Oxidatively generated electrophiles as initiators of epoxide cascade cyclization processes

Article abstract of DOI:10.1016/j.tetasy.2005.08.055

Oxidatively generated oxocarbenium ions are shown to be effective promoters of polyepoxide cascade cyclization reactions to form polyether compounds. The reaction conditions are neutral, ensuring that background acid-mediated processes are not operative a

Full text of DOI:10.1016/j.tetasy.2005.08.055

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 3570–3578
Oxidatively generated electrophiles as initiators of epoxide
cascade cyclization processes
V. Satish Kumar, Shuangyi Wan, Danielle L. Aubele and Paul E. Floreancig*
Department of Chemistry, University of Pittsburgh, Pittsburgh, PA 15260, USA
Received 30 July 2005; accepted 17 August 2005
Available online 2 November 2005
Abstract—Oxidatively generated oxocarbenium ions are shown to be effective promoters of polyepoxide cascade cyclization reac-
tions to form polyether compounds. The reaction conditions are neutral, ensuring that background acid-mediated processes are
not operative and that other acid-sensitive functional groups, such as acetals, can be incorporated into cyclization substrates. While
5-exo pathways are more common that 6-endo pathways, a rational design has been employed to access tetrahydropyranyl ethers.
Ó 2005 Elsevier Ltd. All rights reserved.
1. Introduction
reaction. In this approach, a single epoxide reacts with
an electrophile to form an epoxonium ion,7 with selec-
tivity arising from simple cyclization kinetics. This
epoxonium ion can then react with another epoxide to
continue the cascade. To ensure selectivity in such reac-
tions, a non-acidic method for forming the initial elec-
trophile is desirable. Oxidative electrophile formation
(Fig. 2) is well suited for this purpose, although only
limited examples of this strategy have been reported.
In a recent synthesis of hemibrevetoxin, Holton utilized
selenonium ion formation as a method for intramolecu-
Cascade reactions, in which the formation of a single
reactive intermediate leads to multiple bond forming
events provide significant increases in molecular com-
plexity, making them quite useful for applications in
organic synthesis. Numerous reports of cation, anion,
and radical mediated cascade processes have appeared
in the literature,1 with alkenes or other p-bonds serving
to propagate the reactive intermediate, leading to the
construction of carbocyclic structures. Transition metal
alkene and alkyne complexation is also emerging as an
intriguing method for initiating cascade processes.2 Cyc-
lic ethers have been prepared through the use of epox-
ides in cascade processes under either acidic or basic
conditions (Fig. 1).
8
9
10
´
lar epoxide activation, while Martın and McDonald
have used halonium ions in intramolecular Lewis acid
activation.11 An alternative approach to non-acidic
polyether synthesis involves sequential olefin oxidation
with metal oxides.12
Acid-mediated intramolecular additions into epoxides
have attracted considerable attention in polyether syn-
thesis, because of their capacity to alter the regiochemi-
cal outcome from the stereoelectronically preferred5
exo-pathway to the endo-pathway through substitution.6
In the context of cascade cyclizations, however, treating
a polyepoxide with an acid can result in non-selective
acid–base interactions, leading to potential undesired
pathways and to ambiguous mechanistic interpretations.
An approach to alleviating these problems is to use an
intramolecular acid–base interaction to promote the
We have developed a method for converting homobenz-
ylic ethers and amides to oxocarbenium and acylimin-
ium ions, respectively, through single electron
oxidation (Fig. 3).13 This method of carbon–carbon
bond activation proceeds under non-acidic conditions,
making chemoselectivity for electrophile formation
exceptionally high, even in the presence of functional
groups which are readily activated under conventional
acidic conditions. Therefore, oxidatively formed
oxocarbenium ions should function as excellent initia-
tors for polyepoxide cascade cyclization reactions. The
products of these reactions would contain a latent alde-
hyde group, creating additional benefit to this process.
Herein, we report our studies on oxidatively initiated
polyepoxide cascade reactions to form tetrahydrofurans
*
Corresponding author. Tel.: +1 412 624 8727; fax: +1 412 624
8611; e-mail: florean@pitt.edu
0957-4166/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2005.08.055

V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
3571
O
O
HO
OH
O
O
O
O
CSA
44%
O
O
H
H
H
H
OH
O
HO
HO
OH
HO
OH
O
Saponification
then HOAc
94%
O
O
H
H
H
O
O
O
O
O
OH
HO
Figure 1. Literature examples of epoxide cascade cyclizations.3,4
tert-butylbenzene), yielded tetrahydrofuran 2 in good
yield with high stereocontrol. This process proceeds
through benzylic carbon–carbon bond cleavage to form
oxocarbenium ion 3. Cyclization of the proximal oxygen
of the acetonide provided bicyclic oxonium ion 4, in
which the alkoxy group is preferentially oriented on
the convex face. Hydrolytic breakdown of 4 yields 2.
O
O
Oxidation
X
O
O
O
Intramolecular
epoxide activation
O
The success of this reaction led us to consider the possi-
bility of using other cyclic ethers as nucleophiles in these
reactions. Epoxides were particularly desirable in this
regard due to their expected capacity to serve as potent
electrophiles and engage in further bond formation
upon intramolecular reaction with the oxocarbenium
ion. Indeed, single electron oxidation of epoxide 5
provided hydroxytetrahydropyranyl ether 6 in moderate
yield, but with high stereocontrol through the formation
of epoxonium ion 7, which reacted with adventitious
water at the secondary carbon. In both of these exam-
ples, the intermediacy of the oxocarbenium ion was
inferred from the observation that cyclization reactions
of 8 and 9 yielded identical diastereomer ratios to the
cyclizations of 1 and 5. The efficient cyclization of epoxy
alcohol 10 to yield 11 provided further evidence that
epoxides do not undergo rapid decomposition under
the oxidative conditions.
Figure 2. Oxidatively-initiated intramolecular epoxide activation.
-
-1 e
R
R
Y
Y
R
Y = OR' or NHC(O)R'
+
Y
Figure 3. Electrophile formation through carbon–carbon bond
activation.
and tetrahydropyrans, in which relative and, in some
cases, absolute stereochemistry in the products is con-
trolled by the stereochemistry of the epoxides. In
addition to epoxides, we also highlight the utility of
acid-sensitive acetals as nucleophiles under oxidative
conditions.14
While not proceeding in exceptional yield, the cyclization
of 5 led us to consider the viability of incorporating an
additional nucleophile into the substrate in order to trap
the intermediate epoxonium ion and form a second ring.
Of additional interest herein was the effect of activating
the epoxide through complexation with an oxocarbe-
nium ion on cyclization regiochemistry (exo vs endo)
and stereochemistry (SN1 v s S 2). Thus, racemic epoxy
2. Results and discussion
N
alcohol 15a was prepared through a straightforward se-
quence (Scheme 2) and subjected to the aerobic variant15
of our oxidative cyclization conditions to provide 16 and
17 in a 3:1 ratio in 51% overall yield, indicating that
exo- and endo-pathways are competitive for this sub-
strate. The relationship between the regiochemistry and
Our initial efforts13a in this project (Scheme 1) arose
from our observation that acetonide 1, when exposed
to photoinitiated single electron oxidation conditions
(hm, Pyrex filtration, N-methylquinolinium hexafluoro-
phosphate (NMQPF6), NaOAc, 1,2-dichloroethane,

3572
V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
hν, NMQPF
6
O
NaOAc, DCE
OH
O
Bn
11:1 d.r.
H C O
17 8
tert-butylbenzene
74%
O
OC H
8
17
1
2
H O
2
O
H
C O
8
17
O
O
O
O
OC H
8
17
3
4
h , NMQPF
ν
6
OH
NaOAc, DCE
tert-butylbenzene
49%
Bn
19:1 d.r.
OC H
H
C O
O
8
17
17 8
5
6
H O
2
H
H
C O
O
17
8
7
h , NMQPF
ν
6
O
Bn
C O
NaOAc, DCE
OH
2.6:1 d.r.
tert-butylbenzene
78%
O
H
17
8
H
C O
O
17
8
10
11
O
O
Bn
O
Bn
OC H
8
17
OC H
8
17
8
9
Scheme 1. Cyclic ethers as nucleophiles in oxidative cyclizations.
reaction conditions was studied by changing the base
from NaOAc to soluble, oxidatively stable 2,6-dichloro-
pyridine. In this case, the observed ratio of 16/17 was 1:1
and the overall yield was 64%. Additionally, THP/ether
15b, prepared in accord with our observations that ace-
tals can be surrogates of hydroxyl groups in these pro-
cesses, was subjected to the standard oxidation
conditions to provide 16 in 83% yield with no detectable
endo-product. Therefore, while the effect of changing the
nucleophile on the regiochemistry of the cyclization is
not understood, these results clearly indicate that the
nucleophileꢀs identity does impact the reaction outcome.
ceeds stereoselectively. Evidence for stereospecificity in
the cyclization was provided through the oxidation of
cis-epoxide 22. This reaction provided 23 as a diastereo-
meric mixture again with no detectable endo-cyclization.
Oxidation of 23 provided lactone 24, which was clearly
spectroscopically distinct from 21.
With the basic pathway established, we turned our
attention to incorporate an additional epoxide into the
substrate to create the possibility for an additional ring
forming event and to determine whether epoxonium
ions that form from the coupling of epoxides to second-
ary carbocations differ in reactivity from those formed
from the addition of epoxides into oxocarbenium ions
(Fig. 4). As an additional exploratory element, we
wanted to examine the possibility of using mixed acetals
to open epoxonium ions,16 thereby delivering an alkoxy
group with high regiocontrol.
Bis-tetrahydrofuran 16 was isolated as a 1:1 mixture of
stereoisomers. To determine whether the lack of stereo-
control resulted from a lack of anomeric control or from
non-stereoselective epoxonium ion opening, both iso-
mers were oxidized with Jones reagent to provide lac-
tone 21. Therefore, while anomeric stereocontrol was
not controlled at the high level that was observed in
the cyclization of 5, opening the epoxonium ion pro-
Ester 20, an intermediate in the synthesis of 15, was con-
verted to diene 25, in which the ethoxyethyl ether was

V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
3573
O
O
f-g. or
f,h,g
a-c
d-e
Bn
O
H
OEt
OC H
OC H
8
17
8
17
18
19
20
H
O
Bn
i
OR
+
O
H
C O
8
17
O
O
H
OC H
8
17
H
H
C O
8
O
17
H
15a: R = H
15b: R = THP
16
17
O
i
Bn
OTHP
H
C O
8
17
O
O
H
H
OC H
8
17
22
23
O
O
O
O
O
O
H
H
H
H
21
24
Scheme 2. Prototype oxidative cascade cyclizations. Reagents and conditions: (a) allylmagnesium bromide, Et2O, ꢀ78 °C, 80%; (b) NaH, DMF, then
n-C8H17I, 60%; (c) O3, CH2Cl2, ꢀ78 °C, then Ph3P; (d) vinylmagnesium bromide, THF, ꢀ78 °C, 71% for two steps, (e) (EtO)3CCH3, propionic acid,
xylenes, reflux, 75%; (f) LiAlH4, Et2O, 96%; (g) m-CPBA, NaOAc, CH2Cl2, 83%; (h) dihydropyran, PPTs, CH2Cl2, 76%; (i) see text.
ence for tetrahydrofuran formation over tetrahydropy-
ran formation because of the stereoelectronic prefe-
rence for exo-attack on the epoxonium ion relative to
endo-addition. Based on these considerations, we postu-
lated that tetrahydropyran formation would be possible
through changing the topography of the substrate by
placing the terminal nucleophile between the homobenz-
ylic ether and the epoxide. Therefore, upon epoxonium
ion formation, the nucleophile would be constrained
to react through a fused bicyclic transition state rather
than a bridged transition state, leading to tetrahydropy-
ran formation (Fig. 5).
Nu
R
Nu
RO
O
O
n
n
Figure 4. Epoxonium ions formed from additions to oxocarbenium
and secondary carbenium ions.
incorporated to be the terminal nucleophile in the cas-
cade process. In order to avoid the formation of diaste-
reomeric bis-epoxides, 25 was exposed to Shiꢀs
conditions17,18 to yield 26 (Scheme 3).19
Subjecting 26 to single electron oxidation conditions
resulted in the smooth formation of bis-tetrahydrofuran
27 as a mixture of two diastereomers. Again the ano-
meric position was identified as the site of stereochemi-
cal ambiguity through acylation the primary hydroxyl
group followed by oxidation of with the Jones reagent,
thus forming lactone 28 as a single stereoisomer. This
reaction demonstrated that the epoxonium ion opening
proceeds stereoselectively with epoxide nucleophiles,
that no compelling reactivity difference in reactivity
between epoxonium ions, that were formed from reac-
tion with oxocarbenium ions relative to non-stabilized
carbenium ions could be determined, and that mixed
acetals are effective at delivering alkoxy groups to elec-
trophilic carbons. To further demonstrate the stereo-
selectivity of the process, we converted diene 29 to
diastereomeric bis-epoxide 30 through sequential Sharp-
less20 and Shi epoxidations. This substrate proceeded
efficiently through cyclization to provide 31 as a mixture
of anomers. Acylation and oxidation yielded lactone 32
as a single isomer.
Methyl olivoside 33 was selected as a synthetic target to
test this hypothesis. In our approach (Scheme 4) 33
arises from deprotection of the C5 ether group of 34.
The ether can be installed through the delivery of an alk-
oxy group to an oxidatively generated epoxonium ion
through a mixed acetal that is present in cyclization sub-
strate 35. Phenylacetaldehyde was seen as a suitable pre-
cursor for the stereoselective synthesis of 35.
Thus, commercially available phenylacetaldehyde was
subjected to a Brown allylation reaction21 to provide
the expected homoallylic alcohol, which was converted
to methyl ether 36. Ozonolysis followed in the same
flask by the addition of propynylmagnesium bromide
yielded a 3:1 mixture of propargylic alcohols favoring
the syn-isomer, as determined by subsequent transfor-
mations, in 68% yield (unoptimized). Reduction of the
major product with LiAlH4 gave allylic alcohol 37 in
85% yield. Diastereocontrolled epoxide formation was
best achieved using Sharplessꢀ conditions.20 It is worthy
of note that this reaction never proceeded to completion,
most likely due to the epoxidation performing a kinetic
resolution22 concurrent with the diastereoselective oxi-
dation. In consideration of our objective of delivering
Based on substitution patterns, both carbons of the
epoxonium ions herein should be equally electrophilic.
The cascades that have been described show a prefer-

3574
V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
O
O
a-c
d
Bn
Bn
20
O
OEt
O
OEt
OC H
OC H
8
17
8
17
25
26
O
e
H
C O
OH
17
8
H
C O
O
17
8
O
O
O
O
H
H
H
H
H
H
H
Et
OEt
27
O
O
Bn
f-h
Bn
OH
O
OEt
OC H
8
17
OC H
8
17
29
30
i
H
C O
8
OH
17
O
O
H
H
H
OEt
31
O
O
O
O
O
OAc
OAc
O
O
O
O
H
H
H
H
H
H
O
O
OEt
OEt
O
28
32
Shi catalyst
Scheme 3. Cascade cyclizations of di-epoxides with excellent stereocontrol. Reagents and conditions: (a) DIBAL-H, CH2Cl2, ꢀ78 °C, then
Ph3PCHCO2Et, rt, 87%; (b) DIBAL-H, CH2Cl2, ꢀ78 °C, 81%; (c) Shi catalyst, oxone, K2CO3, Bu4NHSO4, EDTA, H2O, CH3CN, 85%; (d) hm,
˚
NMQPF6, O2, NaOAc, Na2S2O3, PhMe, DCE, 66%; (f) (+)-diisopropyl tartrate, t-BuOOH, Ti(Oi-Pr)4, 4 A MS, CH2Cl2, 93%; (g) 1-chloroethyl
ethyl ether, N,N-dimethylaniline, CH2Cl2, 94%; (h) Shi catalyst, oxone, K2CO3, Bu4NHSO4, EDTA, H2O, CH3CN, 64%; (i) hm, NMQPF6, O2,
NaOAc, Na2S2O3, PhMe, DCE, 69%.
expected to inhibit arene oxidation and trifluoromethyl-
benzyl (TFBn) ethers have been shown to be cleaved
NuX
+
Nu
-X
through hydrogenolysis.23 Thus, TFBnOCH(CH3)Cl,
which was prepared from TFBnOH and acetaldehyde
in the presence of HCl, was coupled with the epoxide
of 37 in the presence of N,N-dimethylaniline to provide
cyclization substrate 38. Exposure of 38 to our standard
oxidative cyclization resulted in the formation of 39 in
48% yield. Variable amounts of tetrahydrofuran 40 were
also observed if the reaction was not conducted under
anhydrous conditions. Stereochemical confirmation of
the final product was obtained through cleaving the
TFBn ether to yield 33, which showed identical NMR
spectra to a literature report (Scheme 5).24
RO
O
RO
O
Figure 5. Strategy for tetrahydropyran formation.
OH
O
OH
O
OH
OR
MeO
MeO
33
34
It is noteworthy that in the cyclization of 38 to 39, no
direct delivery of the TFBnO-group to the initially
formed oxocarbenium ion was observed, even though
the formation of this product would proceed through
a favorable 6-exo cyclization pathway. Two explana-
tions for this are possible. The geometrical constraints
of the epoxide could make the lone pairs on the oxygen
more kinetically accessible even though, from a thermo-
dynamic perspective, they should be less reactive than
standard ethers due to the higher s-character of their
orbitals. Alternatively, the trifluoromethylbenzyloxy
group could serve as the initial nucleophile into the
epoxonium ion to form an oxonium ion intermediate.
If the alkoxy group transfer is slow relative to the refor-
mation of the oxocarbenium ion, then the epoxide could
RO
O
H
O
Ph
O
MeO
Bn
35
Scheme 4. Retrosynthesis of methyl olivoside.
an ether group that would provide a hydroxyl group
upon exposure to mild cleavage conditions while not
being cleaved in the oxidative cyclization. A p-trifluo-
romethylbenzyloxyethyl ether appeared to be well suited
for this purpose since the trifluoromethyl group was

V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
3575
H
Bn
Bn
MeO
a-b
c-d
O
OMe
OH
37
36
O
Bn
RO
O
g
e-f
O
MeO
O
OTFBn
MeO
Bn
38
OTFBn
OH
O
OTFBn
+
MeO
O
H
MeO
O
OH
39
40
Scheme 5. Oxidative carbohydrate synthesis. Reagents and conditions: (a) (+)-Ipc2BCH2CH@CH2, CH2Cl2, ꢀ78 °C; (b) NaH, DMF, then MeI,
82%; (c) O3, CH2Cl2, ꢀ78 °C, then Ph3P, then propynylmagnesium bromide, 51% + 17% anti-diastereomer; (d) LiAIH4, Et2O, reflux, 85%; (e) Ti(Oi-
Pr)4, (+)-diisopropyl tartrate, t-BuOOH, CH2Cl2, ꢀ20 °C, 91%; (f) TFBnOCH(Cl)CH3, N,N-dimethylaniline, CH2Cl2, 93%; (g) hm, NMQPF6, O2,
NaOAc, Na2S2O3, PhMe, DCE, 46%.
O
eventually react to form the epoxonium ion. Due to the
a-c
d-e
EtO
strain of the epoxonium ion, rapid and essentially irre-
versible opening by the trifluoromethylbenzyloxy group
occurs (Fig. 6).
OTBS
O
O
42
O
f-h
Another strategy for promoting tetrahydropyran forma-
tion is to disrupt the degenerate reactivity of the carbons
in the epoxonium ion through substitution. Towards
this goal, trisubstituted epoxide substrate 41 was pre-
pared from butyrolactone as shown in Scheme 6, with
the expectation that the increased cationic character at
the tertiary center, that would be realized during epoxo-
nium ion formation, would promote endo-cyclization.
Indeed, exposing 41 to our standard cyclization condi-
tions resulted in the formation of 44, albeit in 620%
yield. The structural assignment for this compound
was based on the chemical shift of the hydrogen on
the anomeric carbon at 4.65 ppm,25 characteristic of tet-
rahydropyranyl ethers. Interestingly, the THP ether of
44 was a poor cyclization substrate, showing evidence
of an epoxide rearrangement to form a ketone. This
product is likely to result from the intermediate epoxo-
nium ion undergoing a rearrangement faster than nucleo-
philic addition. Therefore, while trisubstituted
epoxonium ions show a greater tendency to react
through the endo-pathway, a sufficiently reactive nucleo-
Bn
Bn
OH
OMe
OTBS
i
OH
43
41
O
MeO
O
H
44
Scheme 6. Regiochemical reversal through substitution. Reagents and
conditions: (a) DIBAL-H, CH2Cl2, ꢀ78 °C, then isopropenylmagne-
sium bromide, 61%; (b) TBSCl, imidazole, DMF, 82%; (c)
(EtO)3CCH3, propionic acid, 140 °C, 88%; (d) DIBAL-H, CH2Cl2,
ꢀ78 °C; (e) BnMgCl, CuCN, BF3ÆOEt2, LiCl, THF, ꢀ78 °C, 68% for
two steps; (f) NaH, DMF, then MeI; (g) Bu4NF, THF, 95% for two
steps; (h) m-CPBA, NaHCO3, CH2Cl2, 0 °C, 85%; (i) hm, NMQPF6,
O2, NaOAc, Na2S2O3, PhMe, DCE, <20%.
phile must be employed to avoid competitive isomeriza-
tion. We are currently pursuing this strategy.
O
O
O
MeO
MeO
MeO
faster
slower
faster
TFBnO
O
TFBnO
O
TFBnO
O
O
O
O
OTFBn
H
O
O
slower
OTFBn
OTFBn
MeO
O
MeO
MeO
Figure 6. Fates of oxonium ion intermediates in the cyclization.

3576
V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
3. Conclusion
80.5, 79.7, 68.5, 67.3, 31.9, 31.8, 29.7, 29.2, 28.1, 26.1,
25.6, 25.4, 22.6, 14.1; IR (neat) 2965, 2856, 1454, 1329,
1193, 1079 cmꢀ1; HRMS (EI) calcd for C12H23O2
(Mꢀ71) 199.1698, found 199.1695. Slower eluting com-
Oxidatively generated electrophiles have been effectively
utilized for intramolecular epoxide activation. This
method employs essentially neutral conditions and offers
the desirable attribute of controlling the sequence of
functional group activation when several acid-sensitive
groups are present. Cascade reactions to form oligo-
tetrahydrofuran products that demonstrated a strong
preference for the exo-cylization pathway were achieved
in good yields when disubstituted epoxides were used as
substrates. High stereoselectivity was observed in these
reactions, with complementary diastereomers being
formed from diastereomeric epoxides. Mixed acetals
were also shown to be suitable nucleophiles for opening
epoxonium ions, resulting in a stereoselective alkoxy
group delivery. Two strategies for tetrahydropyran for-
mation using this strategy have been devised. Installing
the terminal mixed acetal nucleophile between the
oxocarbenium ion and the epoxide was used in the enan-
tioselective synthesis of methyl olivoside. Trisubstituted
epoxides have been employed in these reactions to
promote the endo-cyclization pathway relative to the
exo-pathway. In these reactions, the intermediate
epoxonium ions are prone to isomerization, requiring
that nucleophiles are sufficiently reactive to engage in
cyclization faster than the undesired side reaction.
1
pound (19 mg, 35.4%): H NMR (300 MHz, CDCl3) d
5.06 (dd, J = 3.8, 2.7 Hz, 1H), 3.98 (td, J = 7.8, 6.4,
1H), 3.92–3.79 (m, 2H), 3.78–3.71 (m, 1H), 3.64 (td,
9.5, 6.9 Hz, 1H), 3.33 (td, J = 9.5, 6.6 Hz, 1H), 2.04–
1.81 (m, 8H), 1.55 (quintet, J = 7.4 Hz, 2H), 1.38–1.21
(m, 10H), 0.88 (t, J = 6.5 Hz, 3H); 13C NMR
(75 MHz, CDCl3) d 103.9, 82.2, 82.0, 68.4, 67.2, 32.9,
31.8, 29.6, 29.4, 29.2, 27.9, 27.3, 26.2, 25.8, 22.6, 14.0;
IR (neat) 2965, 2856, 1454, 1329, 1193, 1079 cmꢀ1
;
HRMS (EI) calcd for C12H23O2 (Mꢀ71) 199.1698,
found 199.1695.
4.3. Bis-furan 23
To epoxy-ether 22 (50 mg, 0.11 mmol) in dichloroethane
(6 mL) in a borosilicate flask at room temperature were
added
N-methylquinolinium
hexafluorophosphate
(0.8 mg, 0.003 mmol), sodium acetate (100 mg, 1.21
mmol), anhydrous Na2S2O3 (100 mg, 0.63 mmol),
and toluene (1 mL). The mixture was photoirradiated
with gentle air bubbling for 1.5 h, while stirring at room
temperature. The reaction mixture was filtered through
a small plug of silica gel and then concentrated. The
resulting residue was purified by flash chromatography
(5% EtOAc in hexanes) to provide the desired com-
pound as two diastereomers. Faster eluting compound
4. Experimental
1
(12.5 mg, 41.3%): H NMR (300 MHz, CDCl3) d 5.17
4.1. General procedures
(dd, J = 4.8, 1.4 Hz, 1H), 4.26 (dd, J = 13.4, 6.9 Hz,
1H), 3.95–3.88 (m, 1H), 3.82–3.75 (m, 2H), 3.72 (td,
J = 9.5, 6.9 Hz, 1H), 3.37 (td, J = 9.5, 6.9 Hz, 1H),
2.10–1.81 (m, 5H), 1.62–1.53 (m, 5H), 1.38–1.19 (10H),
0.88 (t, J = 6.9 Hz, 3H); 13C NMR (75 MHz, CDCl3)
d 104.2, 81.4, 80.1, 68.1, 67.4, 32.1, 31.8, 29.7, 29.4,
29.2, 28.1, 26.1, 25.7, 22.6, 14.1; IR (neat) 2926, 2856,
1730, 1459, 1347, 1318, 1197, 1078, 1038, 990,
856 cmꢀ1; HRMS (EI) calcd for C16H29O3 (Mꢀ1)
269.2117, found 269.2125. Slower eluting compound
(11.8 g, 39.0%): 1H NMR (300 MHz, CDCl3) d 5.23
(d, J = 3.8 Hz, 1H), 4.12–3.92 (m, 4H), 3.88 (td,
J = 9.6, 6.9 Hz, 1H), 3.49 (td, J = 9.6, 6.7 Hz, 1H),
2.14–1.96 (m, 5H), 1.95–1.80 (m, 1H), 1.70–1.58 (m,
4H), 1.50–1.36 (m, 10H), 1.02 (t, J = 6.9 Hz, 3H); 13C
NMR (75 MHz, CDCl3) d 103.9, 83.1, 82.8, 68.3, 67.3,
33.1, 31.8, 29.6, 29.4, 29.2, 27.9, 26.2, 25.8, 22.6, 14.1;
IR (neat) 3026, 2928, 2855, 1602, 1495, 1454, 1378,
1346, 1136, 1098, 1057, 1134, 1098, 1057, 943, 930,
894, 870 cmꢀ1; HRMS (EI) calcd for C16H29O3 (Mꢀ1)
269.2117, found 269.2117.
THF and Et2O were dried by distillation over Na/benzo-
phenone under N2. Toluene, 1,2-dichloroethane, and
CH2Cl2 were dried by distillation over CaH2. Photoirra-
diations were performed with a Hanovia 450 W medium
pressure mercury lamp through Pyrex filtration.
Reagents were used, as purchased, without further puri-
fication unless otherwise noted. Chemical shifts (d) are
reported in parts per million.
4.2. Bis-furan 16
To epoxy-ether 15b (90 mg, 0.20 mmol) in dichloroeth-
ane (12 mL) in a borosilicate flask at room temperature
were added N-methylquinolinium hexafluorophosphate
(1.4 mg, 0.005 mmol), sodium acetate (180 mg,
2.19 mmol), anhydrous Na2S2O3 (180 mg, 1.13 mmol),
and toluene (2 mL). The mixture was photoirradiated
with gentle air bubbling for 2 h while stirring at room
temperature. The reaction mixture was filtered through
a small plug of silica gel and the filtrate concentrated.
The resulting residue was purified by flash chromatogra-
phy (5% EtOAc in hexanes) to provide the desired com-
pound as two diastereomers. Faster eluting compound
(21 mg, 37.6%): 1H NMR (300 MHz, CDCl3) d 5.14
(dd, J = 4.9, 1.7 Hz, 1H), 4.03 (td, J = 12.9, 4.5 Hz,
1H), 3.91–3.82 (m, 2H), 3.76 (td, J = 8.2, 6.5 Hz, 1H),
3.66 (td, J = 9.6, 6.9 Hz, 1H), 3.36 (td, J = 9.6, 6.7 Hz,
1H), 2.19–1.84 (m, 6H), 1.74–1.62 (m, 2H), 1.52 (quin-
tet, J = 6.9 Hz, 2H), 1.38–1.21 (m, 10H), 0.88 (t,
J = 6.0 Hz, 3H); 13C NMR (75 MHz, CDCl3) d 104.1,
4.4. Bis-furan 27
To diepoxide 26 (50 mg, 0.10 mmol) in dichloroethane
(6 mL) in a borosilicate flask at room temperature were
added
N-methylquinolinium
hexafluorophosphate
(0.7 mg, 0.0026 mmol), sodium acetate (100 mg, 1.21
mmol), anhydrous Na2S2O3 (10 mg, 0.63 mmol), and
toluene (1 mL). The mixture was photoirradiated with
gentle air bubbling for 1.5 h while stirring at room
temperature. The reaction mixture was filtered through

V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
3577
a small plug of silica gel and the filtrate was concen-
trated. The resulting residue was purified by flash chro-
matography (10% EtOAc in hexanes) to provide the
filtered, and concentrated. The crude residue was dis-
solved in acetone (2 mL) and cooled to 0 °C. Jones
reagent (four drops) was then added. The reaction mix-
ture instantly turned from colorless to green and then
stirred for an additional 10 min. The reaction mixture
was concentrated and purified by flash chromatography
desired compound as a mixture of two diastereomers
25
(28 mg, 66.4%): ½aŠ ¼ ꢀ3:8 (c ꢀ0.45, EtOAc); 1H
NMR (300 MHz, CDDCl3) d 5.12 (dd, J = 4.9, 1.4 Hz,
0.5H), 5.05 (d, J = 4.8 Hz, 0.5H), 4.05–3.88 (m, 2.5H),
3.78–3.59 (m, 5.5H), 3.37–3.29 (m, 2H), 2.12–1.84 (m,
7H), 1.82–1.65 (m, 1H), 1.58–2.50 (m, 2H), 1.39–1.23
(m, 10H), 1.21 (t, J = 7.0 Hz, 3H), 0.88 (t, J = 6.7 Hz,
3H), 13C NMR (75 MHz, CDCl3) d 104.2, 82.5, 82.2,
81.5, 81.4, 81.3, 81.0, 79.6, 67.5, 67.4, 66.2, 66.1, 62.9,
62.8, 32.9, 31.9, 31.8, 29.8, 29.7, 29.4, 29.4, 28.4, 28.1,
27.5, 27.4, 26.9, 26.3, 26.2, 25.8, 22.6, 15.6, 13.9; IR
(neat) 3459, 2974, 2962, 2927, 2875, 2857, 1766, 1730,
1459, 1376, 1347, 1197, 1097, 1077, 1041, 935,
852 cmꢀ1 HRMS (EI) calcd for C12H21O4 (Mꢀ129)
229.1439, found 229.1439.
to provide the desired product (5.5 mg, 70%):
25
½aŠ ¼ ꢀ8:3 (c 0.40, EtOAc); 1H NMR (300 MHz,
D
CDCl3) d 4.42 (dd, J = 13.2, 6.4 Hz, 1H), 4.32 (dd,
J = 11.8, 3.7 Hz, 1H), 4.04 (dd, J = 11.8, 5.4 Hz, 1H),
4.01–3.95 (m, 2H), 3.72 (dq, J = 9.2, 7.0 Hz, 1H), 3.54
(dq, J = 9.2, 7.0 Hz, 1H), 3.48 (m, 1H), 2.58–2.50 (m,
2H), 2.32–2.26 (m, 1H), 2.14–1.93 (m, 4H), 2.08 (s,
3H), 1.85–1.75 (m, 1H), 1.92 (t, J = 6.9 Hz, 3H); 13C
NMR (75 MHz, CDCl3) d 177.0, 170.9, 81.2, 80.3,
79.4, 78.9, 77.8, 77.2, 76.4, 66.5, 63.8, 29.7, 28.1, 27.5,
26.8, 24.0, 15.5; HRMS (EI) calcd for C14H23O6
(M+1) 287.1495, found 287.1503.
4.5. Bis-furan 31
4.7. Lactone 32
To diepoxide 30 (50 mg, 0.10 mmol) in dichloroethane
(6 mL) in a borosilicate flask at room temperature were
To 31 (11 mg, 0.0275 mmol) in CH2Cl2 (2 mL) was
added Et3N, DMAP, and Ac2O. The mixture was stirred
for 3 h and then partitioned between CH2Cl2 and H2O.
The organic layer was dried over MgSO4, filtered, and
concentrated. The crude residue was dissolved in ace-
tone (2 mL) and cooled to 0 °C. Jones reagent (four
drops) was added. The reaction mixture was turned
from colorless to green and then stirred for 10 min.
The reaction mixture was concentrated and purified by
added
N-methylquinolinium
hexafluorophosphate
(1.5 mg, 0.0052 mmol), sodium acetate (100 mg,
1.21 mmol), anhydrous Na2S2O3 (100 mg, 0.63 mmol),
and toluene (1 mL). The mixture was photoirradiated
with gentle air bubbling for 1.5 h, while stirring at room
temperature. The reaction mixture was filtered through a
small plug of silica gel and the filtrate concentrated. The
resulting residue was purified by flash chromatography
(10% EtOAc in hexanes) to provide the desired com-
flash chromatography to provide the desired product
25
(5 mg, 63%): ½aŠ ¼ ꢀ7:8 (c 0.45, EtOAc); 1H NMR
D
pound as a mixture of two diastereomers. Faster eluting
(300 MHz, CDCl3) d 4.43 (dd, J = 12.6, 6.9 Hz, 1H),
4.27 (dd, J = 10.2, 3.9 Hz, 1H), 4.11–3.99 (m, 3H),
3.78–3.65 (m, 1H), 3.63–3.51 (m, 1H), 3.52–3.44 (m,
1H), 2.63–2.47 (m, 2H), 2.36–2.26 (m, 1H), 2.18–1.90
(m, 4H), 2.08 (s, 3H), 1.78–1.65 (m, 1H), 1.19 (t,
J = 7.2 Hz, 3H); 13C NMR (75 MHz, CDCl3) d 177.0,
170.8, 81.6, 80.2, 79.6, 79.1, 66.7, 63.9, 28.1, 27.4, 23.8,
20.9, 15.6. HRMS (EI) calcd for C14H23O6 (M+1)
287.1495, found 287.1505.
25
compound (14 mg, 37.2%): ½aŠ ¼ ꢀ8:5 (c 0.30, EtOAc);
D
1H NMR (300 MHz, CDCl3) d 5.12 (dd, J = 5.1, 1.6 Hz,
1H), 4.01 (dd, J = 13.2, 6.0 Hz, 2H), 3.93 (dt, J = 13.2,
5.8 Hz, 1H), 3.73–3.58 (m, 5H), 3.39–3.30 (m, 2H),
2.10–1.67 (m, 8H), 1.59–1.51 (m, 2H), 1.32–1.24 (m,
10H), 1.20 (t, J = 7.0 Hz, 3H), 0.88 (t, J = 6.8 Hz, 3H);
13C NMR (75 MHz, CDCl3) d 104.2, 81.1, 81.0, 80.7,
79.8, 67.4, 66.3, 66.1, 32.8, 31.9, 31.8, 29.8, 29.7, 29.4,
29.2, 26.2, 25.7, 22.6, 15.7, 14.0; IR (neat) 3456, 2927,
2857, 1459, 1376, 1347, 1197, 1097, 1041, 936,
852 cmꢀ1; HRMS (EI) calcd for C12H21O4 (Mꢀ129)
229.1439, found 229.1430. Slower eluting compound
4.8. Protected methyl olivoside 39
To epoxide 38 (50 mg, 0.11 mmol) in dichloroethane
(6 mL) in borosilicate flask at room temperature were
25
D
1
(12 mg, 31.9%): ½aŠ ¼ ꢀ1:5 (c 0.15, EtOAc); H NMR
(300 MHz, CDCl3) d 5.19 (dd, J = 3.3, 1.2 Hz, 1H),
4.15–4.05 (m, 3H), 3.88–3.73 (m, 5H), 3.49–3.43 (m,
2H), 2.27–1.92 (m, 6H), 1.73–1.63 (m, 2H), 1.32–1.24
(m, 7H), 1.35 (t, J = 7.0 Hz, 3H), 1.04 (t, J = 7.0 Hz,
3H), 0.88 (t, J = 6.8 Hz, 3H), 13C NMR (75 MHz,
CDCl3) d 104.1, 82.2, 82.2, 80.9, 80.6, 67.4, 66.3, 62.9,
32.9, 31.8, 29.7, 29.4, 29.3, 28.7, 28.1, 27.1, 26.3, 22.7,
15.7, 14.1; IR (neat) 3456, 2927, 2857, 1459, 1376,
1347, 1197, 1097, 1041, 936, 852 cmꢀ1; HRMS (EI) calcd
for C12H21O4 (Mꢀ129) 229.1439, found 229.1440.
added
N-methylquinolinium
hexafluorophosphate
(1.6 mg, 0.0057 mmol), sodium acetate (100 mg,
1.21 mmol), anhydrous Na2S2O3 (100 mg, 0.63 mmol),
and toluene (1 mL). The mixture was photoirradiated
with gentle air bubbling for 2.5 h while stirring at room
temperature. The reaction mixture was filtered through
a small plug of silica gel and the filtrate concentrated.
The resulting residue was purified by flash chromatogra-
phy (10% EtOAc in hexanes) to provide the desired
compound as a 10:1 mixture of two diastereomers
25
(15 mg, 46%): ½aŠ ¼ ꢀ129:3 (c 0.18, EtOAc); 1H
D
4.6. Lactone 28
NMR (300 MHz, CDCl3) d 7.62 (d, J = 8.1 Hz, 2H),
7.48 (d, J = 8.1 Hz, 2H), 4.87 (d, J = 12.1 Hz, 1H),
4.81 (d, J = 12.1 Hz, 1H), 4.74 (d, J = 3.3 Hz, 0.9H),
4.42 (dd, J = 9.4, 2.0 Hz, 0.1H), 4.10–4.02 (m, 1H),
3.76–3.67 (m, 1H), 3.50 (s, 0.3H), 3.32 (s, 2.7H), 3.01
(t, J = 9.1 Hz, 1H), 2.25 (ddd, J = 13.1, 4.4, 0.9 Hz,
To 27 (11 mg, 0.0275 mmol) in CH2Cl2 (2 mL) was
added Et3N, DMAP, and Ac2O. The mixture was stirred
for 3 h and then was partitioned between CH2Cl2
and H2O. The organic layer was dried over MgSO4,

3578
V. S. Kumar et al. / Tetrahedron: Asymmetry 16 (2005) 3570–3578
11. For other additions of ethers into halonium ions, see: (a)
Khan, N.; Xiao, H.; Zhang, B.; Cheng, X.; Mootoo, D. R.
Tetrahedron 1999, 55, 8303; (b) Collum, D. B.; McDonald,
J. H., III; Still, W. C. J. Am. Chem. Soc. 1980, 102, 2120.
12. (a) Tang, S. H.; Kennedy, R. M. Tetrahedron Lett. 1992,
33, 5299; (b) McDonald, F. E.; Towne, T. B. J. Am. Chem.
Soc. 1994, 116, 7921; (c) Donohoe, T. J.; Butterworth, S.
Angew. Chem., Int. Ed. 2003, 42, 948; (d) Piccialli, V.;
Caserta, T. Tetrahedron Lett. 2004, 45, 303; (e) Donohoe,
T. J.; Butterworth, S. Angew. Chem., Int. Ed. 2005, 44,
4766.
0.9H), 2.04 (br s, 1H), 1.71 (ddd, J = 15.3, 11.6, 3.7 Hz,
0.9H), 1.38 (d, J = 6.2 Hz, 0.3H), 1.33 (d, J = 6.2 Hz,
2.7H); 13C NMR (75 MHz, CDCl3) d 142.5, 127.7,
127.6, 125.5 (q), 100.5, 98.3, 86.7, 86.3, 74.3, 74.2,
71.1, 69.1, 66.7, 54.7, 37.9, 29.8, 18.4; IR (neat) 3200,
2934, 1616, 1327, 1211, 1168, 1109, 1067, 1051, 1018,
969, 821 cmꢀ1; HRMS (EI) calcd for C12H19O4F3
(M+) 320.1235, found 320.1240.
13. (a) Kumar, V. S.; Floreancig, P. E. J. Am. Chem. Soc.
2001, 123, 3842; (b) Aubele, D. L.; Floreancig, P. E. Org.
Lett. 2002, 4, 3443.
14. For a preliminary account of this work, see: Kumar, V. S.;
Aubele, D. L.; Floreancig, P. E. Org. Lett. 2002, 4, 3443.
15. Kumar, V. S.; Aubele, D. L.; Floreancig, P. E. Org. Lett.
2001, 3, 4123.
Acknowledgments
This work was supported by generous funding from the
National Science Foundation and the National Insti-
tutes of Health.
16. For alkoxy group delivery through mixed acetals in
carbohydrate synthesis, see: Barresi, F.; Hindsgaul, O. J.
Am. Chem. Soc. 1991, 113, 9376.
References
1. Tietze, L. F. Chem. Rev. 1996, 96, 115.
2. For recent examples, see: (a) Kerber, W. D.; Gagne, M. R.
17. (a) Wang, Z.-X.; Tu, Y.; Frohn, M.; Zhang, J.-R.; Shi, Y.
J. Am. Chem. Soc. 1997, 119, 11224; (b) Tu, Y.; Wang,
Z.-X.; Shi, Y. J. Am. Chem. Soc. 1996, 118, 9806.
18. For a related approach to the enantioselective synthesis of
bis-epoxides, see Ref. 7.
19. An exact determination of the relative diastereomeric ratio
of the epoxides was not possible due to the presence of
other randomized stereocenters in the molecule.
20. (a) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.;
Masamune, H.; Sharpless, K. B. J. Am. Chem. Soc. 1987,
109, 5765; (b) Katsuki, T.; Sharpless, K. B. J. Am. Chem.
Soc. 1980, 102, 5974.
21. (a) Jadhav, P. K.; Bhat, K. S.; Perumal, P. T.; Brown, H.
C. J. Org. Chem. 1986, 51, 432; (b) Brown, H. C.; Jadhav,
P. K. J. Am. Chem. Soc. 1983, 105, 2092; (c) Koide, K.;
Finkelstein, J. M.; Ball, Z.; Verdine, G. L. J. Am. Chem.
Soc. 2001, 123, 398.
22. Martin, V. S.; Woodard, S. S.; Katsuki, T.; Yamada, Y.;
Ikeda, M.; Sharpless, K. B. J. Am. Chem. Soc. 1981, 103,
6237.
´
´
Org. Lett. 2005, 7, 3379; (b) Nieto-Oberhuber, C.; Lopez,
S.; Echavarren, A. M. J. Am. Chem. Soc. 2005, 127, 6178.
3. Xiong, Z.; Corey, E. J. J. Am. Chem. Soc. 2000, 122,
4831.
4. Still, W. C.; Romero, A. G. J. Am. Chem. Soc. 1986, 108,
2105.
5. Baldwin, J. E. J. Chem. Soc., Chem. Commun. 1976,
738.
6. (a) Na, J.; Houk, K. N.; Shevlin, C. G.; Janda, K. D.;
Lerner, R. A. J. Am. Chem. Soc. 1993, 115, 8453; (b)
Nicolaou, K. C.; Prasad, C. V. C.; Somers, P. K.; Hwang,
C.-K. J. Am. Chem. Soc. 1989, 111, 5330; (c) Heffron,
T. P.; Jamison, T. F. Org. Lett. 2003, 5, 2339.
7. (a) Valentine, J. C.; McDonald, F. E.; Neiwart, W. A.;
Hardcastle, K. I. J. Am. Chem. Soc. 2005, 127, 4586; (b)
McDonald, F. E.; Bravo, F.; Wang, X.; Wei, X.; Toganoh,
´
M.; Rodrıguez, R.; Do, B.; Neiwart, W. A.; Hardcastle, K.
I. J. Org. Chem. 2002, 67, 2515; (c) McDonald, F. E.;
Wang, X.; Do, B.; Hardcastle, K. I. Org. Lett. 2000, 2,
2917.
23. Liotta, L. J.; Dombi, K. L.; Kelley, S. A.; Targontsidis, S.;
Morin, A. M. Tetrahedron Lett. 1997, 38, 7833.
24. (a) Stanek, J.; Marek, M.; Jary, J. Carbohydr. Res. 1978,
64, 315; (b) Grethe, G.; Mitt, T.; Williams, T. H.;
Uskokovic, M. J. Org. Chem. 1983, 48, 5309.
8. Zakarian, A.; Batch, A.; Holton, R. A. J. Am. Chem. Soc.
2003, 125, 7822.
´
´
9. Alvarez, E.; Dıaz, M. T.; Perez, R.; Ravelo, J. L.; Vera, J.
´
A.; Zurita, D.; Martın, J. D. J. Org. Chem. 1994, 59, 2848.
25. The chemical shift of the hydrogen at the anomeric center
is highly characteristic of tetrahydrofuranyl ethers
(P5.0 ppm) and tetrahydropyranyl ethers (ꢁ4.6 ppm).
10. Bravo, F.; McDonald, F. E.; Neiwart, W. A.; Hardcastle,
K. I. Org. Lett. 2004, 6, 4487.