The design and synthesis of novel anomeric hydroperoxides: Influence of the carbohydrate residue in the enantioselective epoxidation of quinones

Article abstract of DOI:10.1016/j.tetasy.2004.11.015

We present a study of the base (DBU)-catalysed epoxidation of a number of important naturally occurring quinones using a series of pyranose-derived anomeric hydroperoxides. The absolute (viz. d or l) stereochemistry of the carbohydrate, electronic nature of the 6-substituent and ring substitution are all important variables, both for the formation of the hydroperoxide and its reactivity. Reactions studied were the epoxidation of a precursor of the natural antibiotic, alisamycin and a series of naphthoquinones related to Vitamin K. In the best case, an ee of 82% was obtained; either product enantiomer is accessible according to the absolute stereochemistry of the carbohydrate. Finally, a molecular modelling study of the reaction is reported, concluding that the reactions are under kinetic control and that the observed ees cannot be explained by considering transition states that involve only the quinone and peroxide anion. It seems likely that the DBU molecule may play a key role in the transition state.

Full text of DOI:10.1016/j.tetasy.2004.11.015

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 16 (2005) 283–293
The design and synthesis of novel anomeric hydroperoxides:
influence of the carbohydrate residue in the
enantioselective epoxidation of quinones
Abass Bundu,a Neil G. Berry,b Christopher D. Gill,a Catherine L. Dwyer,a,ꢀ
a,
c
Andrew V. Stachulski,b,* Richard J. K. Taylor and John Whittall
*
aDepartment of Chemistry, University of York, Heslington, York YO10 5DD, UK
bRobert Robinson Laboratories, Department of Chemistry, University of Liverpool, Liverpool L69 7ZD, UK
cStylacats Ltd, The Heath Business and Technical Park, Runcorn, Cheshire, WA7 4QX, UK
Received 1 November 2004; accepted 8 November 2004
Available online 8 January 2005
Abstract—We present a study of the base (DBU)-catalysed epoxidation of a number of important naturally occurring quinones
using a series of pyranose-derived anomeric hydroperoxides. The absolute (viz. D or L) stereochemistry of the carbohydrate, elec-
tronic nature of the 6-substituent and ring substitution are all important variables, both for the formation of the hydroperoxide
and its reactivity. Reactions studied were the epoxidation of a precursor of the natural antibiotic, alisamycin and a series of naph-
thoquinones related to Vitamin K. In the best case, an ee of 82% was obtained; either product enantiomer is accessible according to
the absolute stereochemistry of the carbohydrate. Finally, a molecular modelling study of the reaction is reported, concluding that
the reactions are under kinetic control and that the observed ees cannot be explained by considering transition states that involve
only the quinone and peroxide anion. It seems likely that the DBU molecule may play a key role in the transition state.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Background and introduction
and had published a preliminary report of the use of chi-
ral pool epoxidants, namely carbohydrate anomeric
hydroperoxides (AHPs), for the same purpose.9 Chmie-
lewski et al. also prepared AHPs10,11 and used them in
the oxidation of allyl alcohols and alkyl aryl sulfides,
obtaining reasonable ees. In addition, Adam et al.
employed chiral secondary benzylic hydroperoxides in
the Ti-catalysed epoxidation of allylic alcohols and in
the asymmetric epoxidation of chalcones.12
The synthesis of enantiomerically pure epoxides is a fun-
damental challenge in organic chemistry.1 Epoxides are
both versatile building blocks in synthesis and signifi-
cant in their own right as integral functional groups of
important natural products. While impressive advances
have been made in the asymmetric epoxidation of elec-
tron-rich alkenes2 and electron-deficient trans-alkenes
of the chalcone type,3–5 there are few effective proce-
dures for the asymmetric epoxidation of cyclic enones
and quinones. Important examples of this type include
the manumycin family of antibiotic/anticancer agents
and naphthoquinone epoxides of the Vitamin K series.
OR
OPiv
O
OOH
O
OOH
RO
PivO
Previously, we had utilised chiral phase-transfer cata-
lysts6,7 and chiral bases8 to prepare quinone epoxides,
3
1 R = Piv, 2 R = Bn
The above mentioned earlier publication9 reported that
the D-pyranose-derived hydroperoxides 1–3 were the
most effective epoxidants of a number of D-glucose
and D-galactose derivatives screened: the pivaloates
(rather than acetates) were essential to give adequate
*
Corresponding authors. Fax: +44 0151 794 3588 (A.V.S.); fax: +44
1904 434523 (R.J.K.T.); e-mail addresses: stachuls@liverpool.ac.uk;
rjktl@york.ac.uk
ꢀ Present address: Sasol Technology Research & Development, PO Box
1, Sasolburg 1947, South Africa.
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.11.015

284
A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
stability to the epoxidants. The test reactions used were
epoxidations of the alisamycin7,13,14 core amidoquinone
(4!6) and, using 1 only, a series of naphthoquinones
(5!7),15 Schemes 1 and 2.
We set out to obtain a more complete picture by screen-
ing AHPS of other pyranose carbohydrates, varying in
particular the 6-substituent, to determine the impor-
tance of electronic and stereochemical features. In par-
ticular we addressed one clear omission of the earlier
examples, namely the lack of an L-series sugar (since
the effect of anomeric configuration in itself appeared
too small to overturn the ee to a useful extent).9 In
this way we hoped to access either enantiomer of a pair
of epoxides with equal facility. The most accessible L-
series sugar is L-rhamnose, and reactions with rham-
nose-derived AHPs are described herein. Finally, we
present our conclusions from a molecular modelling
study of the epoxidation reaction.
O
O
NHCOR
NHCOR
1-3, DBU
O
PhMe, 20oC
O
O
(-)-6
4
R =
Scheme 1. Epoxidation of (À)-alisamycin precursor. Yields 38–64%, ee
42–64%.9
2. Discussion
O
O
2.1. Preparation of carbohydrate intermediates
R
O
R
1, DBU
PhMe, 20oC
The D-glucose derived AHPs 1 and 2 were readily ob-
tained from the known16 ꢁpseudoglycalꢂ or Ferrier rear-
rangement product 8 (Scheme 3). The switch from the
known acetates to the novel pivaloates, via compounds
9 and 10, was essential to improve the stability of the
products under the basic conditions of the epoxidation
reaction. Acid-catalysed anomeric displacement on 10
with hydrogen peroxide then afforded 1 as a separable
a/b AHP mixture; alternatively, di-O-benzylation of diol
9, followed by treatment with H2O2 and acid, afforded 2
as previously reported.11,17
O
O
i
7a-f
5 a R = Me
b R = Et
c R = Pr
d R = Pr
e R = Ph
f R = c-C6H11
Scheme 2. Epoxidation of naphthoquinones. Yields 30–80%, ees 35–
82%.9
General conclusions from this phase of the work were:
1. DBU was the most effective base: other organic and
inorganic bases gave lower yields and/or ee.
2. Toluene was the most effective solvent in terms of ee.
3. Anomeric purity of the hydroperoxide was vital: the
corresponding b-hydroperoxides gave the opposite
enantiomeric product with lower selectivity.
The di-pivaloate AHP 1 was isolated as a crystalline
material and proved suitable for X-ray structure deter-
mination (Fig. 1). This shows a half-chair conformation,
in agreement with the NMR evidence [especially the
H(4)–H(5) coupling of J = 9.5Hz].
OH
OPiv
OAc
O
OEt
OR
O
OEt
(ii)
(i)
HO
PivO
AcO
9
10 R = Et (α only)
8
(iv), (iii)
1
R = OH
(vi)
(v)
OBn
O
OR
O
OEt
O
OOH
O
O
Si
O
O
O
BnO
R = Et (α only)
2
11
12
13
(iii)
R = OH
Scheme 3. Mono- and bicyclic anomeric hydroperoxides, D-glucose series. Reagents: (i) Et3N, MeOH, cat. K2CO3, quant.; (ii) Me3CÆCOCl, pyridine;
(iii) H2O2, dioxane, H2SO4, 50%, a:b, 2.2:1 for 1; 64%, a:b, 2.7:1 for 2; 87%, a:b, 5:1 for 13; (iv) NaH, BnBr, DMF, 76%; (v) Cl3CÆCOÆCCl3, 59%; (vi)
But2SiðOTfÞ , pyridine, 68%.
2

A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
285
curonic acid-derived glycal 15 was prepared by a known
method.21 However, we could not transform 15 to the
anomeric hydroperoxide 16, either directly or via a
pseudoglycal. It appears that the destabilising effect of
the 6-CO2Me substituent on an incipient anomeric car-
bonium ion is decisive here. This transformation has
been performed previously22 using 85% grade H2O2,
but this is no longer commercially available; we ob-
served no reaction on prolonged reaction with 60%
H2O2.
O
O
O
O
OOH
MeO
AcO
MeO
AcO
OAc
15
16
In complete contrast, the synthesis of the L-rhamnose
analogue proceeded very smoothly, which we attribute
to the presence of the electron-donating 6-Me group.
An effective ꢁone potꢂ preparation23 of L-rhamnal 17
from rhamnose (Scheme 5) was further improved24
by using a non-aqueous version of the Zn–Cu couple-
catalysed reductive elimination from the intermediate
bromosugar to generate 17. Acid-catalysed conversion
of 17 to the desired AHP 18 did not even require the
intermediacy of a pseudoglycal intermediate. In this
series (cf. 1) the acetate 18 proved quite robust; alterna-
tively, using a four-step, but straightforward sequence,
17 could be transformed into the pivaloate 20 via the
ethyl pseudoglycal 19 (Scheme 5). Both 18 and 20 were
obtained very largely (ca. 9:1) as the a-anomers, which
were readily separated by chromatography: for good
yields of 18 and 20, it was essential to use an Et2O–
H2O2–H2SO4 mixture.
Figure 1. X-ray crystal structure of 1. Hydrogen atoms have been
omitted for clarity: ellipsoids are drawn at 50% probability.
In order to probe further the structure–reactivity rela-
tionships in this series, a novel bicyclic hydroperoxide
was also prepared (Scheme 3). Thus Zemplen deprotec-
tion of the Ferrier intermediate 8 gave the presumed 4,6-
diol 9, which without isolation, was transformed by
acylation with triphosgene (59%)18 into the cyclic car-
bonate 11 or by base-catalysed silylation (68%) with
di-t-butylsilyl ditriflate19 into bicyclic ether 12. Acid-
catalysed reaction of 11 with H2O2 gave no reaction
even after several days, whereas 12 reacted smoothly
in 16h to give hydroperoxide 13 in 75% yield (insepar-
able anomeric mixture, a:b = 5:1). The profound differ-
ence in reactivity between 11 and 12 is attributed to
the destabilising effect of an electron-withdrawing
C(6)-substituent on an incipient anomeric carbonium
ion and is parallelled by the behaviour of monocyclic
derivatives (vide infra).
2.2. Enantioselective epoxidations
We studied a range of quinonoid substrates 4, 5a and 5e
to evaluate the best epoxidant, comparing mainly 1, 13,
18 and 20; in the case of 4 the results for AHPs 2 and 3
have been added.9 The quinones were prepared by liter-
ature syntheses.12,25,26 Toluene was shown earlier to be
the optimum solvent and was used exclusively here;9
the hydroperoxides were used in 1:1 molar ratio with 4
and in 20% molar excess with 5a and 5e.9 Considering
first the alisamycin precursor 4 (Table 1), which gave
complete reaction in a few minutes, both D-glucose
derived 1 and L-rhamnose-derived 18 gave good results,
The AHP 3 in the D-galactose series was obtained as an
inseparable a:b mixture from diol intermediate 9 using a
known20 method of Mitsunobu inversion at C(4) fol-
lowed by protecting group interchange, then treatment
of the intermediate ethyl glycoside 14 with acidic H2O2
(Scheme 4).
We also sought to study an AHP bearing a strongly elec-
tron-withdrawing 6-substituent, and to this end the glu-
OPiv
OPiv
OH
OOH
O
OEt
(iv)
O
OEt
(i)-(iii)
PivO
PivO
HO
3
9
14
Scheme 4. Synthesis of AHP in the D-galactose series. Reagents: (i) Ph3P, DEAD, PhCO2H, THF, 20ꢁC, 59%; (ii) NaOMe, MeOH; (iii) Piv-Cl,
pyridine, 28% over two steps; (iv) 60% aq H2O2, H2SO4, dioxane, 43.5%, a:b, 10:1.

286
A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
Me
O
Me
HO
O
OH
OH
Me
O
O-OH
(iv)
(i)-(iii)
AcO
AcO
OAc
OH
L-Rhamnose
17
18
(v)-(vii)
Me
O
OEt
Me
O
O-OH
(iv)
PivO
PivO
19
20
Scheme 5. Anomeric hydroperoxide synthesis, L-rhamnose series. Reagents: (i) Ac2O, DMAP, CH2Cl2, Et3N; (ii) HBr–AcOH, CH2Cl2; (iii) Zn–Cu
couple, AcOH, THF, 83% over three steps; (iv) Et2O–H2O2, H2SO4, 82%, a:b, 10:1 for 18, 86%, a:b, 10:1 for 20; (v) SnCl4, EtOH, CH2Cl2, 88%; (vi)
cat. K2CO3, Et3N, MeOH; (vii) Me3CÆCOCl, pyridine, 96% over two steps.
Table 1. DBU-catalysed epoxidations of quinones with AHPs in toluene at 20ꢁC
Quinone
Alisamycin precursor 4
2-Methyl naphthoquinone 5a
2-Phenyl naphthoquinone 5e
Hydroperoxide
Yield (%)
Ee (%)
T/min
Yield (%)
Ee (%)
+45b
T/h
Yield (%)
Ee (%)
+82b
T/h
1 (D-)
2 (D-)
3 (D-)
18 (L-)
20 (L-)
13 (D-)
55
38
64
74
76
59
À64a
À46
À42
+55
5
5
5
5
6
5
71
5.0
80
6
32
74
57
À78
À49
+40
3.5
4.0
4.5
47
À37.5
25
+22
À30.5
Absolute configuration of each AHP in brackets.
a Absolute configuration of major enantiomer as in 6.
b Absolute configuration of major enantiomer as in 7.
affording access to either enantiomer predominantly,
with the best ee being obtained using 1 [64% ee of the
(À)-enantiomer] and the significantly better yield from
18 being offset by a slightly lower ee [55% of (+)-enan-
tiomer]. The substitution of benzyl ether protection for
esters in 2 and the inversion of C(4) in 3 offered no
advantage over 1, though reasonable ees were observed.
Compared to 18, the pivaloate 20 gave an equivalent
yield (within experimental error) but a disappointing
ee. As expected, the bicyclic D-glucose derived reagent
13 gave the same predominant enantiomer as 1 but with
a much lower ee; it must be remembered that 13 was
used as an inseparable a:b = 5:1 mixture, while other
AHPs were single a-anomers.
gave an outstanding result with 80% yield and 82% ee:
the reaction of 18 was much slower, for reasons that
are unclear, and the yield and ee were moderate. For this
substrate, both 14 and 16 reacted too slowly to be of
value.
We also studied briefly the epoxidation of trans-chal-
cone, the standard substrate for the H2O2–polyleucine
method.3 Out of AHPs 1, 13, 18 and 20 only 1 would re-
act; in a very slow reaction (4d) a yield of 10% and an ee
of 13% were obtained. Other methods5a–c,12 are clearly
superior in acyclic systems.
2.3. Molecular modelling studies
In the case of the 2-methyl naphthoquinone 5a, where
reactions were between 50- and 300-fold slower than
with 4, AHP 18 gave the highest ee yet recorded for
epoxidation in this series, 78% compared to 45% (with
opposite sign) from 1. The yield, however, was disap-
pointing at 32%. Here the pivaloate 20 gave essentially
the same yield as 1 and an almost equal but opposite
ee. The bicyclic AHP 13 was slightly inferior to 1 in both
yield and ee. For the 2-phenylnaphthoquinone 5e, 1
Computational methods were employed to attempt to
explain the observed product enantioselectivity for the
reaction of 1 with 2-methyl naphthoquinone. It was
hoped that by comparing the calculated energy differ-
ences between the transition states when attack of the
peroxide anion occurs from either side of the plane of
the quinone system, the observed product distribution
may be explained.

A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
287
The reactions were modelled in silico assuming that they
proceeded via a Weitz–Scheffer mechanism with attack
of the peroxide anion occurring at the least hindered
end of the double bond. The rationale for this assump-
tion is outlined in a paper by Pluim and Wynberg.27
Additional evidence for this hypothesis was shown by
MOPAC28 calculations of the LUMO of the quinone
and the HOMO of the peroxide anion of 1 (see Section
4 for details). The LUMO of 2-methyl naphthoquinone
is largely located on the quinone portion of the molecule
and the HOMO of the attacking peroxide anion is al-
most entirely located on the peroxide oxygens (data
not shown). Thus, it seemed reasonable that the reaction
was likely to proceed via an interaction of these two
molecular orbitals. The reactions were considered to
be under kinetic control, and the first step (attack of
the peroxide anion to the b position of the double bond)
was initially taken to be the rate determining step.
Kinetic control was confirmed, as the calculated heats
of formation of enantiomeric products of 1 with 2-
methyl naphthoquinone were very similar (data not
shown). In order to investigate whether the enantioselec-
tivity resulted from kinetic control, it was necessary to
locate the transition states for each pair of reactants
attacking from either face of the quinone.
veal why the enantiomeric excesses were observed. We
can compare the heats of formation of the TS rather
than the heats of activation as the heats of formation
of the starting materials are the same, independent of
which side of the quinone the reaction occurs from. Dis-
appointingly, the difference in heat of formation of the
two transition states was very small and does not explain
the enantioselectivity observed (Table 2).
Table 2. Heats of formation of the transition states of 1 reacting from
the re- and si-faces of the 2-methyl naphthoquinone
re-Face
si-Face
Heat of formation (kcal/mol)
À280.785127
À280.785040
Up to this point the first step of the reaction was taken
as the rate determining step. The second step, ring clo-
sure and elimination of the alkoxide anion, was hypo-
thesised to be not rate determining.32 In order to
investigate whether this second step was in fact rate
determining, an analogous study was undertaken to
locate the TS for this second step of the reaction.
The TS’s were located by working from a published TS
for a similar reaction.32 The approximate TS was input
into SYBYL 6.91 and various constraints30 were applied
to the system (see Section 4) to generate an approximate
TS. This rough TS was refined in the same way as de-
scribed above and vibrational analysis revealed one nega-
tive force constant, which led to the product epoxide
and alkoxide anion, confirming that it was a true TS.
A similar analysis was carried out for the case when
the peroxide attacked the other face of the quinone
and the TS was again confirmed by the single negative
vibration, which led to the expected products. As before,
there was very little difference between the two heats of
formation, thus the enantiomeric excess cannot be ex-
plained by this second step of the reaction (Table 3).
Unfortunately, a literature search did not reveal a pro-
posed transition state (TS) for the epoxidation of qui-
nones with a peroxide anion. As a reasonable structure
could not be taken from the literature, an alternative
method was employed. The molecules were entered into
the SYBYL 6.9129 molecular modelling program, a con-
straint30 was placed between the b-carbon and the per-
oxide anion (see Section 4 for details) and the system
was subjected to molecular mechanics energy minimisa-
tion.30 This partially refined structure was then inserted
into MOPAC28 and was minimised using semi-empirical
quantum mechanical methods, with the distance be-
tween the b-carbon and the peroxide anion kept fixed.
The output of this calculation was taken as a crude esti-
mate of the TS. This approximate TS was refined so that
the total energy gradient was very close to zero, a condi-
tion of a true TS.31 Once refined, the TS was confirmed
to be a saddle point as vibrational analysis revealed that
the TS had one single negative vibration along the reac-
tion coordinate31 and the presence of six very small
eigenvalues (less than 5cmÀ1). Two Intrinsic Reaction
Coordinate (IRC) calculations confirmed that the TS
did connect the reactants to the products (data not
shown).
Table 3. Heats of formation of the transition states of ring closure of 1
reacting from the re- and si-face of the 2-methyl naphthoquinone
Re-Face
Si-Face
Heat of formation (kcal/mol)
À278.257059
À278.256986
In summary, it does not appear that the enantiomeric
excess observed for this system can be explained by a
transition state that involves two species, that is, the
peroxide and quinone. It may be that the conjugate acid
of DBU is involved in the transition state by forming an
ion pair with the hydroperoxide anion. This hypothesis
agrees with the findings that ꢁnon-polar solvents gave
the highest enantiomeric excessesꢂ and ꢁchanging the
base resulted in a reduction in the yield and/or enantio-
meric excessesꢂ9 and also the successful use of enantio-
pure guanidine bases for epoxidation.8 Previous work
on a,b-enones with optically active hydroperoxides has
proposed that enantioselectivity has been induced
through hydrogen bonded transition states that involve
DBU.33 Although the systems under investigation
here are similar, they do differ in one important
Having successfully located the TS for peroxide 11 add-
ing to quinone 5a when attack occurred from the re-face
of the quinone, attempts were made to locate the TS for
the reaction occurring from the si-face of the quinone.
This was achieved in a similar manner as before and
was confirmed to be a true TS by vibrational analysis
and IRC calculations.
If the reaction proceeded via the proposed mechanism
and under kinetic control, then the differences in heat
of formation of the TS of the attack from the one face
of the quinone and attack from the other face should re-

288
A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
respect—the quinone system places the enone function-
ality in a transoid configuration rather than cisoid. Thus
it may be necessary to account for three species (qui-
none, peroxide and DBU) in the transition state to ex-
plain the enantioselectivity. Many attempts were made
to locate this TS: however, they all proved unsuccessful,
due to the very large problem space that exists for this
type of molecular arrangement.
ꢁflashꢂ silica gel at enhanced pressure. NMR spectra were
obtained on JEOL EX270 or 400 spectrometers using
the indicated solvents, operating at 270 or 400MHz
1
for H or at 67.5MHz for 13C spectra. Chemical shifts
(d) are relative to Me4Si and coupling constants J are
quoted in hertz. IR spectra were recorded on a Thermo-
Nicolet FT-IR 100 spectrometer for the physical form
noted. Optical rotations were measured on a JASCO
DIP-370 digital polarimeter at 20ꢁC (Na D-line). Mass
spectra [by electron impact (EI), chemical ionisation
(CI), electrospray (ES) or fast-atom bombardment
(FAB)] were obtained on a Fisons analytical (VG) auto-
spec instrument. HPLC data were recorded on a Gilson
Autoinjector 234, Gilson 170 Diode Array Detector and
Gilson 321 pump. For chiral HPLC, DAICEL CHIR-
ALPAK AD or OD columns were used for the 2-substi-
tuted naphthoquinone epoxides or alisamycin
precursors, respectively.
3. Conclusions
We have described a series of AHPs suitable for the
epoxidation of some important quinones with fair to
good yields and ees. With regard to the ease of forma-
tion of the AHPs, the electronic character of the carbo-
hydrate 6-substituent is a determining factor. In the
rhamnose series (6-Me), acid-catalysed hydroperoxida-
tion at the anomeric carbon is facile, even on the glycal,
while in the methyl glucuronate series it appears to be
impossible on either the glycal or ethyl pseudoglycal
with the strength of commercial H2O2 now available.
The ꢁnormalꢂ pyranoses, glucose and galactose, show
intermediate reactivity.
4.1. 2,3-Dideoxy-1-O-oxidanyl-4,6-di-O-pivaloyl-a-D-
erythro-hex-2-enopyranose 1a and 2,3-dideoxy-1-O-oxi-
danyl-4,6-di-O-pivaloyl-b-D-erythro-hex-2-enopyranose
1b
Ethyl 2,3-dideoxy-4,6-di-O-pivaloyl-a-D-erythro-hex-2-
enopyranoside 1034 (0.95g, 2.8mmol) in 1,4-dioxane
(5mL) was treated with 67% aq H2O2 (1mL) and a cata-
lytic amount of concd H2SO4, then the reaction mixture
was stirred for 4d. Partition between CH2Cl2 and water
(10mL each), followed by extraction of the aqueous
layer with further CH2Cl2 (2 · 10mL), then combina-
tion of the organic extracts, drying (MgSO4) and evap-
oration afforded the title compound 1 and its anomer
1b (462mg, 50.4%, a:b = 2.2:1). This mixture was par-
tially separable by column chromatography (PE/EtOAc
9:1 as eluent), yielding a colourless oil, a mixture of ano-
mers (a:b = 1:1, 253mg) and a white solid (a:b = 19:1,
209mg), which was recrystallised from EtOAc to give
pure a-hydroperoxide 1 (114mg); mp 98–100ꢁC; [found
C 58.9%, H 8.3%, C16H26O7 requires C 58.2%, H 7.9%];
Rf (9:1 PE/EtOAc) 0.15; [a]D +133.9 (c 1, CHCl3); mmax
(CHCl3)/cmÀ1 3429 (OOH), 2974 (CH), 1732 (C@O),
1481, 1283, 1152, 996; dH (270MHz CDCl3) 5.98–5.93
(1H, m, H-3), 5.71 (1H, ddd, J1,2 3.0, J2,3 10.0, J2,4
2.0Hz, H-2), 5.48 (1H, m, H-1), 5.35–5.29 (1H, ddd,
J2,4 2.0, J3,4 3.5, J4,5 9.5Hz, H-4), 4.34–4.08 (3H, m,
H-5, H-6, H-60), 1.17 (9H, s, OPiv), 1.15 (9H, s, OPiv);
dC (67.5MHz, CDCl3) 178.6, 177.7 (C, OPiv), 133.1
(CH, C-3), 123.0 (CH, C-2), 98.5 (CH, C-1), 67.6, 64.4
(CH, C-4, C-5), 62.3 (CH2, C-6), 38.8 (C, OPiv), 26.9,
26.6 (CH3, OPiv); CIMS m/z 348 (MNHþ4 , 18), 330
(M+, 77), 297 (M+ÀHO2, 100), HRMS (CI): found
MNHþ4 , found 348.2024. C16H30NO7 requires
348.2022. 0.5ppm error. Combination of the mixed frac-
tions from a number of reactions followed by flash col-
umn chromatography (PE/EtOAc 9:1 as eluent) gave the
b-anomer 1b as a colourless oil; Rf (PE/EtOAc, 9:1)
0.19; [a]D = +102.5 (c 1, CHCl3); dH (270MHz, CDCl3)
9.60 (1H, br s, OOH), 6.15–6.07 (1H, m, H-3), 5.95 (1H,
ddd, J1,2 2.5, J2,3 10.5, J2,4 1.0Hz, H-2), 5.59 (1H, m, H-
1), 5.11 (1H, m, H-4), 4.42–4.10 (3H, m, H-5, H-6, H-60),
1.23–1.20 (18H, m, 2· OPiv); dC (67.5MHz, CDCl3)
178.5, 177.7 (C, OPiv), 128.7 (CH, C-3), 125.7 (CH, C-
The situation with regard to the epoxidation results is
harder to assess. Certainly the configuration of the
major enantiomer obtained is determined by the abso-
lute configuration of the AHP, in practical synthesis a
very useful feature, but rationalisation of the results
beyond this is difficult. While the pivaloate groups in 1
are essential to confer stability, the pivaloate 19 in the
rhamnose series gave poorer ees than acetate 17, and
reacted too slowly to be useful with quinone 5e; in yield
terms, however, it offered the best results for 4 and 5a.
The bicyclic AHP 13 gave reasonable results but offered
no advantage over 1.
Molecular modelling failed to explain the observed ee
for the reaction of 1 with 2-methyl naphthoquinone
assuming that it proceeded via a transition state, which
involved two species (peroxide and quinone). However,
it was shown with fair certainty that the epoxidation is
kinetically controlled and that the transition state prob-
ably intimately involves the DBU.
Further examples of the application of these AHP re-
agents in natural product synthesis will be presented in
due course.
4. Experimental
Reactions were carried out under a nitrogen atmo-
sphere. THF and diethyl ether (Et2O) were distilled from
benzophenone ketyl prior to use; CH2Cl2 was dried over
CaH2 and distilled immediately prior to use. Other sol-
vents, of analytical grade, were used as purchased. Melt-
ing points were recorded on an Electrothermal 9100 and
are uncorrected. Analytical TLC was performed using
aluminium plates coated with Merck silica gel 60 F254
Preparative chromatography was carried out using Fish-
.
˚
er matrix silica gel 60 @ 760mmHg or ICN 33–63 (60A)

A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
289
2), 97.9 (CH, C-1), 73.7, 63.0 (CH, C-4, C-5), 63.9 (CH2,
C-6), 38.9 (C, OPiv), 27.0 (CH3, OPiv).
(20mL), dried (MgSO4), filtered and concentrated in
vacuo to give a crude pale yellow oil. The product
was purified by flash column chromatography on silica
gel (PE/EtOAc 3:1 as eluent), yielding the title com-
pound 12 (626mg, 68%) as a colourless oil; Rf (PE/
EtOAc 1:1) 0.80; [a]D = +47.2 (c 1.15, CHCl3); mmax
neat/cmÀ1 2934, 2861 (2 · O–SiR), 1473, 1392, 1306,
1057; dH (400MHz, CDCl3) 6.01 (1H, d, J2,3 10.0Hz,
H-3), 5.63 (1H, dt, J1,2 2.4, J2,4 2.4, J2,3 10.0Hz, H-2),
4.93 (1H, br d, J1,3 1.2Hz, H-1), 4.34 (1H, ddd, J3,4
1.5, J2,4 2.1, J4,5 9.0Hz, H-4), 4.12 (1H, dd, J5,6 5.0,
4.2. 4,6-Di-O-benzyl-2,3-dideoxy-1-O-oxidanyl-a-D-ery-
thro-hex-2-enopyranose 2
Ethyl
4,6-di-O-benzyl-2,3-dideoxy-a-D-erythro-hex-2-
enopyranoside17 (400mg, 1.1mmol) was treated with
67% aq H2O2 (0.5mL) in dioxane following the above
procedure for the synthesis of hydroperoxide 1. The
product 2 was identical to that previously reported.11
0
0
0
J6,6 9.8Hz, H-6), 3.86 (1H, dd, J5,6 10.0, J6,6 9.8Hz,
H-60), 3.85–3.77 (2H, m, OCHACHBCH3, H-5), 3.57–
3.49 (1H, dq, Jgem 9.8, J 7.0Hz, OCHACHBCH3), 1.22
(3H, t, J 7.0Hz, OCH2CH3), 1.05 (9H, s, C4H9), 0.99
(9H, s, C4H9); dC (100MHz, CDCl3) 134.4, 125.3 (CH,
C-2, C-3), 94.8 (CH, C-1), 70.7, 67.4 (CH, C-5, CH2,
C-6), 67.3, 64.3 (CH2, OCH2CH3, CH, C-4), 27.6, 27.2
(CH3, CMe3), 22.9, 20.2 (C, CMe3), 15.5 (CH3,
OCH2CH3); CIMS m/z 315 (MH+, 15), 269
(M+ÀC2H5O, 100); HRMS (CI): MH+, found
315.1992. C16H31O4Si requires 315.1991. 0.3ppm error.
4.3. (4aR,6S,8aS)-6-Ethoxy-4,4a,6,8a-tetrahydropyrano-
[3,2-d][1,3]dioxin-2-one 11
Ethyl 4,6-di-O-acetyl-2,3-dideoxy-a-D-erythro-hex-2-eno-
pyranoside 8 (3.04g, 11.8mmol), in 1:1 MeOH/Et3N
(60mL), together with a catalytic amount of K2CO3,
was heated at 50ꢁC under N2 for 18h. The solvent was
then filtered through Celite and the solvent removed in
vacuo to give the diol 9 as a pale brown viscous oil
(1.95g), which was used without purification. A portion
of 9 (504mg, 2.89mmol) was dissolved in pyridine
(1.4mL, 17.34mmol) and CH2Cl2 (13.8mL) under N2
and cooled to À70ꢁC, then a solution of triphosgene
(430mg, 1.45mmol) in CH2Cl2 (5.8mL), was added drop-
wise. Once addition was complete, the reaction was al-
lowed to warm to 20ꢁC over 18h. The resultant
homogenous solution was quenched with satd aq NH4Cl
and the aqueous portion was separated and extracted
with CH2Cl2 (3 · 10mL). The organic phases were com-
bined, washed with 1M HCl (2 · 5mL), satd aq NaH-
CO3 (2 · 5mL), brine (2 · 5mL), dried (Na2SO4),
filtered and concentrated in vacuo to give a crude cream
coloured oil. Purification by flash column chromatogra-
phy on silica gel (PE/EtOAc 5:1 as eluent), afforded 11
(341mg, 59%) as a white solid, mp 79–81ꢁC; Rf (PE/
EtOAc 1:1) 0.53; [a]D = À17.3 (c 1.02, CHCl3); mmax
CHCl3/cmÀ1 1763 (C@O), dH (400MHz, CDCl3) 6.10
(1H, d, J2,3 10.0Hz, H-3), 5.85 (1H, dt, J1,2 2.4, J2,3
10.0, J2,4 2.0Hz, H-2), 5.07 (1H, br s, H-1), 4.62 (1H,
4.5. (4aR,6R,8aS)-2,2-Di(tert-butyl)-4,4a,6,8a-tetrahy-
dropyrano[3,2-d ][1,3,2]dioxasilan-6-yl hydroperoxide 13
To ethyl 4,6-O-(di-tert-butyl)silanediyl-2,3-dideoxy-a-D-
erythro-hex-2-enopyranoside 12 (105mg, 0.33mmol) in
Et2O–H2O2 (1.0mL) at 0ꢁC was added concentrated
H2SO4 (two drops). The reaction mixture was stirred
for 16h and allowed to warm to 20ꢁC. The reaction mix-
ture was diluted with CH2Cl2 (5mL) and water (5mL),
then the aqueous phase was separated and washed with
CH2Cl2 (2 · 5mL). The organic phases were combined,
dried (MgSO4), filtered and concentrated in vacuo to
give crude 13 as a colourless oil (88mg, 87%). Flash col-
umn chromatography (PE/EtOAc 8:1 as eluent) afforded
1
13 as a colourless oil (75mg, 75%), which by H NMR
was an inseparable 5:1 a:b anomeric mixture; Rf (PE/
EtOAc 1:1) 0.71; [a]D = +24.4 (c 2.74, CHCl3); mmax
neat/cmÀ1 3366 (OOH), 2936, 2859 (2 · O–SiR), 1473,
1395, 1297, 1183; major anomer; dH (400MHz, CDCl3)
8.75 (1H, br s, OOH), 6.18 (1H, d, J2,3 10.0Hz, H-3),
5.58 (1H, dt, J1,2 2.4, J2,4 2.4, J2,3 10.0 Hz, H-2), 5.44
(1H, br d, J1,2 1.8Hz, H-1), 4.38 (1H, dd, J2,4 2.0, J4,5
0
dd, J3,4 1.5, J4,5 9.2Hz, H-4), 4.52 (1H, dd, J5,6 6.0, J6,6
0
9.8Hz, H-6), 4.26 (1H, dd, J5,6 10.7, J6,6 9.8Hz, H-60),
0
0
4.13 (1H, ddd, J5,6 6.0, J4,5 9.2, J5,6 10.7Hz, H-5),
3.88–3.80 (1H, dq, Jgem 9.5, J 7.0Hz, OCHACHBCH3),
3.62–3.54 (1H, dq, Jgem 9.5, J 7.0Hz, OCHACHBCH3),
1.23 (3H, t, J 7Hz, OCH2CH3); dC (100MHz, CDCl3)
148.0 (C, O@CO2), 128.8, 127.0 (CH, C-2, C-3), 95.0
(CH, C-1), 72.4, 70.3 (CH, C-5, CH2, C-6), 65.0, 60.9
(CH2, OCH2CH3, CH, C-4), 15.3 (CH3, OCH2 CH3);
CIMS m/z 218 (MNH4þ, 100), 201 (MH+, 20), 155
(M+ÀC2H5O, 10); HRMS (CI): MH+, found 201.0763.
C9H13O5 requires 201.0765. 0.9ppm error.
0
9.0Hz, H-4), 4.22 (1H, dd, J5,6 5.0, J6,6 10.0Hz, H-6),
0
3.92 (1H, dd, J5,6 10.0, J6,6 10.0Hz, H-60), 3.79 (1H,
0
0
ddd, J5,6 5.0, J4,5 9.0, J5,6 10.0Hz, H-5), 1.05 (9H, s,
C4H9), 0.99 (9H, s, C4H9); dC (100MHz, CDCl3)
138.2, 120.3 (CH, C-3, C-2), 98.9 (CH, C-1), 70.3,
67.9, 67.1 (CH, C-5, CH2, C-6, C-4), 27.6, 27.2 (CH3,
CMe3), 22.9, 20.2 (C, CMe3); CIMS m/z 302 (M+,
100), 285 (M+ÀOH, 56), 269 (M+ÀHO2, 70); HRMS
(CI): MNHþ4 , found 320.1893. C14H30NO5Si requires
320.1892. 0.3ppm error.
4.4. Ethyl 4,6-O-(di-tert-butyl)silanediyl-2,3-dideoxy-a-D-
erythro-hex-2-enopyranoside 12
4.6. Ethyl 2,3-dideoxy-4,6-di-O-pivaloyl a-D-threo-hex-2-
enopyranoside 14
The diol 9 (510mg, 2.93mmol) was dissolved in CH2Cl2
(20mL) under N2 and cooled to À60ꢁC. 2,6-Lutidine
(2.6mL, 22.6mmol) and di-tert-butylsilyl ditriflate
(2.97g, 6.74mmol) were added at À60ꢁC. The reaction
mixture was stirred at this temperature for 3h, then
diluted with EtOAc (25mL) and washed with H2O
Ethyl
4,6-di-O-benzoyl-2,3-dideoxy-a-D-threo-hex-2-
enopyranoside was obtained from diol 9 using a
previously described20 Mitsunobu procedure in 59%
yield. To this intermediate (4.50g, 11.8mmol) in MeOH

290
A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
(30mL) at rt under N2, was added sodium methoxide
(1mL, 25% solution in MeOH). The reaction was stirred
at rt until deprotection was complete (2d, TLC), then
concentrated in vacuo to give a yellow residue. Purifica-
tion by column chromatography (PE/EtOAc 1/1 then
EtOAc as eluent) gave the diol as a yellow solid. This
was dissolved in pyridine (20mL) and cooled to 0ꢁC, un-
der N2 followed by dropwise addition, over 30min, of
pivaloyl chloride (4.0mL, 32.7mmol). The reaction was
stirred at rt for 4d with further pivaloyl chloride
(4.0mL) added after 2d. On completion the reaction
was concentrated in vacuo and the residue obtained dis-
solved in CH2Cl2 (30mL), washed with water (30mL),
satd aq CuSO4 solution (5 · 30mL) and water again
(30mL), then dried (MgSO4), filtered and concentrated
in vacuo to give a brown oil. Purification by column
chromatography (PE/EtOAc 3:2 as eluent) gave the title
compound 14 (1.12g, 27.8%) as a colourless oil; Rf
(PE/EtOAc 7:1) 0.74; [a]D = À123.8 (c 0.9, CHCl3); mmax
(CHCl3)/cmÀ1 2975 (CH), 1734 (C@O), 1481, 1283, 1153,
1049; dH (270MHz, CDCl3) 6.06 (1H, dd, J2,3 10.0, J3,4
5.0Hz, H-3), 5.98 (1H, dd, J1,2 3.0, J2,3 10.0Hz, H-2),
5.03 (1H, d, J1,2 3.0Hz, H-1), 4.94 (1H, dd, J3,4 5.0, J4,5
2.5Hz, H-4), 4.35 (1H, dt, J4,5 2.5, J5,6 6.5Hz, H-5),
4.26–4.10 (2H, m, H-6, H-60), 3.83 (1H, dq, Jgem 9.5, J
7.0Hz, OCHAHBCH3), 3.54 (1H, dq, Jgem 9.5, J 7.0Hz,
OCHAHBCH3), 1.22 (9H, s, OPiv), 1.20 (3H, t, J
7.0Hz, OCH2CH3) 1.15 (9H, s, OPiv); dC (67.5MHz,
CDCl3) 177.9, 177.5 (C, OPiv), 130.5 (CH, C-3), 125.0
(CH, C-2), 93.4 (CH, C-1), 66.8, 62.5 (CH, C-4, C-5),
63.6, 62.7 (CH2, C-6, OC2CH3), 38.8, 38.6 (C, OPiv),
27.0, 26.4 (CH3, OPiv), 15.1 (CH3, OCH2C3); CIMS
C-4b, C-5b), 67.1, 61.6 (CH, C-4a, C-5a), 62.2 (CH2,
C-6b), 61.9 (CH2, C-6a), 38.9, 38.7 (C, OPiv), 27.0
(CH3, OPiv); CIMS m/z 348 (MNHþ4 , 8), 330 (M+,
66), 313 (M+ÀHO, 9) 297 (M+ÀHO2, 100), 195
(M+ÀC5H11O4, 92); HRMS (CI): MNHþ4 , found
348.2030. C16H30NO7 requires 348.2022. 2.3ppm error.
4.8. (2S,3R,6S)-6-Hydroperoxy-2-methyl-3,6-dihydro-
2H-pyran-3-yl acetate 18a and (2S,3R,6R)-6-hydroper-
oxy-2-methyl-3,6-dihydro-2H-pyran-3-yl acetate 18b
To di-O-acetyl-L-(+)-rhamnal 1723,24 (1.29g, 6.02mmol)
in Et2O–H2O2 [6.0mL, prepared by extracting 50% aq
H2O2 (5mL) into Et2O (2 · 10mL); CARE!]35 at 0ꢁC
was added concentrated H2SO4 (two drops, cat.). The
reaction mixture was stirred overnight with slow warm-
ing to 20ꢁC. The reaction mixture was diluted with
CH2Cl2 (20mL) and water (20mL), then the aqueous
phase was separated and washed with CH2Cl2
(2 · 15mL). The organic phases were combined, dried
(MgSO4), filtered and concentrated in vacuo to give a
very pale yellow oil (929mg, 82%), which was an a:b=
10:1 mixture of anomers by NMR. Separation by flash
chromatography (PE/EtOAc 4:1 as eluent) afforded pure
18a and 18b. a-Anomer 18a: Rf (PE/EtOAc 1:1) 0.38;
[a]D = À138.5 (c 0.84, CHCl3); mmax CHCl3/cmÀ1 3379
(OOH), 2983 (CH), 1742 (C@O), 1449, 1374, 1238,
1197, 1098; CIMS m/z 188 (M+, 100), 155 (M+ÀHO2,
60); HRMS (CI): MNHþ4 , found 206.2156. C8H16NO5
requires 206.2157. 0.5ppm error. dH (400MHz, CDCl3)
8.57 (1H, br s, OOH), 6.00 (1H, d, J2,3 10.0Hz, H-3),
5.72 (1H, dt, J1,2 2.7, J2,3 10.0, J2,4 2.0Hz, H-2), 5.48
(1H, br s, H-1), 5.08 (1H, dq, J1,4 1.4, J2,4 2.0, J3,4 1.6,
J4,5 9.2Hz, H-4), 3.97 (1H, dq, J4,5 9.2, J5,Me 6.2Hz,
H-5), 2.10 (3H, s, OAc), 1.28 (3H, d, J5,Me 6.2, Me);
dC (100MHz, CDCl3) 170.5 (C@O, OAc), 133.4, 122.9
(@CH, C-3, C-2), 98.6 (CH, C-1), 70.3, 65.6 (CH, C-5,
C-4), 21.0 (CH3, OAc); 17.8 (Me). b-Anomer 18b: Rf
(PE/EtOAc 1:1) 0.39; [a]D = À116.5 (c 0.84, CHCl3);
mmax CHCl3/cmÀ1 3379 (OOH), 2983 (CH), 1742
(C@O), 1449, 1374, 1238, 1197, 1098; dH (400MHz,
CDCl3) 8.60 (1H, br s, OOH), 6.05 (1H, ddd, J1,3 1.6,
J3,4 3.5, J2,3 10.0Hz, H-3), 5.87 (1H, dt, J1,2 1.6, J2,3
10.0, J2,4 1.8Hz, H-2), 5.60 (1H, br d, J1,2 1.6Hz, H-
1), 5.02 (1H, m, H-4), 3.99 (1H, dq, J4,5 5.4, J5,Me
6.5Hz, H-5), 2.09 (3H, s, OAc), 1.36 (3H, d, J5,Me
6.5Hz, Me); dC (100MHz, CDCl3) 170.6 (C@O, OAc),
129.9, 126.0 (@CH, C-3, C-2), 98.7 (CH, C-1), 70.8,
68.2 (CH, C-5, C-4), 20.6 (CH3, OAc); 18.2 (Me). CIMS
m/z 188 (M+, 100), 155 (M+ÀHO2, 60); HRMS (CI):
MNHþ4 , found 206.2156. C8H16NO5 requires 206.2157.
0.5ppm error.
m/z 360 (MNHþ, 6), 297 (M+ÀC2H5O, 100), 241
4
(M+ÀC6H13O, 8), 195 (M+ÀC7H15O3, 92), HRMS
(CI): MNHþ4 , found 360.2385. C18H34NO6 requires
360.2386. 0.3ppm error.
4.7. 2,3-Dideoxy-4,6-di-O-pivaloyl-D-threo-hex-2-enopyr-
anosyl hydroperoxide 3
Ethyl
2,3-dideoxy-4,6-di-O-pivaloyl-a-D-threo-hex-2-
enopyranoside 14 (1.07g, 3.1mmol) was treated with
67% aq H2O2 (3mL) as described for the preparation
of 1. After 5d the reaction was worked-up as previously
described, followed by purification by flash column
chromatography (PE/EtOAc 7:1 as eluent) to give the
title compound 3 (454mg, 43.5%), as a colourless oil,
an inseparable mixture of anomers (a:b = 10:1); Rf
(PE/EtOAc 7:1) 0.44; mmax (film)/cmÀ1 3407 (OOH),
2964 (C-H), 1725 (C@O), 1282, 1216, 1156; dH
(270MHz, CDCl3) 6.22 (1H, ddd, J1,3 1.5, J2,3 10.0,
J3,4 5.5Hz, H-3a), 6.12 (1H, ddd, J 1.5, J 4.5, J2,3
10.0Hz, H-3b), 5.94 (1H, dd, J1,2 3.0, J2,3 10.0Hz, H-
2a), 5.89 (1H, ddd, J 1.0, J 1.5, J2,3 10.0Hz, H-2b),
5.56–5.54 (1H, m, H-1b), 5.53 (1H, dd, J1,2 3.0, J1,3
1.5Hz, H-1a), 5.17–5.13 (1H, m, H-4b) 4.96 (1H, dd,
J3,4 5.5, J4,5 2.5Hz, H-4a), 4.40 (1H, dt, J4,5 2.5, J5,6
7.0Hz, H-5a), 4.36–4.30 (3H, m, H-5b, H-6a, H-6b),
4.9. Ethyl-4-O-pivaloyl-2,3,6-trideoxy-a,b-L-erythro-hex-
2-enopyranoside 19
To a solution of di-O-acetyl-L-(+)-rhamnal 17 (530mg,
2.47mmol) in CH2Cl2 (10.6mL) containing dry EtOH
(0.3mL, 5.14mmol) was added a solution of SnCl4 in
CH2Cl2 (22.7lL, 1.0M; 5mol%). The resulting mixture
was stirred at rt for 20min, quenched with satd aq NaH-
CO3 (ꢀ2.6mL) and then extracted (3 · 8.0mL) with
CH2Cl2. The organic phases were combined, washed
4.22–4.17 (1H, m, H-60b), 4.16 (1H, dd, J5,6 7.0, J6,6
0
10.5Hz, H-6a), 1.17 (9H, s, OPiv), 1.16 (9H, s, OPiv);
dC (67.5MHz, CDCl3) 178.3, 177.6 (C, OPiv), 128.7,
127.7 (CH, C-2b, C-3b), 128.6, 126.0 (CH, C-2a, C-
3a), 99.7 (CH, C-1b), 97.8 (CH, C-1a), 70.9, 63.3 (CH,

A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
291
with brine (ꢀ2.6mL), dried (MgSO4) and evaporated to
furnish a crude brown syrup as a mixture of anomers
a:b = 5.9:1 by 1H NMR (438mg, 88%). The crude mate-
rial [2.3g, 11.5mmol, Rf 0.54 (PE/EtOAc 1:1)] had spec-
troscopic data in accordance with literature values36 and
was progressed without further purification; the com-
pound in 1:1 MeOH/Et3N (16.0mL) along with a spat-
ula of K2CO3 under N2, was heated at 50ꢁC for 6h.
The solvent was then removed in vacuo to give a pale
brown oil. This was immediately taken up in pyridine
(10mL) under N2 and cooled to 0ꢁC, then pivaloyl chlo-
ride (2.4mL, 19.6mmol) was added dropwise. The reac-
tion mixture was stirred at rt for 24h then diluted with
water (15mL) and extracted with EtOAc (3 · 15mL).
The organic phases were combined, washed with satd
aq CuSO4 solution (5 · 15mL), water (2 · 15mL), satd
aq NaHCO3 (2 · 15mL), water again (2 · 15mL), dried
(MgSO4), filtered and concentrated in vacuo to give the
title compound 19 as a crude brownish orange oil. This
was purified by flash column chromatography (PE/
EtOAc 3:1 as eluent) furnishing 19 as an orange oil
(2.68g, 96%) Rf (PE/EtOAc 1:1) 0.74; [a]D = À124.7
(c 1.06, CHCl3); mmax neat/cmÀ1 1733 (C@O), 1480,
1280, 1156, 1058; dH (400MHz, CDCl3) 5.78–5.82 (2H,
m, H-2,3), 5.01 (1H, d, J4,5 9.5Hz, H-4), 4.95 (1H,
br s, H-1), 4.01–3.94 (1H, dq, J5,Me 6.4, J4,5 9.2Hz,
H-5), 3.86-3.77 (1H, dq, Jgem 9.5, J 7.0Hz, OCHAHB-
CH3), 3.57–3.49 (1H, dq, Jgem 9.5, J 7.0Hz, OCHA-
HBCH3), 1.24–1.19 (6H, m, OCH2CH3 + CHCH3),
1.17 (9H, OPiv); dC (100MHz, CDCl3) 178.0 (C, OPiv),
130.0, 128.6 (CH, C-2, C-3), 94.4 (CH, C-1), 70.7,
64.9 (CH, C-4, C-5), 64.2 (CH2, OCH2CH3), 38.9 (C,
OPiv), 27.1 (CH3, OPiv), 18.1 (Me), 15.5 (CH3,
OCH2CH3); CIMS m/z 197 (M+ÀC2H5O, 100). HRMS
(CI): MNHþ4 , found 260.1862. C13H26NO4 requires
260.1858. 1.5ppm error.
2.1Hz, H-2), 5.49 (1H, br s, H-1), 5.05 (1H, dt, J2,4
1.8, J3,4 1.5, J4,5 9.2Hz, H-4), 3.98 (1H, dq, J4,5 9.2,
J5,Me 6.4Hz, H-5), 1.26 (3H, d, J5,Me 6.4Hz, Me), 1.21
(9H, s, OPiv); dC (100MHz, CDCl3) 178.0 (C@O, OPiv),
133.8, 122.9 (@CH, C-3, C-2), 98.8 (CH, C-1), 70.1, 65.8
(CH, C-5, C-4), 39.0 (C, OPiv), 27.2 (CH3, OPiv), 18.0
(Me); CIMS m/z 230 (M+, 72), 213 (M+ÀOH, 85), 197
(M+ÀHO2, 100); HRMS (CI): MNHþ4 , found
248.1498. C11H22NO5 requires 248.1492. 2.4ppm error.
4.11. 2-(5-Cyclohexylpenta-20E,40E-dienamido)-5,6-
epoxy-1,4-benzoquinone 6:37 general procedure
To a solution of hydroperoxide (see Table 1, 0.14mmol)
in toluene (20mL) at room temperature was added DBU
(20mg, 0.14mmol). The mixture was stirred for 10min,
then
2-(5-cyclohexylpenta-20E,40E-dienamido)-1,4-
benzoquinone 4 (40mg, 0.14mmol) in toluene (20mL)
was added. The reaction was stirred until completion
(see Table 1), deemed to be the point when the solution
no longer produced a yellow colour on a TLC plate.
Water (5mL) was then added to quench the reaction.
Following addition of water and transfer to a separating
funnel, the two layers were separated and the aqueous
phase washed with CH2Cl2 (2 · 30mL). The organic
phases were combined, dried (MgSO4), filtered and con-
centrated in vacuo to give a brown residue. Purification
by flash column chromatography (CH2Cl2 as eluent)
gave the title compound 6 (for yields see Table 1) as a
pale yellow solid; mp 158–159ꢁC (lit.37 mp 159–
160ꢁC); Rf (PE/EtOAc 7:3) 0.82; dH (270MHz, CDCl3)
7.85 (1H, br s, NH), 7.61 (1H, d, J3,5 2.5Hz, H-3), 7.32
(1H, m, vinyl-H), 6.022–6.09 (2H, m, vinyl-H), 5.91 (1H,
0
0
d, J2 ,3 15.0Hz, H-2 ), 3.91 (1H, d, J5,6 3.5Hz, H-6), 3.82
(1H, dd, J3,5 2.5, J5,6 3.5Hz, H-5), 2.10 (1H, m, cy-H),
1.76–1.65 (4H, m, cy-H), 1.21 (6H, m, cy-H); dC
(67.5MHz, CDCl3) 191.1, 188.2 (C, C-1, C-4), 165.0
(C, C-10), 152.4 (C, C-2), 146.0, 125.4, 122.4, 120.0,
115.2 (CH, vinyl), 53.8, 52.5 (CH, C-5, C-6), 41.2
(CH, cy), 32.1, 25.9, 25.7 (CH2, cy). The data were con-
sistent with the literature values.37 Ee values (Table 1)
were determined both by polarimetry and by HPLC
analysis using a Daicel AD column eluted with n-hex-
ane/EtOH mixtures and UV detection.
4.10. (2S,3R,6R)-6-Hydroperoxy-2-methyl-3,6-dihydro-
2H-pyran-3-yl pivalate 20a, (2S,3R,6S)-6-hydroperoxy-
2-methyl-3,6-dihydro-2H-pyran-3-yl pivalate 20b
To ethyl-4-O-pivaloyl-2,3,6-trideoxy-a-L-erythro-hex-2-
enopyranoside 19 (1.01g, 4.17mmol) in Et2O–H2O2
(6.0mL) at 0ꢁC was added concentrated H2SO4 (two
drops). The reaction mixture was stirred overnight
slowly warming to 20ꢁC. The reaction mixture was di-
luted with CH2Cl2 (15mL) and water (15mL). The
aqueous phase was separated and washed with CH2Cl2
(2 · 10mL). The organic phases were combined, dried
(MgSO4), filtered and concentrated in vacuo to give a
colourless oil (825mg, 86%). Compounds 20a and 20b
were obtained as a 10:1 mixture of anomers, respec-
4.12. 2-Substituted-2,3-epoxy-1,4-naphthoquinones 7a
and 7e:15 general procedure
To a solution of hydroperoxide (see Table 1, 0.28mmol)
in toluene (20mL) at 20ꢁC, was added DBU (35mg,
0.23mmol). The mixture was stirred for 10min then
the
appropriate
2-substituted-1,4-naphthoquinone
1
tively, by HNMR. The crude mixture of hydroperox-
(0.23mmol) in toluene (20mL) was added. The reaction
was stirred until completion (see Table 1) then water
(5mL) was added to quench the reaction. Work-up,
purification and chiral analysis were the same as above
for 2-(5-cyclohexylpenta-20E,40E-dienamido)-5,6-epoxy-
1,4-benzoquinone 6.
ides was partially separable by flash column
chromatography (PE/EtOAc 8:1 as eluent), yielding
pure a-anomer 20a as a colourless oil (367mg, 38%),
which crystallised on standing in the freezer, as well as
a mixture of anomers a:b = 10:1 (400mg, 52%); a-ano-
mer 20a: Rf (PE/EtOAc 1:1) 0.52; [a]D = À159.6 (c
1.01, CHCl3); mmax neat/cmÀ1 3379 (OOH), 2983 (CH),
1742 (C@O), 1449, 1374, 1238, 1197, 1098; dH
(400MHz, CDCl3) 8.67 (1H, br s, OOH), 5.96 (1H, d,
J2,3 10.0Hz, H-3), 5.72 (1H, dt, J1,2 2.8, J2,3 10.0, J2,4
4.13. 2-Methyl-2,3-epoxy-1,4-naphthoquinone 7a
2-Methyl-1,4-naphthoquinone 5a25 was treated with sev-
eral hydroperoxides (see Table 1) as described above to

292
A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
yield the title compound 7a in the yields shown, as a yel-
low solid; mp 95–96ꢁC (lit.38 mp 96–97ꢁC); Rf (PE/
EtOAc 9:1) 0.72; dH (270MHz, CDCl3) 8.03–7.89 (2H,
m, Ar–H), 7.77–7.69 (2H, m, Ar–H), 3.85 (1H, s, H-3),
1.72 (3H, s, CH3).
Acknowledgements
We are grateful to Stylacats Ltd and SOCSA for finan-
cial support (studentship to AAB).
4.14. 2-Phenyl-2,3-epoxy-1,4-naphthoquinone 7e
References
2-Phenyl-1,4-naphthoquinone 5e26 (54mg) was treated
with several hydroperoxides (see Table 1) as described
above to yield the title compound 7e in the yields shown,
as a yellow solid; mp 61–62ꢁC (lit.38 mp 62–63ꢁC); Rf
(PE/EtOAc 9:1) 0.77; dH (270MHz, CDCl3) 8.12–7.98
(2H, m, Ar–H), 7.79–7.72 (2H, m, Ar–H), 7.52–7.36
(5H, m, Ph–H), 3.96 (1H, s, H-3).
1. Reviews: Jacobsen, E. N. In Catalytic Asymmetric Syn-
thesis; Ojima, I., Ed.; VCH: New York, 1993; p 159;
Wang, Z.-W.; Miller, S. M.; Anderson, O. P.; Shi, Y. J.
J. Org. Chem. 1999, 64, 6443; Katsuki, T.; Martin, V. S.
Org. React. 1996, 48, 1.
2. Katsuki, T.; Sharpless, K. B. J. Am. Chem. Soc. 1980, 102,
5974; Zheng, W.; Loebach, J. L.; Wilson, S. R.; Jacobsen,
E. N. J. Am. Chem. Soc. 1990, 112, 2801.
3. Review: Porter, M. J.; Skidmore, J. Chem. Commun. 2000,
22, 1215.
4. Julia, S.; Masanna, J.; Guixer, J.; Rocas, J.; Colonna, S.;
Annunziata, R.; Molinari, H. J. Chem. Soc. Perkin Trans.
1 1982, 1317, and references cited therein.
5. (a) Bentley, P. A.; Bergeron, S.; Cappi, M. W.; Hibbs, D.
E.; Hursthouse, M. B.; Nugent, T. C.; Pudilo, R.; Roberts,
S. M.; Wu, L. E. Chem. Commun. 1997, 739; (b) Allen, J.
V.; Drauz, K. H.; Flood, R. W.; Roberts, S. M.;
Skidmore, J. Tetrahedron Lett. 1999, 40, 5417; (c) Dhanda,
A.; Drauz, K. H.; Geller, T.; Roberts, S. M. Chirality
2000, 12, 313.
6. Wynberg, H.; Marman, B. J. Org. Chem. 1980, 45, 158.
7. Macdonald, G.; Alcaraz, L.; Lewis, N. J.; Taylor, R. J. K.
Tetrahedron Lett. 1998, 39, 5433.
4.15. Computational methods
All calculations were performed on a PC running Red-
Hat Linux 7.3.39 Molecular mechanics energy minimisa-
tions were performed using the default settings in SYBYL
6.91, expect that Gasteiger–Huckel charges were ac-
¨
counted for and 100,000 iterations were used.29 Con-
straints were used to help locate the transition states
and default settings were employed. For identifying
the approximate first transition state (peroxide attacking
˚
quinone), a constraint of 2A was chosen between the b-
carbon of the quinone and the peroxide anion. For the
generation of the approximate second TS (formation
of epoxide via ring closure and elimination of alkoxide),
8. Genski, T.; McDonald, G.; Wei, X.; Lewis, N.; Taylor, R.
J . K.Synlett 1999, 795.
˚
two distance constraint of 1.9A were placed between
9. Preliminary communication: Dwyer, C. L.; Gill, C. D.;
Ichihara, O.; Taylor, R. J. K. Synlett 2000, 704.
10. Hamann, H.-J.; Hoft, E.; Mostowicz, D.; Mishnev, A.;
Urbanczyk-Lipkowska, Z.; Chmielewski, M. Tetrahedron
1997, 53, 185.
11. Mostowicz, D.; Jurczak, J.; Hamann, H.-J.; Hoft, E.;
Chmielewski, M. Eur. J. Org. Chem. 1998, 2617.
12. Adam, W.; Korb, M. N. Tetrahedron: Asymmetry 1997, 8,
1131, and references cited therein; Adam, W.; Rao, P. B.;
Degen, H.-G.; Saha-Mo¨ller, C. R. J. Am. Chem. Soc.
2000, 122, 5654.
13. Taylor, R. J. K.; Alcaraz, L.; Kapfer-Eyer, I.; Macdonald,
G.; Wei, X.; Lewis, N. Synthesis 1998, 775.
14. Alcaraz, L.; Macdonald, G.; Ragot, J. P.; Lewis, N.;
Taylor, R. J. K. J. Org. Chem. 1998, 63, 3526.
15. Arai, S.; Oku, M.; Miura, M.; Shiori, T. Synlett 1998,
1201.
16. Ferrier, R. J.; Prasad, N. J. Chem. Soc. (C) 1969, 570;
Pedretti, V.; Mallet, J.-M.; Sinay, P. Carbohydr. Res. 1993,
244, 247.
17. Descotes, G.; Martin, J.-C. Carbohydr. Res. 1977, 56, 168.
18. Burk, R. M.; Roof, M. B. Tetrahedron Lett. 1993, 34, 395.
19. Speybroeck, R. V.; Guo, H.; Van der Eycken, J.;
Vandewalle, M. Tetrahedron 1991, 47, 4675.
20. Grynkiewicz, G. Bull. Acad. Pol. Sci. Ser. Sci. Chim. 1978,
26, 537; Georges, M.; Mackay, D.; Fraser-Reid, B. Can. J.
Chem. 1984, 62, 1539.
the terminal peroxide oxygen and the carbons at either
end of the double bond. An angle constraint was applied
between the carbon of the double bond, terminal perox-
ide oxygen and the adjacent oxygen of 146ꢁ and the
whole system then underwent energy minimisation.30
Input files for MOPAC calculations (atom types and
connectivity) were obtained using SYBYL 6.91.29 Semi-
empirical calculations were carried out in the gas phase
using MOPAC with the AM1 Hamiltonian.28 Molecules
were modelled in vacuo as MOPAC does not model tolu-
ene, a solvent with a low dielectric constant,40 very well.
MOPAC 7 is freely available from internet sources.41 All
structures were optimised using the eigenvector follow-
ing algorithm. For each molecule and transition state,
the vibrational spectrum was produced by a FORCE
calculation. Both ground states and transition states
(TS) were confirmed by the presence of six eigenvalues
that were very small (less than ꢀ5cmÀ1). A TS was fur-
ther characterised by one, and only one, negative force
constant corresponding to atomic motion along the
reaction coordinate.31 It was also necessary to show that
a proposed TS connects the desired reactant(s) to the
product(s). This was achieved by performing two intrin-
sic reaction coordinate (IRC) calculations. The first IRC
began with an initial perturbation of the atomic coordi-
nates in the direction of the single negative frequency
and the second IRC calculation with a negative pertur-
bation of the atomic coordinates along the same normal
coordinate. By joining the results of the two calculations
together it was easy to tell if the TS does indeed link
reactants to products.
21. Wyss, P. C.; Kiss, J.; Arnold, W. Helv. Chim. Acta 1975,
58, 1847.
22. Fehlhaber, H. W.; Snatzke, G.; Vlahov, I. Liebigs Ann.
Chem. 1987, 637.
23. Renneberg, B.; Li, Y.-M.; Laatsch, H.; Fiebig, H.-H.
Carbohydr. Res. 2000, 329, 861.
24. Bredenkamp, M. W.; Holzapfel, C. W.; Toerien, F. Synth.
Commun. 1992, 2459.

A. Bundu et al. / Tetrahedron: Asymmetry 16(2005) 283–293
293
25. Arnold, T.; Lawson, R. J. J. Org. Chem. 1940, 5, 240.
26. Best, W. M.; Sims, C. G.; Winslade, M. Aust. J. Chem.
2001, 54, 401.
27. Pluim, H.; Wynberg, H. J. Org. Chem. 1980, 45, 2498.
28. Stewart, J. J. P. J. Comp. Chem. 1989, 10, 209.
29. SYBYL version 6.91, 1699, South Hanley Road, St. Louis,
MO 63144-2319.
35. Orosz, G.; Given, R. S.; Schowen, R. L. Anal. Chim. Acta.
1992, 266, 219; Saito, I.; Nagata, R.; Yuba, K.; Matsuura,
T. Tetrahedron Lett. 1983, 24, 1737–1740; Schmidt, M.;
Bornmann, P. Zeitschrift fu¨r Anorg. Allgem. Chemie 1964,
330, 309–310. Extreme caution should be exercised using
this procedure; see: Heaney, H. Aldrichimica Acta 1993,
26, 35–45.
30. Force Field Manual in SYBYL version 6.91, 1699, South
Hanley Road, St. Louis, MO 63144-2319.
31. Leach, A. R. Molecular Modelling; Longman Singapore
Publishers, 1996; p 240.
36. Priebe, W.; Bhate, P.; Horton, D. Carbohydr. Res. 1985,
144, 331.
37. Wei, X.; Grove, J. J. C.; Taylor, R. J. K. J. Chem. Soc.,
Perkin Trans. 1 1999, 1143.
32. Grigg, R.; Hadjisoteriou, M.; Kennewell, P.; Markandu,
J.; Thornton-Pett, M. J. Chem. Soc., Chem. Commun.
1993, 1340.
33. Adam, W.; Rao, P. B.; Degen, H. G.; Saha-Mo¨ller, C. R.
Eur. J. Org. Chem. 2002, 630.
38. Wynberg, H.; Marsman, B. G. J. Org. Chem. 1980, 45,
158; Snatzke, G.; Wynberg, H.; Feringa, B.; Marsman,
B. G.; Greydanus, B.; Pluim, H. J. Org. Chem. 1980, 45,
4094.
39. http://www.redhat.com.
40. http://www.asiinstr.com/dc1.html#SECTION-T.
41. http://www.webmo.net/support/mopac_linux.html.
34. Prasad, J. S.; Clive, D. L. J.; da Silva, G. V. J. J. Org.
Chem. 1986, 51, 2717.