carbon
bond that can later be cleaved from the final product, was
described by Ellman in his synthesis of a library of
1,4-benzodiazepines, in which the final products were
obtained by HF cleavage of the carbon–silicon bond.1
Another tactic for traceless cleavage is via cyclization. This
method has the advantage that usually only the desired
compounds and not side products are cleaved from the
polymer, thus making purification of the products much
Keywords: cyclization; ethers; isoxazolines; nitrile oxides; polymer
support.
* Corresponding author; e-mail: neschore@ucdavis.edu
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(99)01063-7

712
D. A. Ockey et al. / Tetrahedron 56 (2000) 711–717
Scheme 1.
simpler. Fig. 1 shows two examples of this type that have
been reported.2
homogeneous solution. As a bonus, a “site-isolation” effect
is observed under solid-phase conditions, obviating the need
to employ a very large excess of diene in the dipolar
cycloaddition step that forms the isoxazoline ring.
The conversion of isoxazolines into cyclic ethers first
reported by Kurth has been of particular recent interest to
us. Cyclic ethers of various sizes can be synthesized by
electrophilic cyclization between the oxygen of the isoxazo-
line and an activated olefin. While four- and eight-
membered cyclic ethers are not accessible in this way,
five-, six-, and seven-membered rings may be obtained in
reasonable yields.3
The cyclization reaction is thought to proceed via the cyclic
isoxazolinium intermediates shown in Fig. 2.5 A preference
for the formation of the trans product is usually observed, a
consequence of a nonbonding interaction between the
electrophile and the isoxazoline ring: the exo-intermediate
should be the more stable. Diastereoselectivity is solvent-
dependent, presumably because of endo/exo equilibration
between the isoxazolinium intermediates. Subsequent
irreversible loss of “Rϩ” and N–O bond cleavage gives
the cyclic ether.6
In our laboratory Beebe showed that the Kurth methodology
for the synthesis of 2,5-disubstituted tetrahydrofurans
(THFs) could be adapted to the solid phase and incorporate
a cyclization-based traceless linker strategy. In this study
ICl was employed to initiate an electrophile-promoted
cyclative cleavage process from a polymer-bound isoxazo-
line bearing a pendant alkene moiety as shown in Scheme 1.4
The goal of this project was to expand the range of cyclic
ethers that could be synthesized in homogeneous solution as
well as by SPOS, utilizing the electrophile-promoted
cyclization strategy on a variety of specifically designed
isoxazolines. Herein we describe the results of these efforts.
There are several advantages to using SPOS techniques
rather than conventional solution-phase methodology in
synthesizing cyclic ethers via isoxazolines, the most notable
being ease in separation of byproducts at several stages.
During isoxazoline synthesis the diphenylurea formed can
easily be washed away from the polymer using excess
solvent. Furthermore, only the desired cyclic ethers are
cleaved from the polymer in the final step. Side products
derived from addition to the alkene without concomitant
cyclization remain polymer-bound under SPOS conditions.
These side products are very difficult to separate from the
desired cyclic ethers when the reaction is carried out in
Results and Discussion
Our initial efforts were aimed at designing routes by which
tetrahydrofurans of varying substitution patterns could be
synthesized via an appropriate precursor isoxazoline 1,
shown retrosynthetically in Scheme 2.
We sought first to prepare a library of isoxazolines via
alkylation a to the activating group ‘G’. For this purpose
Figure 2. Proposed mechanism for the electrophilic cyclization.

D. A. Ockey et al. / Tetrahedron 56 (2000) 711–717
713
Scheme 2.
we chose allyl phenyl sulfone as the dipolarophile for the
isoxazoline-forming cycloaddition. Prior to attempting the
chemistry on the polymer, the solution equivalent was
carried out.
diastereotopic methylene protons coupled to the hydrogen
at d 6.44. Evidently, formation of the anion a to the sulfone
gave rise to ring-opening of the isoxazoline leading to the
trans-a,b-unsaturated sulfone oxime 5. In the 13C NMR,
resonances corresponding both to the carbon of the oxime
functional group and to those of the a,b-unsaturated sulfone
appeared. When formed in the presence of D2O, deuterium
was incorporated at the methylene a to the oxime group.
This material underwent isomerization to at least two other
substances upon storage for several hours.
The sulfone-substituted isoxazoline was prepared via the
nitrile oxide by treatment of nitrosilyl ether 2 with phenyl-
isocyanide, triethylamine, and allyl phenyl sulfone to give 3
in 84% yield as an inseparable mixture of diastereomers
(Scheme 3). Separation of the diphenylurea byproduct
from the desired isoxazoline was problematic, requiring
tedious chromatography.
It having become clear that the generation of an anion a to
the isoxazoline opens the ring, we turned instead to a
scheme designed to synthesize substituted tetrahydropyrans
via introduction of a malonate-derived diester moiety that
could be alkylated with suitable unsaturated species. We
began by using diethyl allylmalonate as a dipolarophile.
Under solution-phase conditions, 1,3-dipolar cycloaddition
between 2 and diethyl allylmalonate with phenyl isocyanate
as the dehydrating agent gave compound 6 in 44% yield, Eq.
(1).
Alkylation was first attempted in the usual way7 by treat-
ment with n-BuLi in THF at Ϫ78ЊC followed by addition of
allyl bromide in THF. Alkylation was not observed; instead,
only the deprotected starting material was isolated. We
therefore changed the protecting group to the TBDMS
ether to improve stability, converting 3 into 4 in 92%
yield, as shown.8 However, alkylation of sulfone 4 with
allyl bromide still failed: some starting material was
isolated, along with a second substance. The same result
was obtained using methyl iodide. Interestingly, treatment
with n-BuLi followed by quenching with D2O gave no
deuterium incorporation in the recovered starting material.
Using sec-BuLi under the original alkylating conditions, we
found by TLC that all of the starting material was
consumed. Allyl bromide was added, but again no alkylated
product was obtained. In this case, the new compound
observed in the previous alkylation attempts was obtained
as a single product in virtually quantitative yield. Analysis
by 1H NMR revealed that the isoxazoline ring was no longer
present; instead, there appeared a doublet at d 6.13, a ddd
pattern at d 6.44, consistent with a trans disubstituted
double bond, and two signals at d 3.56 and d 3.72 for
ꢀ1†
In initial alkylation attempts we chose LDA rather than the
more commonly used NaH,9 reasoning that the soluble base
would be more suitable for the polymer-based system.
Malonate-substituted isoxazoline 6 was treated with LDA
in THF at 0ЊC and then allyl bromide. After stirring over-
night, TLC showed only the starting material. Further
Scheme 3.

714
D. A. Ockey et al. / Tetrahedron 56 (2000) 711–717
Scheme 4.
attempts with a variety of bases under several different sets
of conditions also returned only the unreacted starting mate-
rial. Treatment of compound 6 with the dimsyl anion
followed by quenching with D2O did not result in any
deuterium incorporation, indicating, again, a system oddly
resistant to deprotonation.
vary appreciably.
ꢀ2†
Failing to develop alkylation-based approaches, we returned
to diene-based methodology. Following the published
procedure,4 polymer-bound 2-nitro-1-phenylethan-1-ol
was prepared by reaction of polymer-bound benzaldehyde
with CH3NO2 in THF and protected as the TMS ether to
give the cycloaddition precursor. We repeated the published
preparation of the simplest system, 2-(cyanomethyl)-6-
(iodomethyl)tetrahydropyran. The required polymer-bound
isoxazoline was synthesized by a 1,3-dipolar cycloaddition
with 1,6-heptadiene.
We then turned to the synthesis of 4-substituted 1,6-dienes,
to provide access to 4-substituted THP derivatives (Scheme
4), hoping in addition that substitution in the diene would be
beneficial in promoting the desired cyclization process.
Although ICl can be used successfully in the polymer-
supported electrophile-induced cyclization, competing
addition to the terminal double bond is a persistent
problem.4 In this study and in a related work we explored
the use of several electrophile sources, including IBr,
N-iodosuccinimide, PhSeCl, and PhSe(phthalimide),
finding that the various sources of “Iϩ” showed only mini-
mal differences in yields and stereoselectivities for cyclic
ether products, while the Se-based reagents did not induce
cyclization at all. As a result, we chose to use I2, which is an
effective cyclization promoter under homogeneous con-
ditions and gives clean products provided that light is
excluded and unconsumed I2 is removed completely.6 The
polymer-bound isoxazoline was swollen in CH2Cl2, and
5 equiv. of I2 were added in one portion. After the
appropriate reaction time and workup, a 2.7% yield of
2,6-disubstituted THP was obtained (Eq. (2)), an average
per step yield of 40%, somewhat poorer than earlier
results with ICl. A cis/trans ratio of 1:2.6 was determined
by GC analysis, identical to that found with ICl. Several
optimization attempts were made, but the yields did not
Diene preparation was based on work of Greeves and Lee,
who prepared d,e-unsaturated aldehydes from bis-allyl
ethers by tandem 2,3-Wittig–anionic oxy-Cope rearrange-
ments.10 We prepared bis-allyl ether 8 by Williamson
synthesis using cinnamyl bromide and allyl alcohol in
84% yield. Next, ether 8 was dissolved in THF, cooled to
0ЊC, and treated with 2.5 equiv. KH and 1.5 equiv.
18-crown-6, Scheme 5. After warming to rt, quenching,
and work-up, aldehyde 9 was isolated in 96% yield. Finally,
diene 10 was prepared in a 75% yield by Wittig olefination
using standard conditions.
We proceeded directly with the polymer-supported 1,3-
dipolar cycloaddition. Polymer bound 2-nitro-1-phenyl-
ethan-1-ol was treated with phenyl isocyanate, triethyl-
amine, and 4-phenyl-1,6-heptadiene 10 in refluxing
benzene, giving isoxazoline-containing polymer 11, which
was then swollen in dry CH2Cl2 and treated with 5 equiv. I2.
The mixture was stirred in the dark overnight, and the
excess I2 quenched with aqueous Na2S2O3. After filtration
Scheme 5.

D. A. Ockey et al. / Tetrahedron 56 (2000) 711–717
715
Figure 3. Diastereomers of substituted tetrahydropyrans 12.
and removal of solvent an oil remained which was found by
GC to consist of a 1:1:1:1 mixture of the four possible
diastereomers 12, Fig. 3. The most characteristic absor-
bances in the 1H NMR were the pairs of doublets of doublets
that corresponded to the hydrogens a to the iodide and
nitrile groups. The four-step yield was a respectable 7.8%,
an average per step yield of 53%. This particular system
therefore provides proof of the concept for an approach to
libraries of 4-substituted tetrahydropyrans beginning with
appropriately substituted allylic halides.
previously in this way. The starting diene, 16, was prepared
in one step from dihydropyran utilizing the reported pro-
cedure.11 Following 1,3-dipolar cycloaddition and subse-
quent electrophilic cyclization, 18 was obtained in a 7.4%
overall yield, 52% per step (Scheme 7). The diastereomer
ratio was determined by GC analysis again to be 1:1.
Conclusion
We have expanded the scope of the polymer-supported
multistep preparation of cyclic ethers to include both
seven-membered rings and spirocyclic derivatives, and to
demonstrate that libraries of substituted tetrahydropyrans
are in principle accessible given the relative ease of prepa-
ration of the precursors. Diene-based approaches analogous
to those previously reported were successful, although the
benefits of reducing the need for large excesses of diene and
eliminating multiple tedious purification steps were partly
tempered by the modest overall yields for the four-step
sequence on the polymer. Alternative approaches based on
alkylations of two different isoxazoline-based systems were
unsuccessful; a novel ring opening process was found to
arise from deprotonation a to the 5-position of the isoxazo-
line ring.
We next attempted to synthesize a seven-membered ring
ether utilizing both solution-phase and polymer-supported
conditions. The 1,3-dipolar cycloadditions were success-
fully carried out between 1,7-octadiene and both solution-
phase and polymer-bound nitrile oxides. Following electro-
philic cyclization on the products 13 and 14, respectively,
oxepane 15 was isolated from 14 in solution in 52% yield
(final step yield only), and in 8.1% overall yield from the
polymer, corresponding to 53% per step (Scheme 6). GC
analysis revealed that both products consisted of a 1:1
mixture of cis and trans diastereomers.
Finally, we investigated formation of a spiroketal using this
methodology. Spiroketals have been synthesized in solution
Scheme 6.
Scheme 7.

716
D. A. Ockey et al. / Tetrahedron 56 (2000) 711–717
Experimental
Jˆ8.1, 14.4 Hz), 3.25 (dd, 0.5H, Jˆ9.9, 17.7 Hz), 3.38 (dd,
0.5H, Jˆ4.9, 14.4 Hz), 3.51 (dd, 0.5H, Jˆ4.8, 14.4 Hz),
4.81 (m, 0.5H), 4.89 (m, 0.5H), 5.67 (s, 1H), 7.25–7.90
(m, 10H); 13C NMR: d Ϫ5.0, Ϫ3.5/Ϫ3.0, 25.0, 37.2/37.3,
59.3/59.5, 72.5, 122.7, 125.3, 125.8, 126.4, 131.6, 137.0/
137.3, 157.8/158.0; HRMS: calcd for C22H28NO4SSi
(MϪCH3)ϩ 430.1501, found 430.1508.
General procedures
See Ref. 4b for general information regarding purification of
solvents and reagents, carrying out of reactions, polymer
handling, and analytical procedures. Fourier Transform IR
spectra were obtained on a Matson Genesis II with the
Golden Gate Attenuated Total Reflectance (ATR) attach-
ment. Radial chromatography was done on a Chromatotron
using radial plates of either 1-, 2-, or 4-mm thickness of
silica gel 60 PF; bands were visualized by UV. Capillary
GC was done on a Shimadzu 14A using a DB-101 column.
Conditions: Initial temp 150ЊC, hold time 2.0 min, ramp
5Њ/min, final temp 240ЊC, hold time 5.0 min.
E-1-Phenyl-5-[(phenyl)sulfonyl]-1-[(tert-butyldimethyl-
silyl)oxy]-4-penten-2-one oxime (5). A solution of 460 mg
(1.07 mmol) of 4 in 3 mL of THF was cooled to Ϫ78ЊC,
treated with 1.8 mL of sec-BuLi (1 M in THF), and stirred at
Ϫ78ЊC for 10 min. TLC (4:1 hexane/ethyl acetate) showed
complete loss of the starting material. The mixture was
quenched with water, the layers separated, the aqueous
layer extracted with Et2O, and the combined Et2O layers
dried (Na2SO4). The solvent was removed to give a viscous
red oil, a quantitative yield of a 1:1 mixture of (presumably)
syn and anti oximes, one of which could by obtained in
relatively pure form by a sequence of column chroma-
tography using hexane followed by radial chromatography
using 1:1 Et2O/hexane: FTIR (neat) 3408 (br), 3060, 3030,
2950, 2935, 2891, 2856, 1636, 1593, 1458, 1446, 1398,
1313, 1255, 1147, 1097, 974, 862, 845, 777, 739,
4,5-Dihydro-5-{[(phenyl)sulfonyl]methyl}-3-{phenyl[(tri-
methylsilyl)oxy]methyl}isoxazole (3). To a solution of
3.0 g (12.5 mmol) 2 in 50 mL Et2O was added 3.4 g
(18 mmol) allyl phenyl sulfone and 0.1 mL triethylamine.
Over a period of three days a total of 3.5 g (36 mmol) phenyl
isocyanate was added in portions. The reaction mixture was
quenched with 10 mL water, filtered, and the filtrates dried
(MgSO4), and evaporated to give a yellow oil. Chroma-
tography (3:2 Et2O/hexane) gave 4.05 g of a light yellow
oil estimated by NMR to contain 3.5 g (72% yield) of a ca.
1:1 mixture of diastereomers of 3, the remainder being
1
700 cmϪ1; H NMR d Ϫ0.02 (s, 3H), 0.02 (s, 3H), 0.85
(s, 9H), 3.56 (dd, 1H, Jˆ8.7, 14.1 Hz), 3.72 (dd, 1H,
Jˆ6.5, 14.1 Hz), 5.38 (s, 1H), 6.13 (d, 1H, Jˆ15.9 Hz),
6.44 (ddd, 1H, Jˆ6.5, 8.7, 15.9 Hz), 7.15–7.85 (m, 10H),
8.1 (br s, 1H); partial 13C NMR for the isomer mixture
d 133.5, 137.6, 140.5, 141.1, 155.2, 158.7; HRMS calcd
for C19H22NO4SSi (MϪC4H9)ϩ 388.1039, found
388.1022.
1
residual allyl phenyl sulfone: H NMR d 0.12 (s, 9H),
2.56 (dd, 0.5H, Jˆ7.4, 17.2 Hz), 2.87 (dd, 0.5H, Jˆ10.5,
17.7 Hz), 3.04 (m, 1H), 3.25 (dd, 0.5H, Jˆ10.3, 16.9 Hz and
dd, 0.5H, Jˆ7.6, 14.2 Hz), 3.39 (dd, 0.5H, Jˆ4.8, 14.1 Hz),
3.52 (dd, 0.5H, Jˆ5.1, 14.4 Hz), 4.79 (m, 0.5H), 4.89 (m,
0.5H), 5.66 (s, 1H), 7.25–7.95 (m, 10H). Loss of the
trimethylsilyl group accompanied all preparative scale
purification attempts of 3, necessitating the use of the
crude material in subsequent experiments.
5-[2,2-Bis(ethoxycarbonyl)]ethyl-4,5-dihydro-3-{phenyl-
[(trimethylsilyl)oxy]methyl}isoxazole (6). To a solution of
5.76 g (24.0 mmol) of 2 and 9.64 g of diethyl allylmalonate
(48.0 mmol) in 50 mL of benzene was added 17.86 g phenyl
isocyanate (99.6 mmol) and 50 drops of triethylamine. The
solution was refluxed for two days, cooled, water was
added, and the solution was stirred for 18 h at rt. The
mixture was filtered, and the solvent evaporated. The
product was purified by column chromatography (10%
EtOAc/hexane) to yield 6 as a viscous yellow oil (4.41 g,
41%) as a 1:1 mixture of diastereomers, which proved to be
inseparable: FTIR (neat) 1749, 1733, 1252, 1071, 877,
4,5-Dihydro-5-{[(phenyl)sulfonyl]methyl}-3-{phenyl[(tert-
butyldimethylsilyl)oxy]methyl}isoxazole (4). Crude 3
(5.67 g total, ca. 4.0 g 3, remainder allyl phenyl sulfone)
was deprotected by being dissolved in 10 mL MeOH
containing 1.25 g citric acid. After 1 h at rt the solvent
was removed, the residue taken up in 1:1 Et2O/hexane,
filtered, and run down a short silica column. Allyl phenyl
sulfone was eluted, the solvent was changed to MeOH, and
the product was removed. Evaporation of solvent gave an oil
which was taken up in Et2O, dried (MgSO4), and evaporated
to give 2.47 g (ca. 72% yield) of the alcohol, which was
1
844 cmϪ1; H NMR d 0.04 (s, 4.5H), 0.06 (s, 4.5H), 1.14
(m, 6H), 1.88 (m, 1H), 2.05 (t, 1H Jˆ7.1 Hz), 2.15 (dd,
0.5H, Jˆ7.8, 17.1 Hz), 2.60 (m, 1H), 3.01 (dd, 0.5H,
Jˆ10.5, 17.1 Hz), 3.43 (dd, 0.5H, Jˆ5.7, 9.0 Hz), 3.50 (t,
0.5H, Jˆ7.2 Hz), 4.07 (m, 4H), 4.35 (m, 0.5H), 4.43 (m,
0.5H), 5.60 (s, 1H), 7.12–7.25 (m, 5H); 13C NMR: d Ϫ0.33/
Ϫ0.30, 13.8, 33.3/33.6, 37.1, 47.8/47.9, 60.8, 68.7/68.8,
76.0/76.4, 124.6, 124.7, 126.9, 127.4, 139.3, 159.9, 168.0;
HRMS: calcd for C20H28NO6Si (MϪCH3)ϩ 406.1686, found
406.1687.
1
used without further purification: H NMR d 2.62 (dd,
0.5H, Jˆ7.5, 17.7 Hz), 2.95–3.20 (m, 3H), 3.25 (dd, 0.5H,
Jˆ7.5, 14.4 Hz), 3.43 (dd, 0.5H, Jˆ4.8, 14.1 Hz), 3.49 (dd,
0.5H, Jˆ5.1, 14.1 Hz), 4.82 (m, 1H), 5.53 (s, 0.5H), 5.56 (s,
0.5H), 7.25–7.95 (m, 10H). A mixture of 5.63 g of this alco-
hol, 2.90 g (42.7 mmol) imidazole, and 3.14 g (20.8 mmol)
tert-butyldimethylchlorosilane in 30 mL CH2Cl2 was stirred
at rt for 18 h. The solvent was removed, and the residue was
taken up in 1:1 Et2O/hexane and filtered through a short
silica plug. Removal of solvent gave 6.75 g (92% yield)
essentially pure 4: FTIR (neat) 1321, 1251, 1151, 1070,
Polymer-supported 5-(4-pentenyl)-4,5-dihydro-3-{phenyl-
[(trimethylsilyl)oxy]methyl}isoxazole (7). To 1.1 g of
polymer-bound nitrosilyl ether4 and 10 mL of benzene
was added 1 mL phenyl isocyanate (9.2 mmol), 1.0 g
1,6-heptadiene (10.4 mmol), and 0.5 mL triethylamine
(3.6 mmol). The solution was refluxed for four days, cooled,
treated with water, and stirred for 18 h at rt. The usual
877, 844 cmϪ1; H NMR d 0.05 (s, 3H), 0.08 (s, 3H), 0.92
1
(s, 4.5H), 0.93 (s, 4.5H), 2.57 (dd, 0.5H, Jˆ7.4, 17.6 Hz),
2.86 (dd, 0.5H, Jˆ10.2, 17.7 Hz), 3.00 (dd, 0.5H, Jˆ8.1,
14.4 Hz), 3.06 (dd, 0.5H, Jˆ6.3, 17.7 Hz), 3.22 (dd, 0.5H,

D. A. Ockey et al. / Tetrahedron 56 (2000) 711–717
717
workup gave polymer 7 as a yellow solid: FTIR (powder)
1253, 1071, 877, 842 cmϪ1
2-(Cyanomethyl)-6-(iodomethyl)-4-phenyltetrahydro-
.
pyran (12). In 7.1% yield via polymer 11; FTIR (neat)
1
2251 cmϪ1; H NMR d 0.78–2.00 (m, 5H), 2.52 (dd, 1H,
Polymer-supported 5-(2-phenyl-4-pentenyl)-4,5-dihydro-
3-{phenyl[(trimethylsilyl)oxy]methyl}isoxazole (11). To
1.1 g of polymer-bound nitrosilyl ether and 10 mL of
benzene was added 1 mL phenyl isocyanate (9.2 mmol), 9
(1 g, 5.81 mmol), and 1.0 mL triethylamine (7.2 mmol).
The solution was refluxed for four days, cooled, treated
with water, and stirred for 18 h at rt. The usual workup
gave polymer 11 as a yellow solid: FTIR (powder) 1635,
Jˆ6.0, 10.5 Hz), 2.58 (dd, 1H, Jˆ1.5, 6.3 Hz), 3.17 (m,
2H), 3.50–3.80 (m, 2H), 4.00–4.50 (m, 2H), 6.99–7.68
(m, 5H); 13C NMR d 8.2, 24.6, 29.7, 37.5, 38.0, 38.4,
40.8, 120.6/121.3, 124.4, 126.3/128.0, 129.0/129.3, 130.5;
HRMS: calcd for C14H16INO 341.0277, found 341.0265.
(2Rء,7Rء)- and (2Rء,7Sء)-2-(Cyanomethyl)-7-(iodomethyl)-
oxepane (15). Yield in solution from 13, 52%; yield via
1
1253, 1066, 915, 868, 841 cmϪ1
.
polymer 14, 8.1%; FTIR (neat) 2251 cmϪ1; H NMR d
0.75–1.61 (m, 8H), 2.47–2.55 (m, 1H), 2.97–3.06 (m,
2H), 3.11–3.17 (m, 1H), 3.60–4.00 (m, 2H); 13C NMR: d
11.2, 11.3, 25.1, 25.2, 26.1, 36.3, 72.7, 81.4, 119.0; HRMS:
calcd for C9H14INO 279.0129; found 279.0120; Anal. calcd
for C9H14INO: C, 38.73; H, 5.06; N, 5.02; found: C, 38.47;
H, 5.09; N, 4.97.
5-(5-Hexenyl)-4,5-dihydro-3-{phenyl[(trimethylsilyl)oxy]-
methyl}isoxazole (13). A solution of 1.11 g of 2-nitro-1-
phenyl-1-[(trimethylsilyl)oxy]ethane (4.6 mmol), 3.71 g of
phenylene diisocyanate (23.0 mmol), 2.04 g of 1,7-octa-
diene (8.5 mmol), and 0.5 mL of triethylamine in 50 mL
benzene was refluxed for three days, cooled, treated with
water, and stirred for 18 h at rt. The mixture was filtered,
and the volatiles evaporated. The product was purified by
column chromatography (10% EtOAc/hexane) to yield
0.61 g (40% yield) of a clear liquid, which proved to be a
1:1 mixture of diastereomers of 13: 1H NMR d 0.05 (s, 9H),
1.40–1.90 (m, 6H), 2.10–2.25 (m, 2H), 2.41 (dd, 1H, Jˆ8.7,
17.0 Hz), 2.82 (d, 1H, Jˆ9.3 Hz), 3.22 (dd, 1H, Jˆ10.2,
16.0 Hz), 4.65 (m, 1H), 5.13 (m, 2H), 5.90 (m, 1H), 7.40–
7.65 (m, 5H); 13C NMR d Ϫ0.34, 24.4, 24.6, 28.3, 33.3,
34.9, 36.9, 69.4, 80.4, 114.3, 120.5, 128.0, 128.9, 136.9,
140.0, 160.0; HRMS: calcd for C19H29NO2Si 331.1968,
found 331.1956.
(5Rء,6Sء,8Sء)- and (5Rء,6Sء,8Rء)-8-(Cyanomethyl)-5-iodo-
1,7-dioxaspiro[5.4]decane (18).6 Yield via polymer 17,
1
7.4%; FTIR (neat) 2251 cmϪ1; H NMR d 1.32–2.06 (m,
8H), 2.51 (dd, 1H, Jˆ5.7, 10.5 Hz), 2.60 (dd, 1H, Jˆ1.5,
6.0 Hz), 2.72–2.82 (m, 1H), 3.15–3.19 (m, 2H), 3.49–3.76
(m, 2H), 4.40 (m, 1H); 13C NMR d 21.0/25.3, 26.0/26.8,
29.9/30.7, 34.4/34.5, 38.2/39.4, 62.8, 72.8, 75.3, 106.7,
117.6/117.7; HRMS: calcd for C10H14NO2 (MϪI)ϩ
180.1024, found 180.1024.
Acknowledgements
This work was supported by a grant from the National
Science Foundation (CHE-97-10286).
Polymer-supported 5-(5-hexenyl)-4,5-dihydro-3-{phenyl-
[(trimethylsilyl)oxy]methyl}isoxazole (14). To 1.1 g of
polymer-bound nitrosilyl ether and 10 mL of benzene was
added 1 mL phenyl isocyanate (9.2 mmol), 1.1 g 1,7-octa-
diene (10.0 mmol), and 0.5 mL triethylamine (3.6 mmol).
The solution was refluxed for four days, cooled, treated
with water, and stirred for 18 h at rt. The usual workup
gave polymer 14 as a yellow solid: FTIR (powder) 1645,
References
1. Ellman, J. A.; Plunkett, M. J. J. Org. Chem. 1995, 60, 6006.
2. (a) Mayer, J. P.; Zhang, J.; Bjergarde, K.; Lenz, D.; Gaudino, J.
Tetrahedron Lett. 1996, 37, 8081. (b) Tietze, L. F.; Steinmetz, A.
Synlett 1996, 667.
1254, 1069, 876, 840 cmϪ1
.
3. (a) Kurth, M. J.; Rodriguez, M. J. J. Am. Chem. Soc. 1987, 109,
7577. (b) Kurth, M. J.; Rodriguez, M. J. Tetrahedron 1989, 45,
6963. (c) Kurth, M. J.; Rodriguez, M. J.; Olmstead, M. M. J. Org.
Chem. 1990, 55, 283. (d) Kurth, M. J.; Kantorowski, E. J. Mol.
Diversity 1996, 2, 207.
4. (a) Beebe, X.; Kurth, M. J.; Schore, N. E. J. Am. Chem. Soc.
1992, 114, 10 061. (b) Beebe, X.; Schore, N. E.; Kurth, M. J.
J. Org. Chem. 1995, 60, 4196.
5. (a) Evans, D. E.; Sacks, W. A.; Kleschick, W. A.; Taber, T. R.
J. Am. Chem. Soc. 1979, 101, 6789. (b) Hughes, D. L. Tetrahedron
Lett. 1978, 19, 3959. (c) Cresp, T. M.; Probery, C. L.; Sondheimer,
F. Tetrahedron Lett. 1978, 19, 3955. (d) Evans, D. E.; Sacks, W.
A.; Whitney, R. A.; Mandel, N. G. Tetrahedron Lett. 1978, 19, 727.
6. Rodriguez, M. J. Ph.D. Thesis, University of California, Davis,
1988.
7. Julia, M.; Ward, D. Bull. Soc. Chim. Fr. 1973, 3065.
8. Hanessian, S.; Lavallee, P. Can. J. Chem. 1975, 53, 2975.
9. Corey, E. J.; Chaykovsky, M.; Greenwald, R. J. Org. Chem.
1963, 28, 1128.
Polymer-supported 4,5-dihydro-5-[2-(4,5-dihydro-2H-py-
ranyl)ethyl]-3-{phenyl[(trimethylsilyl)oxy]methyl}isoxa-
zole (17). To 1.5 g of polymer-bound nitrosilyl ether and
10 mL of benzene was added 1 mL phenyl isocyanate
(9.2 mmol), 0.9 g 6-(3-butenyl)-3,4-dihydro-2H-pyran
(6.5 mmol), and 1.0 mL triethylamine (7.2 mmol). The
solution was refluxed for four days, cooled, treated with
water, and stirred for 18 h at rt. Workup gave polymer 17
as a yellow solid: FTIR (powder) 1268, 1067, 876,
841 cmϪ1
.
General procedure for polymer-supported cyclic ether
formation
The appropriate polymer was swollen in dry CH2Cl2 and
treated with 5 equiv. of I2. The reaction was then stirred
for 18 h at rt. The formation of cyclic ethers was monitored
by TLC (9:1 hexane/EtOAc). Sat. aq. Na2S2O3 was added,
and the mixture stirred overnight. The polymer was filtered,
and the organic layer was washed with sat. aq. Na2S2O3,
dried (MgSO4), and evaporated to yield an oily residue.
10. Greeves, N.; Lee, W. Tetrahedron Lett. 1997, 38, 6449.
11. Boeckman, R. K.; Bruza, K. J. Tetrahedron 1981, 37, 3997.