Dihydroisoquinolinium salts: Catalysts for asymmetric epoxidation

Article abstract of DOI:10.1039/b002156n

A range of dihydroisoquinolinium salts were prepared and tested as asymmetric epoxidation catalysts. The factors affecting enantioselectivity of catalytic asymmetric epoxidation process were also discussed. The asymmetric epoxidation reaction employed alkaline hydrogen peroxide and a chiral polypeptide derived from a naturally occurring amino acid, leucine.

Full text of DOI:10.1039/b002156n

View Article Online / Journal Homepage / Table of Contents for this issue
Dihydroisoquinolinium salts: catalysts for asymmetric epoxidation
Philip C. Bulman Page,*a Gerasimos A. Rassias,a David Barros,a Donald Bethellb and
Mark B. Schillingc
1
a Department of Chemistry, Loughborough University, Loughborough, Leicestershire,
UK LE11 3TU. E-mail: p.c.b.page@lboro.ac.uk
b Robert Robinson Laboratories, Department of Chemistry, University of Liverpool,
Oxford Street, Liverpool, UK L69 3BX
c Glaxo Wellcome Research & Development, Gunnels Wood Road, Stevenage, Hertfordshire,
UK SG1 2NY
Received (in Cambridge, UK) 16th March 2000, Accepted 12th July 2000
First published as an Advance Article on the web 11th September 2000
A range of dihydroisoquinolinium salts has been prepared and tested as asymmetric epoxidation catalysts
in an investigation of the reaction mechanism and the factors affecting enantioselectivity in this process.
derived from a naturally occurring amino acid, typically leu-
cine. Good to excellent enantioselectivities are obtained, but
essentially only for α,β-unsaturated ketone substrates with aryl
substituents. Further, the process is not homogeneous, but
multiphasic. These features again limit the potential of this
method as a synthetic tool for the construction of chiral
epoxides from a more general pool of substrates.16
Introduction
Optically active epoxides are important building blocks in
asymmetric synthesis.1 Due to their high versatility, they have
frequently been employed as key intermediates in synthesis. In
addition, the epoxide functionality is itself found as part of the
structure of many biologically active compounds and natural
products.2 Over recent years, a number of research groups
around the world have worked towards the development of
methodologies to access chiral epoxides with high enantioselec-
tivity. The greatest impact in the construction of chiral epoxides
has been made with the introduction of catalytic systems. The
most popular precursors remain olefins, but approaches have
been also made which involve carbonyl compounds as the start-
ing materials.3 Undoubtedly the best-known catalytic process is
that of Sharpless for the epoxidation of allylic alcohols, usually
with greater than 90% ee.4
Attempts have also been made to mimic the biosynthetic
pathway of epoxide formation using cytochrome P450 anal-
ogues, which are based on chiral porphyrin–transition metal
complexes.5 Although a plethora of chiral complexes based
upon cyclic polyamines/polyimines has been synthesized,6 the
enantioselectivities have remained low. Limited generality has
been observed, styrenes usually being the best substrates, and
this restricts the applicability of the method.7
Chiral dioxiranes, generated by the reaction of OxoneTM
(Caroate) with chiral ketones, have also been developed as
reagents for asymmetric epoxidation.17 Until recently, however,
the enantioselectivities obtained with these chiral oxidants were
only moderate.18 Over the last three years, independent work by
Yang, Shi, and Armstrong has identified chiral ketones whose
derived dioxiranes are among the best chiral epoxidizing agents.
Yang and co-worker’s C2 symmetric ketone in the presence of
alkaline OxoneTM, was shown to epoxidize trans-stilbene with
84% ee.19 Under similar conditions, Armstrong and Hayter’s
chiral α-fluoro ketone mediates epoxidation of triphenylethyl-
ene with 83% ee.20 Shi and co-worker’s fructose derived chiral
ketone catalyses the asymmetric epoxidation of a wide variety
of alkenes, including allylic and homoallylic alcohols and
ethers, with good to excellent enantioselectivities.21
Oxaziridines are the nitrogen analogues of dioxiranes,
and constitute an important class of organic oxidants under
non-aqueous conditions. Indeed, Davis and co-worker’s
oxaziridines22 and related systems23 are very successful in
the asymmetric oxidation of sulfides to sulfoxides, but are
much less successful with less potent nucleophilic substrates
such as alkenes.
Oxaziridinium salts, a related electrophilic oxygen source first
reported in 1976,24 and subsequently investigated by Lusinchi
and co-workers,25 have been shown to be extremely reactive for
oxygen transfer to nucleophilic substrates, including sulfides
and alkenes.26 These organic salts were prepared either by
quaternization of the corresponding oxaziridines or by peracid
oxidation of iminium salts. More importantly, dihydroiso-
quinolinium salts 1 were shown to catalyse the epoxidation of
simple olefins, the corresponding racemic oxaziridinium salts
2 (R = RЈ = H) being presumed to be the active oxidants.27
Chiral salen complexes of transition metals,8 which bear
some similarity to porphyrin systems, have been developed by
Jacobsen,9 Katsuki and others as catalysts for asymmetric
epoxidation. A high degree of substrate specificity is observed
however, for example, tetrasubstituted10 and acyclic E-alkenes11
tend to give lower ees than trisubstituted and aryl alkenes.12
More seriously, the reaction is not stereospecific, leading to loss
of the geometry around the former double bond, particularly in
the case of aryl alkene substrates. Jacobsen and co-workers
have, however, developed a modification to allow the generation
of trans epoxides as the major products derived from Z-alkene
substrates by addition of cinchona alkaloid-derived salts.13
Catalytic asymmetric epoxidation has also been studied using
purely organic molecules as the mediators. Chiral peracids are
of limited value for asymmetric epoxidation,14 and an asym-
metric version of the analogous Payne epoxidation procedure,
mediated by nitrile and hydrogen peroxide, has provided high
ees only when using as mediator a nitrile derived from a
relatively inaccessible chiral helicene.15
The Julia–Colonna asymmetric epoxidation reaction
employs alkaline hydrogen peroxide and a chiral polypeptide
DOI: 10.1039/b002156n
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334
This journal is © The Royal Society of Chemistry 2000
3325

View Article Online
The first enantiomerically pure oxaziridinium salt 2 (R = Ph,
RЈ = Me; X = BF4Ϫ) was prepared by quaternization of an
oxaziridine derived from a chiral imine, prepared in turn in
four steps from norephedrine.28 This salt was shown to induce
asymmetric epoxidation of alkenes, and furthermore the corre-
sponding iminium salt was shown to catalyse epoxidation using
OxoneTM as the stoicheiometric oxidant; ees of up to ca. 40%
have been obtained. No cis-epoxide products are observed,
suggesting a single-step oxygen transfer process. More recently,
Aggarwal and Wang have reported the similar catalytic
asymmetric epoxidation of simple alkenes mediated by a
binaphthalene-derived iminium salt.29 This system provided
1-phenylcyclohex-1-ene oxide with 71% ee and trans-stilbene
oxide with 31% ee, also with complete retention of stereo-
chemistry. Armstrong and co-workers have shown that even
ing to the procedure of Rieche and Schmitz:33 Bromination of
isochromane yields the expected 1-bromoisochromane which
may be isolated by distillation if required. Treatment of the
initial product with concentrated aqueous HBr furnishes the
desired bromoaldehyde in good yield (85%); it may also be
purified by distillation, but the crude product (ca. 80% pure,
contaminated principally by 2-vinylbenzaldehyde) can be used
with equal success in most cases. Cyclocondensation directly
with the appropriate amine takes place, generally at room
temperature, to give the iminium salts in good yields.
This sequence has proved invaluable for the screening of
asymmetric catalysts derived from a wide variety of chiral
amines. Furthermore, the overall process employed in the syn-
thesis of the catalyst is rapid, inexpensive, involves simple syn-
thetic steps, does not require chromatography, and is easily
scaled up. The catalyst preparations described below have been
performed on scales of up to 70 g without any difficulties.
Chiral primary amines react smoothly and rapidly with 2-
(2-bromoethyl)benzaldehyde to furnish the corresponding
dihydroisoquinolinium bromide salts. These organic salts, with
a few exceptions, are oils, and difficulties were encountered on
attempting to purify them by conventional methods. This prob-
lem was solved by counter-anion exchange, which takes place
readily simply by addition of an appropriate inorganic salt to
the reaction mixture before work up. Fluoroborate, hexafluoro-
phosphate, perchlorate, and periodate salts of these derivatives
did not prove ideal, but the tetraphenylborate salts 3, produced
acyclic iminium salts can catalyse epoxidation by OxoneTM 30
.
Perhaps because this process is relatively new, no thorough
investigation of the parameters involved in this class of
catalytic asymmetric process has yet been reported. We have
previously described our approach to a new type of cyclic chiral
iminium salt containing the asymmetric centres in an exocyclic
substituent at nitrogen, and these iminium salts have been suc-
cessfully employed in the catalytic asymmetric epoxidation of
simple alkenes.31 Herein we describe our examination of some
of the parameters that influence the yield, enantioselectivity
and reaction rate, and we discuss a possible mechanistic ration-
ale for the process.
Results and discussion
Catalyst design and synthesis
The exocyclic group attached to the nitrogen atom of the
oxaziridinium salt is invariably methyl or ethyl in the examples
published prior to our own work and discussed above. The
approach we have taken towards the design of this class of
chiral catalysts is entirely different in concept. We reasoned that
attachment of the controlling asymmetric centres (e.g. on exo-
cyclic carbon atoms) to the iminium nitrogen atom would bring
the asymmetric centre of the catalyst nearer to the site of the
reaction, and might therefore be expected to lead to higher ees.
Our preliminary experiments revealed that acyclic iminium salts
derived for example from benzaldehyde and 2,4-dichlorobenz-
aldehyde do not promote the desired reaction to any reasonable
extent because of their instability towards hydrolysis under the
alkaline conditions employed in the reaction. These findings
are in accord with the work of Armstrong et al.30 We therefore
chose cyclic iminium salts for investigation, and we selected the
dihydroisoquinolinium nucleus as a basis for study.
by use of sodium tetraphenylborate, were most suitable: they
are all solids, with the degree of crystallinity varying with the
alkyl group of the parent amine. As a consequence of a side
reaction, the elimination of hydrogen bromide from the bromo-
ethyl moiety of the precursor, the yields are generally between
30 and 80%. Very hindered amines give inferior conversions,
typically of the order of 20–30%, presumably due to their
increased tendency to act as bases rather than nucleophiles, and
indeed increased amounts of 2-vinylbenzaldehyde (and the
corresponding imine) are observed in these cases. Nevertheless,
the very small quantities required to catalyse the desired reac-
tions more than compensate for the moderate yields and cost of
some of the chiral amines screened. Typically, for the epoxid-
ation of 1-phenylcyclohex-1-ene on a 3 mmol scale, less than
eight milligrams of most catalysts were used for quantitative
conversions within 1 hour. Overall, the synthesis of the catalyst
and the asymmetric synthesis of epoxides together may usually
be completed within eight hours.
Formally, such catalysts might be prepared through conden-
sation of primary amines with a bromoaldehyde as shown in
Scheme 1, a hitherto little-explored synthetic approach,32 but
Epoxidation reactions
After some experimentation, optimum conditions were estab-
lished for the catalytic asymmetric epoxidation of 1-phenyl-
cyclohex-1-ene 4: 0 ЊC in water–acetonitrile (1:1); typically 0.5
to 10 mol% of the dihydroisoquinolinium salt; two equivalents
of OxoneTM and four equivalents of sodium carbonate. The
results are summarized in Table 1 for a selection of primary
amines 5–14. Reactions were in general complete within the one
hour timescale allowed, entries c and g (fenchyl)† being excep-
tions. It is important to note that blank reactions carried out in
parallel under the same conditions, in the presence of OxoneTM
,
but without catalyst, gave no reaction over up to eight hours
when four equivalents of sodium carbonate were present.
Scheme 1
† The IUPAC name for fenchyl is 1,3,3-trimethylbicyclo[2.2.1]heptyl; for
myrtanyl is 6,6-dimethylbicyclo[2.2.1]hept-2-ylmethyl; for isobornyl
is exo-bornyl and for isopinocampheyl is 2,6,6-trimethylbicyclo-
[3.1.1]heptyl.
one with the great advantage that a wide variety of readily
available chiral primary amines can be used. The key inter-
mediate, 2-(2-bromoethyl)benzaldehyde, was prepared accord-
3326
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334

View Article Online
Table 1 Epoxidation of 1-phenylcyclohex-1-ene with dihydroisoquinolinium tetraphenylborate salts derived from chiral aminesa
Amine
Catalyst load
(mol%)
Yield
(%)
Ee
Entry
precursor39
(%)40
Configuration41
a
b
c
d
e
f
g
h
i
5
6
5.0
5.0
1.0
0.5
0.5
0.5
0.5
0.5
0.5
0.5
54
70
39
63
68
66
45
60
58
47
0
0
7
25
19
27
12
32
18
8
(Ϫ)-(S,S)
(ϩ)-(R,R)
(ϩ)-(R,R)
(Ϫ)-(S,S)
(ϩ)-(R,R)
(ϩ)-(R,R)
(ϩ)-(R,R)
(Ϫ)-(S,S)
8
9
10
11
12
13
14
j
14
a Conditions: 4 equiv. OxoneTM; 8 equiv. Na2CO3; 0 ЊC; H2O–MeCN 1:2; reactions were monitored by TLC.
Scheme 3
(Scheme 3). Each may deliver the oxygen atom to either of the
prochiral faces of the alkene substrate with a different degree of
enantiocontrol, and they may be in competition for the alkene
substrate. Further, even if oxaziridinium ion is the oxidative
intermediate, it remains unclear whether these epoxidation
reactions proceed through spiro or planar transition states,34
although recent theoretical work provides evidence for the spiro
transition state.35
With the first two entries in Table 1, using the structurally
simplest amines, no asymmetric induction was observed, and it
became clear that a conformationally more defined and rigid
system was required to impart reasonable enantioselectivities.
The catalyst corresponding to entry c (the cis-3-amino-2-
phenylpiperidine system) appears to decompose under the
reaction conditions, perhaps accounting for the incomplete
conversion. Acyclic iminium salts are unstable towards
hydrolysis under our reaction conditions. It may be in this case
that the cyclization to give the iminium salt is less favourable
due to steric and/or conformational restrictions in the piper-
idine derivative.
There is therefore no significant competition under these
reaction conditions from achiral epoxidation processes, such as
Payne-type oxidation mediated by the acetonitrile solvent, and
indeed the epoxidation reaction does proceed in dichloro-
methane as solvent in the absence of acetonitrile.
A possible catalytic cycle for an oxaziridinium ion as the
oxidative intermediate is depicted in Scheme 2. The first stage is
Both camphor-based (entries h, i) and menthyl (entry d) sys-
tems gave disappointing ees, although these are two of the more
common systems upon which chiral auxiliaries and catalysts
have been based, often imparting high selectivity. The factors
which determine enantioselectivities remain unclear, since the
myrtanyl† system (entry f), which has its asymmetric centres
relatively remote from the reaction site, provides greater select-
ivity than does the isobornyl† system (entry i). It is clear
from the cases examined that increased steric hindrance near
the reaction site is important, but is not the only factor which
governs enantioselectivity. For example, the fenchyl derivative
(entry g), which is the most sterically demanding, gave the best
enantioselectivity under these conditions, albeit at the expense
of the apparent rate of the process. The N-(isopinocampheyl)†
dihydroisoquinolinium salt (entry e), which has considerably
less steric requirements than the fenchyl, is however almost as
selective, although in a faster reaction. Further, it is signifi-
cantly more selective than the camphor, menthyl and steroidal
systems, all of which are of similar or higher steric demand,
Scheme 2
the formation of an initial adduct 15, uncharged at nitrogen,
formed by (probably reversible) nucleophilic attack of the oxid-
ant on the iminium salt 3; this reaction would be expected to be
retarded in polar solvents. This is followed by irreversible expul-
sion of sulfate to give the oxaziridinium ion, a reaction which
would be expected to be accelerated in polar solvents, but which
is we believe the rate-determining step under our reaction con-
ditions. Oxygen may then be transferred to a substrate in a
subsequent step, the rate of which would not be expected to
have any great solvent dependence. An interesting but compli-
cating feature of these processes is that it is not one but two
diastereoisomeric oxaziridinium salts which may be formed, by
attack of oxidant at the Si or Re face of the iminium species
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334
3327

View Article Online
Table 2 Effect of counter-ion on epoxidation of 1-phenylcyclohex-1-
ene with dihydroisoquinolinium salts derived from amine 9a
tion rate. The effect is more pronounced when small amounts
of catalysts are used. Thus, with the isopinocampheylamine
derivative (entry e) (0.5 mol%), the yield of 1-phenylcyclohex-1-
ene oxide was approximately 30% after one hour at 0 ЊC when a
1:1 ratio of the two solvent was used, but the yield was essen-
tially quantitative at 2:1 (water–acetonitrile). Reducing the
amount of OxoneTM and base by a factor of 2 (i.e. using one
equivalent of OxoneTM and two equivalents of sodium carbon-
ate), resulted in incomplete conversion after one hour in the
improved (2:1) solvent system. In accord with previous sugges-
tions,17,19,21,36 this may result from competitive decomposition
of OxoneTM under the basic conditions, and hence, in a faster
reaction, more of the unstable oxidant is consumed in the
desired oxygen transfer process. Higher catalyst loadings
accelerate the rate of the reaction to an extent that outweighs
the effect of water content. Thus, when the water content was
varied successively from 10 to 50 to 66 and finally to 90% in
acetonitrile with 5 mol% catalyst loading (perchlorate salt), all
reactions indicated complete consumption of 1-phenylcyclo-
hex-1-ene within 20 minutes. This change in solvent com-
position did not however influence the enantioselectivity of
the reaction: for example, all reactions of the perchlorate salt
in acetonitrile–water produced 1-phenylcyclohex-1-ene oxide
with 18–20% ee with remarkable consistency.
Entry
Counter-ion
Ee (%)40
Configuration41
a
b
c
d
e
BPh4
BF4
PF6
ClO4
IO4
40
28
28
20
35
(ϩ)-(R,R)
(ϩ)-(R,R)
(ϩ)-(R,R)
(ϩ)-(R,R)
(ϩ)-(R,R)
a Conditions: 5 mol% catalyst; 2 equiv. OxoneTM; 4 equiv. Na2CO3; 0 ЊC;
H2O–MeCN 1:1; reactions were monitored by TLC. Reactions were
judged to be complete when all alkene had been consumed (р45 min in
all cases).
despite the similarity in the yields obtained after a one hour
reaction time between all these systems.
The reaction parameters
An examination of some of the reaction parameters was carried
out in order to optimize the reaction conditions with respect to
the enantioselectivity of the oxygen transfer process. The iso-
pinocampheylamine derivative exhibited one of the best pro-
files in terms of ees and rates; further, both enantiomeric forms
of the chiral amine are commercially available. For these
reasons it was chosen as the model catalyst for optimization
studies and for investigation of parameters that may influence
the enantioselectivity of the catalytic reaction.
This dramatic effect of the presence of water on the rate of
the reaction is perhaps a result of increasing OxoneTM solubil-
ization with increasing water concentration, such that nucleo-
philic attack by persulfate on the iminium species is speeded up,
leading to a concentration effect coupled with better solvation
of departing sulfate ion.
Effect of counter-ion. Although in several cases we were
unable to isolate tetrafluoroborate salts by anion exchange of
the iminium bromides, the isopinocampheylamine-derived
dihydroisoquinolinium salt was successful in forming precipi-
tates with a variety of counter-ions. Thus, in addition to the
original tetraphenylborate, the corresponding tetrafluoro-
borate, hexafluorophosphate, perchlorate and periodate salts
were also prepared. The latter two were formed in somewhat
better yields than the fluorine-containing examples.
Such an apparent rate change without change in enantio-
selectivity suggests that the rate-determining step does not
involve oxygen transfer to the substrate, i.e. that the sub-
sequent enantiocontrolling oxygen transfer to alkene is not the
rate-determining step under these conditions, and is consistent
with the catalytic cycle outlined in Scheme 2.
We have briefly investigated any potential correlation of reac-
tion rates and extent of asymmetric induction with the polarity
of the co-solvent, as this could prove useful in clarification of
the reaction mechanism and therefore the factors that control
the enantioselectivity of the process. The co-solvents used were
selected so that they differed significantly in dielectric constant
(ε, indicated by the values in brackets): dichloromethane
(8.9), trifluoroethanol (26.7), acetonitrile (37.5), water (78.4),
formamide (111).
The epoxidation of 1-phenylcyclohex-1-ene with the isopino-
campheyl catalyst (entry e) (10 mol%) was tested using these
co-solvents with water in 1:1 ratio (Table 3). In order also to
examine the counter-ion effect, the catalyst was tested both as
its perchlorate and tetraphenylborate salts (20 and 40% ee
respectively, in acetonitrile). The perchlorate salt exhibited an
interesting trend. In trifluoroethanol, it mediated the epoxid-
ation of 1-phenylcyclohex-1-ene within 30 minutes with quanti-
tative conversion and 26% ee, while in dichloromethane the
reaction was unusually slow (50% conversion after three hours),
but the ee of the epoxide was still increased to 33%. In form-
amide however there was no reaction, and this was also the case
for the tetraphenylborate salt. The lack of reaction in form-
amide regardless of the counter-ion involved perhaps suggests
that the iminium species are too well stabilized/solvated, and
the possibility of an irreversible attack by the formamide can-
not be dismissed. Oddly, while the reaction proceeds slowly in
DMF (65% conversion after 4 hours), the epoxide product is
racemic. The tetraphenylborate salt mediated the same reaction
in trifluoroethanol within the expected time scale (20–30 min-
utes) and with the same ee as the perchlorate salt (26%). In the
chlorinated solvent system however, no reaction was manifested
even after three hours. The difference in reactivity for both
counter-ions with dichloromethane as the co-solvent reflects
the poor miscibility of the two solvents, which must severely
All of the salts were tested in the asymmetric catalytic epox-
idation of 1-phenylcyclohex-1-ene (Table 2). A catalyst load-
ing of 5 mol% was used in 1:1 water–acetonitrile solvent in
the presence of two equivalents of OxoneTM and four equiv-
alents of sodium carbonate at 0 ЊC. The enantioselectivities
obtained exhibited an interesting trend. The periodate salt
produced an enantioselectivity (35%) comparable to those
obtained with the tetraphenylborate species (40% ee), while the
fluoride-containing counter-ions afforded lower ees (28%). The
perchlorate salt furnished inferior enantioselectivities (20% ee).
All of the salts however invariably produced the same enantio-
mer of the epoxide product (R,R) as the major component, and
all of the reactions were complete within the same time scale,
ca. 45 minutes. Control experiments were carried out to prove
that neither perchlorate nor periodate were acting as oxidants
towards either the substrate or the iminium salt under the reac-
tion conditions. Although the differences in ees observed are
not large, they are reproducible, and this behaviour would be
consistent with the existence of the catalyst in the reaction
medium as ion pairs, with the tightness of electrostatic bonding
depending upon the relative polarizability of each counter-ion.
Single crystal X-ray analysis of the isopinocampheyl tetra-
phenylborate catalyst suggests the presence of π-stacking
between the dihydroisoquinoline unit and one phenyl group
of the counter-ion. Presumably the counter-ion moderates the
intrinsic enantioselectivity of the oxaziridinium species by
providing an additional contribution to the activation barrier
leading to the two diastereoisomeric transition states.
Effect of the solvent system. An increase in water to
acetonitrile ratio is accompanied by an increase in the reac-
3328
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334

View Article Online
Table 3 Effect of cosolvents on epoxidation of 1-phenylcyclohex-1-ene with dihydroisoquinolinium salts derived from amine 9a
Entry
Cosolvent
Counter-ion
Time/h
Conversion (%)
Ee (%)40
Configuration41
a
b
c
d
e
f
g
h
CH2Cl2
CF3CH2OH
CH3CN
H2NCHO
CH2Cl2
CF3CH2OH
CH3CN
ClO4
ClO4
ClO4
ClO4
BPh4
BPh4
BPh4
BPh4
3
50
100
100
0
33
26
20
—
—
26
40
—
(ϩ)-(R,R)
(ϩ)-(R,R)
(ϩ)-(R,R)
0.5
0.5
3
3
0
0.5
0.5
3
100
100
0
(ϩ)-(R,R)
(ϩ)-(R,R)
H2NCHO
a Conditions: 10 mol% catalyst; 2 equiv. OxoneTM; 4 equiv. Na2CO3; 0 ЊC; H2O–solvent 1:1; reactions were monitored by TLC. Conversion measured
by 1H NMR spectroscopy.
limit the availability of the inorganic oxidant in the organic
phase.
The epoxidation reaction was also tested using the periodate
salt in trifluoroethanol, and again produced the epoxide with
26% ee, leading inexorably to the conclusion that since the
other parameters were kept constant, generation of a trifluoro-
ethoxide counter-ion may occur in all three cases under the
reaction conditions (pKA of CF3CH2OH = 12.37).37
Effect of temperature. The effect of this parameter on the
reaction is also difficult to study over a wide range. This is
partly due to the instability of OxoneTM under alkaline condi-
tions, which increases at higher temperatures, and partly due to
the freezing point of aqueous solvent mixtures and to reagent
solubility below 0 ЊC.
When the solvent system used was of high acetonitrile to
water ratio (1:1), the reaction at Ϫ10 ЊC was sluggish as the
Fig. 1
solubility of the inorganic oxidant and base in water is dramat-
using a fixed concentration of 1-phenylcyclohex-1-ene as sub-
ically decreased at that temperature. As indicated above, we
strate; it indicates that the enantioselectivity increases with
have found that a high water content in the solvent system does
increasing catalyst concentration, reaching a maximum at
not adversely affect the enantioselectivities and also speeds up
around 2 mol%.
The variation of ee in Fig. 1 suggests that the oxidation
pathway is in some way dependent upon the mediator concen-
the reaction; we therefore repeated the oxidation at Ϫ10 ЊC
using an increased proportion of water (3:1 water–aceto-
nitrile); a separate acetonitrile phase is seen at both 0 and
tration, giving low enantioselectivity at lower catalyst loadings.
Ϫ10 ЊC. Oxidation of 1-phenylcyclohex-1-ene mediated by the
However, as indicated above, blank reactions, carried out in the
presence of OxoneTM but without catalyst under the same con-
isopinocampheylamine-derived dihydroisoquinolinium tetra-
phenylborate salt (5 mol%) resulted, within 45 minutes, in com-
ditions gave no reaction. In the absence of any such achiral
oxidation process, this behaviour is consistent with a break-
down of catalyst–counter-ion ion pairs at low catalyst concen-
tration. Two diastereoisomeric adducts are possible upon
(probably reversible) reaction of the catalyst with persulfate;
the ratio of these diastereoisomeric adducts may vary with the
plete consumption of the starting material, and afforded the
corresponding epoxide in slightly improved yield. The enantio-
selectivity (ca. 35%) was very little reduced below that obtained
at 0 ЊC (40%).
In order to double check on the apparent insensitivity of the
reaction selectivity to temperature, the oxidation was also con-
degree of ion pairing in the catalyst. This ratio would be
ducted at a lower catalyst loading. Thus, with 0.5 mol% of the
reflected in the ratio of the diastereoisomeric oxaziridine salts
formed from the adducts by the irreversible loss of sulfate, pre-
sumed to be the rate determining step, and would affect the ee
same catalyst, the selectivities exhibited at 0 and Ϫ10 ЊC were
again essentially the same (26–28% ee). In the absence of the
cooling bath, the temperature of the reaction mixture reaches
observed as enantioselectivity in the (fast) epoxidation step
up to 32 ЊC, perhaps due to the exothermicity of the neutraliz-
then arises from competition between these diastereoisomeric
oxaziridine salts for the two enantiotopic faces of the alkene
ation of OxoneTM by the sodium carbonate, and ambient tem-
perature is then slowly attained. When the oxidation was
(Scheme 3). Doubling the alkene concentration from one to two
repeated over this temperature range (27–32 ЊC), negligible
molar while keeping the equivalence of base constant and
the catalyst loading fixed at 5 mol% had a negligible effect upon
the rate and enantioselectivity of the process, again support-
conversion to the epoxide after 1 hour was detected. This
unexpected failure of the catalysed process at moderately
increased temperature may be due to instability of the oxazirid-
ing the conclusion that the alkene is not involved in the rate-
inium species or of OxoneTM at higher temperatures.38 Indeed,
determining step.
in catalytic processes for alkene epoxidation where OxoneTM is
used as the stoicheiometric oxidant (e.g. mediated by oxaziridin-
Effect of substrate structure. The isopinocamphenylamine-
derived catalyst described here was tested with several aryl
alkenes 16–21 as substrates (Table 4), chosen to provide a range
of alkene substitution patterns. It is interesting that this system
ium salts or dioxiranes), reactions performed at above 0 ЊC are
rare.17–21 We have observed that OxoneTM gradually loses its
efficacy on storage, even when kept cool.
appears to offer a similar pattern of substrate structure–enantio-
Effect of catalyst loading. Catalyst loading was expected to
be an important parameter in the reaction system in relation to
conversion and perhaps also enantiomeric excess. The relation-
ship in Fig. 1 was obtained for the isopinocampheyl derivative
selectivity relationship to that of the previously described
iminium salt-catalysed epoxidation systems of Lusinchi and
Aggarwal; thus trans-stilbene is a poor substrate, and 1-phenyl-
cyclohex-1-ene the best, of those tested with our system.
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334
3329

View Article Online
Table 4 Epoxidation of unfunctionalized aryl alkenes with dihydro-
isoquinolinium salt derived from amine 9a
nm, at the temperatures indicated and are reported in units
of 10Ϫ1 deg cm2 gϪ1. Melting points were carried out on an
Electrothermal-IA 9100 apparatus and are uncorrected. Infra-
red absorption spectra were recorded on a Perkin-Elmer FT-IR
spectrometer Paragon 2001 instrument in the range of 4000–
600 cmϪ1. Mass spectra were recorded using Kratos MS-80 or
JEOL-SX102 instruments using the electron impact ionization
technique. Electrospray mass spectrometry was carried out
using a Hewlett Packard Autoplatform instrument or by the
EPSRC National Service. Proton nuclear magnetic resonance
spectra were recorded on Bruker AC 200, AC 250, AC 300, and
DPX 400 instruments, operating at 200.10, 250.13, 300.17, and
400.13 MHz respectively. Carbon-13 nuclear magnetic reson-
ance spectra were recorded on Bruker AC 250 and DPX 400
instruments, operating at 62.86 and 100.62 MHz respectively.
Enantiomeric excesses were determined either by 1H
NMR spectroscopy in the presence of tris[3-(heptafluoro-
propylhydroxymethylene)-(ϩ)-camphorato]europium(III), [(ϩ)-
Eu(hfc)3], as the chiral shift reagent, or by chiral HPLC analysis
using Chiracel OD and Chiracel OJ columns mounted on a
TSP Thermo-Separating-Products Spectra Series P200 instru-
ment. The solvent system used was hexane–propan-2-ol (90:10
or 95:5), operating at a flow rate of 0.50 mL per minute, (pump
pressure equivalent to ca. 80–135 psi), with the UV detector set
at 254 nm.
Entry
Alkene
Yield (%)
Ee (%)40
Configuration41
a
c
d
e
f
g
h
16
17
18
19
4
20
21
68
75
72
43
68
34
73
8
(ϩ)-(R)
1042
(ϩ)-(R,R)b
(ϩ)-(R,R)
(ϩ)-(S)
15
5
40
(ϩ)-(R,R)
(ϩ)-(1S,2R)
(Ϫ)c
3
20–6343
a Conditions: 5 mol% catalyst; 2 equiv. OxoneTM; 4 equiv. Na2CO3; 0 ЊC;
H2O–MeCN 1:1; reactions were monitored by TLC. Reactions were
judged to be complete when all alkene had been consumed (р45 min in
all cases). b 10 mol% catalyst. c Absolute configuration unknown.
2-(2-Bromoethyl)benzaldehyde33
Conclusion
Bromine (60 g, 0.37 mol) was added slowly over a period of
10 minutes to an ice cooled solution of isochromane (50 g,
0.37 mol) in carbon tetrachloride (200 mL) with stirring.
After the exothermic reaction subsided, the cooling bath was
removed and the dark brown solution heated under reflux
until the reaction mixture became pale yellow, and liberation
of HBr ceased (ca. 1.5 hours). The solution was then allowed
to reach ambient temperature and the solvent removed under
reduced pressure. To the yellow oil obtained (1-bromo-
isochromane), aqueous hydrobromic acid (48%, 75 mL) was
added and the reaction mixture heated under reflux. After
10–15 minutes the solution was allowed to cool and extracted
with diethyl ether (4 × 50 mL). The organic extracts were
washed with water (2 × 30 mL) and dilute aqueous sodium
bicarbonate, and dried over magnesium sulfate. Evaporation
of the solvent under reduced pressure furnished crude 2-(2-
bromoethyl)benzaldehyde as an orange oil approximately 85–
90% pure (67.5 g, 65% yield). Analytically pure samples may
be obtained by distillation under reduced pressure at ca.
150 ЊC and 0.5 mbar; chromatography is not recommended.
Both the crude and the distilled compound can be used in the
synthesis of dihydroisoquinolinium salts. νmax/cmϪ1 (neat)
2742, 1697, 1600, 1575, 1260, 1193, 755; δH (CDCl3, 400
MHz) 3.54–3.63 (4 H, m), 7.33 (1 H, d, J 8.0 Hz), 7.48 (1 H,
t, J 7.5 Hz), 7.54 (1 H, t, J 8.0 Hz), 7.80 (1 H, d, J 7.5 Hz),
10.14 (1 H, s); δC (62.50 MHz) 33.17, 36.70, 128.10, 132.51,
134.14, 134.33, 134.88, 140.95, 193.33; m/z 211.98370; calcd
for C9H9BrO: 211.98373.
In conclusion, we have examined a range of dihydroisoquin-
olinium salt epoxidation catalysts, effective at catalyst load-
ings as low as 0.5 mol%, providing up to ca. 40% ee in the
epoxidation of alkenes. Investigation of some of the reaction
parameters has thrown some light on the reaction mechanism,
including the conclusion that the enantiocontrolling step
involving transfer of oxygen to the substrate is not rate-
determining. This conclusion suggests that the prospects for
optimization of the process are good, since tuning of enantio-
selectivity may be possible without adverse effects upon the
overall rate.
Experimental
General
Light petroleum ether (bp 40–60 ЊC), was distilled from calcium
chloride prior to use. Ethyl acetate was distilled over calcium
sulfate or chloride. Dichloromethane was distilled over phos-
phorus pentaoxide or calcium hydride. Tetrahydrofuran was
distilled under a nitrogen atmosphere from the sodium–
benzophenone ketyl radical or from lithium aluminium
hydride. Triethylamine and diisopropylethylamine were stored
over potassium hydroxide pellets. Commercially available
reagents were used as supplied, without further purification,
unless otherwise stated. Air- and moisture-sensitive reactions
were carried out using glassware that had been dried overnight
in an oven at 240 ЊC. This was allowed to cool in a desiccator
over self-indicating silica gel pellets, under a nitrogen atmos-
phere. The reactions were carried out under a slight positive
static pressure of nitrogen. Flash chromatography was carried
out using Merck 9385 Kieselgel 60-45 (230–400 mesh). Thin
layer chromatography (TLC) was carried out on glass or
aluminium plates coated with a silica gel layer of 0.25 mm
thickness, containing fluorescer. Compounds were visualized
by UV irradiation at a wavelength of 254 nm, or stained by
exposure to an ethanolic solution of phosphomolybdic acid
(acidified with concentrated sulfuric acid), followed by charring
where appropriate.
General procedure for the synthesis of dihydroisoquinolinium
salts from 2-(2-bromoethyl)benzaldehyde and primary amines
A solution of the amine or amino alcohol in ethanol (10 mL
per g of amine) is added dropwise to 2-(2-bromoethyl)benz-
aldehyde (1.2 equiv.; 1.8 equiv. if crude material is used), exter-
nally cooled using an ice bath. The reaction mixture is stirred
for a few hours or overnight (depending on the amine used)
while attaining ambient temperature. A solution of sodium
tetraphenylborate (or any other anion exchanging salt, 1.10
equiv.) in the minimum amount of acetonitrile is added in one
portion, and after 5 minutes the organic solvents are evaporated
under reduced pressure. Ethanol is added to the residue, fol-
lowed by water. The resulting solid is collected by filtration and
Microanalyses were performed on Carlo Erba or Perkin-
Elmer instruments. Optical rotation values were measured with
an Optical Activity-polAAar 2001 instrument, operating at 589
3330
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334

View Article Online
washed with additional ethanol followed by diethyl ether. If no
solid materializes after the addition of water the suspension is
allowed to settle, the ethanol–water phase is decanted, and the
residue is macerated in hot ethanol. The organic salt may then
precipitate but in rare cases does so only upon slow cooling of
the hot alcoholic solution. If solubility problems arise, small
amounts of acetonitrile may be added during this process.
(1 H, m), 6.82 (4 H, t, J 7.2 Hz), 6.97 (8 H, t, J 7.7 Hz), 7.25–
7.28 (8 H, m), 7.44 (1 H, d, J 7.6 Hz), 7.51 (1 H, t, J 7.6 Hz),
7.75 (1 H, t, J 7.6 Hz), 7.85 (1 H, d, J 7.7 Hz), 8.67 (1 H, s);
δC (100 MHz) 13.03, 17.35, 18.63, 24.59, 26.32, 26.63, 31.24,
43.85, 50.11, 51.54, 51.63, 76.34, 121.44, 124.63, 125.31, 127.90,
128.05, 133.71, 135.38, 137.19, 137.87, 163.49, 164.71; m/z
268.20660; calcd for C19H26N: 268.20651.
(؊)-2-[(1S)-1-Phenylethyl]-3,4-dihydroisoquinolinium
tetraphenylborate
(؉)-2-[(1R,2R,4S)-1,3,3-Trimethylbicyclo[2.2.1]hept-2-yl]-3,4-
dihydroisoquinolinium tetraphenylborate
Amine 5 (1.00 g, 8.26 mmol) was treated with crude 2-(2-
bromoethyl)benzaldehyde (3.00 g, 14 mmol) according to the
general procedure to give the salt as a pale yellow solid (1.92 g,
42%), mp 168–169 ЊC. [α]2D0 Ϫ9.4 (c = 1.57, CH3CN); νmax/cmϪ1
(Nujol) 1647, 1605, 1572; δH (CD3CN, 400 MHz) 1.82 (3 H, d,
J 6.9 Hz), 2.99 (2 H, t, J 7.6 Hz), 3.65–3.71 (2 H, m), 5.24 (1 H,
q, J 6.8 Hz), 6.82 (4 H, t, J 7.2 Hz), 6.97 (8 H, t, J 7.4 Hz), 7.25–
7.29 (8 H, m), 7.37 (1 H, d, J 7.6 Hz), 7.45 (5 H, m), 7.54 (1 H, t,
J 7.6 Hz), 7.75 (1 H, t, J 7.6 Hz), 7.83 (1 H, d, J 7.6 Hz), 8.97
(1 H, s); δC (100 MHz) 17.61, 24.57, 46.96, 68.86, 121.53,
124.34, 125.30, 127.22, 127.96, 128.10, 129.05, 129.29, 130.83,
133.77, 135.37, 136.73, 138.01, 163.51, 164.57; m/z 236.14190;
calcd for C17H18N: 236.14392.
(ϩ)-endo-Fenchylamine 11 (1.00 g, 6.52 mmol) was treated with
crude 2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) accord-
ing to the general procedure to give the salt as a colourless solid
(2.33 g, 61%), mp 153–155 ЊC. [α]D20 ϩ12.5 (c = 1.50, CH3CN);
νmax/cmϪ1 (Nujol) 1649, 1605, 1577; δH (CD3CN, 400 MHz) 1.04
(3 H, s), 1.29 (6 H, s), 1.40–1.93 (7 H, m), 3.15 (2 H, t, J 8.0 Hz),
3.89–3.92 (2 H, m), 3.65 (1 H, s), 6.83 (4 H, t, J 7.2 Hz), 6.98
(8 H, t, J 7.5 Hz), 7.25–7.29 (8 H, m), 7.43 (1 H, d, J 7.6 Hz),
7.52 (1 H, t, J 7.6 Hz), 7.76 (1 H, t, J 7.6 Hz), 7.91 (1 H, d, J 7.6
Hz), 8.75 (1 H, s); δC (62.50 MHz) 18.48, 20.30, 24.34, 24.94,
25.52, 30.74, 40.56, 44.25, 48.07, 49.29, 51.25, 83.48, 121.50,
124.36, 126.43, 127.93, 128.10, 134.03, 135.87, 137.09, 137.98,
163.50, 164.94; m/z 268.20650; calcd for C19H26N: 268.20651.
(؉)-(؊)-2-[(1S)-1-Cyclohexylethyl]-3,4-dihydroisoquinolinium
tetraphenylborate
(؊)-2-[(1R,2R,3R,5S)-2,6,6-Trimethylbicyclo[3.1.1]hept-2-yl]-
3,4-dihydroisoquinolinium tetraphenylborate
Amine 6 (1.00 g, 7.86 mmol) was treated with crude 2-(2-
bromoethyl)benzaldehyde (3.00 g, 14 mmol) according to the
general procedure to give the salt as a colourless solid (2.55 g,
58%), mp 178–180 ЊC. [α]D20 ϩ22.5 (c = 1.36, CH3CN); νmax/cmϪ1
(Nujol) 1641, 1603, 1573; δH (CD3CN, 400 MHz) 1.00–1.04
(1 H, m), 1.18–1.27 (3 H, m), 1.44 (3 H, d, J 2.7 Hz), 1.60–1.79
(6 H, m), 1.92–1.94 (1 H, m), 3.16 (2 H, t, J 7.9 Hz), 3.83–3.87
(3 H, m), 6.83 (4 H, t, J 7.2 Hz), 6.98 (8 H, t, J 7.4 Hz), 7.25–
7.29 (8 H, m), 7.44 (1 H, d, J 7.6 Hz), 7.51 (1 H, t, J 7.6 Hz),
7.74–7.79 (2 H, m), 8.67 (1 H, s); δC (100 MHz) 14.71, 24.41,
24.96, 25.17, 25.24, 27.89, 29.33, 39.55, 45.24, 72.60, 121.48,
124.21, 125.33, 128.00, 128.12, 133.48, 135.44, 136.96, 137.93,
163.53, 165.15; m/z 242.19090; calcd for C17H24N: 242.19086.
Isopinocampheylamine 9 (1.00 g, 6.52 mmol) was treated with
crude 2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) accord-
ing to the general procedure to give the salt as a colourless solid
(2.67 g, 70%), mp 238–240 ЊC. [α]D20 Ϫ24.7 (c = 1.36, CH3CN);
νmax/cmϪ1 (Nujol) 1639, 1600, 1574, 735, 710; δH (CD3CN, 400
MHz) 1.12 (3 H, s), 1.18 (3 H, d, J 7.0 Hz), 1.33 (3 H, s), 1.95–
2.00 (3 H, m), 2.12–2.15 (2 H, m), 2.42–2.60 (2 H, m), 3.24 (2 H,
t, J 7.6 Hz), 4.08 (2 H, t, J 7.7 Hz), 4.54 (1 H, m), 6.86 (4 H, t,
J 7.2 Hz), 7.01 (8 H, t, J 7.4 Hz), 7.28–7.32 (8 H, m), 7.49 (1 H,
d, J 7.5 Hz), 7.56 (1 H, t, J 7.6 Hz), 7.80–7.84 (2 H, m), 9.09
(1 H, s); δC (100 MHz) 18.71, 22.15, 25.00, 26.95, 31.38, 33.07,
38.72, 39.66, 40.76, 44.62, 46.73, 89.18, 121.46, 124.71, 125.32,
127.93, 128.12, 133.42, 135.49, 136.96, 137.89, 163.55, 166.44;
m/z 268.20590; calcd for C19H26N: 268.20651.
(؉)-2-[(1R,2R,4R)-1,7,7-Trimethylbicyclo[2.2.1]hept-2-yl]-3,4-
dihydroisoquinolinium tetraphenylborate
(؊)-2-[(1R,2R,3R,5S)-2,6,6-Trimethylbicyclo[3.1.1]hept-2-yl]-
3,4-dihydroisoquinolinium periodate
exo-Bornylamine 13 (1.00 g, 6.52 mmol) was treated with crude
2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) according to
the general procedure to give the salt as a colourless solid (0.65
g, 17%), mp 196–197 ЊC. [α]D20 ϩ6.9 (c = 1.22, CH3CN); νmax/cmϪ1
(Nujol) 1649, 1601, 1572; δH (CD3CN, 400 MHz) 0.84 (3 H, s),
0.89 (3 H, s), 1.12 (3 H, s), 1.25–1.31 (1 H, m), 1.36–1.40 (1 H,
m), 1.67–1.72 (1 H, m), 1.85–1.91 (2 H, m), 1.97 (1 H, t, J 4.3
Hz), 2.39–2.48 (1 H, m), 3.06–3.14 (2 H, m), 3.81–3.87 (1 H, m),
3.89–3.95 (2 H, m), 6.82 (4 H, t, J 7.2 Hz), 6.98 (8 H, t, J 7.4
Hz), 7.25–7.29 (8 H, m), 7.41 (1 H, d, J 7.6 Hz), 7.50 (1 H, t,
J 7.6 Hz), 7.75 (1 H, t, J 7.6 Hz), 7.81 (1 H, d, 7.6 Hz), 8.81
(1 H, s); δC (100 MHz) 11.92, 19.09, 19.92, 24.52, 25.84, 33.19,
36.95, 44.39, 47.99, 50.92, 52.00, 77.91, 121.44, 124.64, 125.31,
127.89, 128.07, 133.71, 135.38, 135.45, 137.91, 163.57, 164.20;
m/z 268.20690; calcd for C19H26N: 268.20651.
Isopinocampheylamine 9 (1.00 g, 6.52 mmol) was treated with
crude 2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) accord-
ing to the general procedure to give the salt as a yellow solid
(2.00 g, 67%), mp 159–161 ЊC (dec.). [α]D20 Ϫ14.5 (c = 3.50,
CH3CN); νmax/cmϪ1 (Nujol) 1648, 1608, 1576, 844; δH, δC (CD3-
CN, 400 and 100 MHz respectively) identical with tetraphenyl-
borate derivative within 0.06 ppm (except for the tetraphenyl-
borate signals); m/z 268.20600; calcd for C19H26N: 268.20651.
(؊)-2-[(1R,2R,3R,5S)-2,6,6-Trimethylbicyclo[3.1.1]hept-2-yl]-
3,4-dihydroisoquinolinium hexafluorophosphate
Isopinocampheylamine 9 (1.00 g, 6.52 mmol) was treated with
crude 2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) accord-
ing to the general procedure to give the salt as a colourless solid
(1.56 g, 58%), mp 239–241 ЊC. [α]D20 Ϫ34.8 (c = 2.3, CH3CN);
νmax/cmϪ1 (Nujol) 1646, 1608, 1576, 877, 834; δH, δC (CD3CN,
400 and 100 MHz respectively) identical with tetraphenylborate
derivative within 0.09 ppm (except for the tetraphenylborate
signals); m/z 268.20610; calcd for C19H26N: 268.20651.
(؊)-2-[(1R,2S,4R)-1,7,7-Trimethylbicyclo[2.2.1]hept-2-yl]-3,4-
dihydroisoquinolinium tetraphenylborate
endo-Bornylamine 12 (1.00 g, 6.52 mmol) was treated with
crude 2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) accord-
ing to the general procedure to give the salt as a colourless solid
(1.00 g, 26%), mp 198–200 ЊC. [α]D20 Ϫ33.2 (c = 1.53, CH3CN);
νmax/cmϪ1 (Nujol) 1645, 1600, 1575; δH (CD3CN, 400 MHz) 0.96
(3 H, s), 0.98 (3 H, s), 1.00 (3 H, s), 1.22–1.28 (1 H, m), 1.47–
1.51 (2 H, m), 1.65–1.70 (1 H, m), 1.85–1.91 (2 H, m), 2.29–2.36
(1 H, m), 3.11 (2 H, t, J 7.8 Hz), 3.81–3.88 (2 H, m), 4.24–4.27
(؊)-2-[(1R,2R,3R,5S)-2,6,6-Trimethylbicyclo[3.1.1]hept-2-yl]-
3,4-dihydroisoquinolinium tetrafluoroborate
Isopinocampheylamine 9 (1.00 g, 6.52 mmol) was treated
with crude 2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol)
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334
3331

View Article Online
according to the general procedure to give the salt as a colour-
less solid (1.20 g, 52%), mp 186–188 ЊC. [α]D20 Ϫ33.6 (c = 2.5,
CH3CN); νmax/cmϪ1 (Nujol) 1645, 1606, 1577, 1062, 933, 776,
767; δH, δC (CD3CN, 400 and 100 MHz respectively) identical
with tetraphenylborate derivative within 0.07 ppm (except
for the tetraphenylborate signals); m/z 268.20600; calcd for
C19H26N: 268.20651.
125.36, 125.75, 127.72, 127.96, 128.07, 128.67, 133.65, 135.50,
136.24, 137.78, 139.07, 163.52, 167.40; m/z 291.18850; calcd for
C20H23N2: 291.18611.
(؊)-2-[(3R,10S,13S,14R,17S)-3-Hydroxy-10,13-dimethyl-
perhydrocyclopenta[a]phenanthren-17-yl]-3,4-dihydroiso-
quinolinium tetraphenylborate
Amine 14 (1.50 g, 5.15 mmol) was treated with 2-(2-bromo-
ethyl)benzaldehyde (2.00 g, 9.38 mmol) according to the general
procedure to give the salt as a colourless solid (0.41 g, 11%), mp
125–127 ЊC. [α]D20 Ϫ1.2 (c = 1.33, CH3CN); νmax/cmϪ1 (Nujol)
3420, 1644, 1605, 1578; δH (CD3CN, 400 MHz) 0.74 (3 H, s),
0.79 (3 H, s) 1.11–1.93 (22 H, m), 2.44 (1 H, m), 3.13 (2 H, m),
3.87–3.92 (4 H, m), 6.82 (4 H, t, J 7.2 Hz), 6.98 (8 H, t, J 7.4
Hz), 7.25–7.29 (8 H, m), 7.47 (1 H, d, J 7.3 Hz), 7.51 (1 H, t,
J 7.4 Hz), 7.76–7.79 (2 H, m), 8.78 (1 H, s); δC (100 MHz) 10.92,
11.24, 20.07, 22.75, 24.13, 24.87, 28.10, 28.55, 32.02, 32.64,
35.37, 35.59, 35.84, 36.68, 38.82, 46.15, 51.34, 52.88, 53.92,
65.27, 79.72, 121.68, 124.68, 125.46, 128.18, 128.34, 133.82,
135.64, 137.37, 138.10, 163.48, 164.43; m/z 406.31100; calcd for
C28H40NO: 406.31097.
(؊)-2-[(1R,2R,3R,5S)-2,6,6-Trimethylbicyclo[3.1.1]hept-2-yl]-
3,4-dihydroisoquinolinium perchlorate
Isopinocampheylamine 9 (1.00 g, 6.52 mmol) was treated with
crude 2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) accord-
ing to the general procedure to give the salt as a colourless solid
(1.47 g, 63%), [α]D20 Ϫ36.0 (c = 2.4, CH3CN); νmax/cmϪ1 (Nujol)
1645, 1606, 1576, 1086, 1051, 768, 623; δH, δC (CD3CN, 400 and
100 MHz respectively) identical with tetraphenylborate deriv-
ative within 0.05 ppm (except for the tetraphenylborate signals);
m/z 268.20590; calcd for C19H26N: 268.20651.
(؊)-2-[(1R,2S,5R)-5-Methyl-2-(1-methylethyl)cyclohexyl]-3,4-
dihydroisoquinolinium tetraphenylborate
Menthylamine 8 (1.00 g, 6.44 mmol) was treated with crude
2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) according to
the general procedure to give the salt as a pale yellow solid (1.44
General procedure for the catalytic asymmetric epoxidation of
simple alkenes mediated by iminium salts
g, 38%), mp 166–168 ЊC. [α]D20 Ϫ28.1 (c = 1.71, CH3CN); νmax
/
OxoneTM (2 equiv.), is added to an ice-cooled solution of
sodium carbonate (4 equiv.) in water (12 mL per 1.20 g of
sodium carbonate or as appropriate to adjust the MeCN–H2O
ratio) with vigorous stirring, and the resulting foaming suspen-
sion is stirred for ca. 5 minutes until most of the effervescence
subsides. The iminium salt (5–10 mol% with respect to the sub-
strate) is added as a solution in acetonitrile (7 mL per 100 mg of
catalyst), followed by the alkene substrate (1 equiv.), also as a
solution in acetonitrile to adjust the MeCN–H2O ratio (5 mL
per 400 mg of alkene). Solid substrates are added directly to the
reaction mixture after addition of further acetonitrile (5 mL per
400 mg of alkene). The suspension is stirred at 0 ЊC until the
substrate is completely consumed as indicated by TLC. The
reaction mixture is diluted with water until most of the
inorganic material dissolves, and extracted four times with
diethyl ether. The organic extracts are washed with water and
brine, and then dried over sodium sulfate. Filtration and evap-
oration of the solvents provides the crude epoxides. Chromato-
graphy on a short column of silica gel, eluting initially with
light petroleum to remove non polar impurities and/or
unreacted parent alkene, followed by light petroleum–ethyl
acetate (95:5), affords the purified epoxides.
cmϪ1 (Nujol) 1643, 1605, 1576; δH (CD3CN, 400 MHz) 0.79
(3 H, d, J 6.9 Hz), 0.94 (3 H, d, J 6.8 Hz), 0.96 (3 H, d, J 6.5 Hz),
1.21–1.25 (1 H, m), 1.45–1.49 (1 H, m), 1.51–1.68 (2 H, m),
1.70–1.82 (3 H, m), 1.84–2.00 (2 H, m), 3.13 (2 H, t, J 7.9 Hz),
3.83–3.89 (3 H, m), 6.81 (4 H, t, J 7.1 Hz), 6.98 (8 H, t, J 7.4
Hz), 7.25–7.28 (8 H, m), 7.42 (1 H, d, J 7.1 Hz), 7.50 (1 H,
t, J 7.3 Hz), 7.74–7.77 (2 H, m), 8.74 (1 H, s); δC (100 MHz)
14.30, 19.87, 20.81, 22.33, 24.56, 26.01, 31.23, 32.93, 72.97,
121.47, 124.47, 125.31, 127.99, 128.13, 133.50, 135.44, 137.09,
138.06, 163.53, 165.76; m/z 270.22220; calcd for C19H28N:
270.22216.
(؊)-2-{[(1S,2R,5S)-6,6-Dimethylbicyclo[3.1.1]hept-2-yl]-
methyl}-3,4-dihydroisoquinolinium tetraphenylborate
Myrtanylamine 10 (1.00 g, 6.52 mol) was treated with crude
2-(2-bromoethyl)benzaldehyde (2.5 g, 11.7 mmol) according
to the general procedure to give the salt as a pale green solid
(2.87 g, 75%), mp 173–174 ЊC. [α]D20 Ϫ10.7 (c = 1.80, CH3CN);
νmax/cmϪ1 (Nujol) 1651, 1604, 1574; δH (CD3CN, 400 MHz) 1.13
(3 H, s), 1.28 (3 H, s), 1.57–1.58 (1 H, m), 1.94–2.05 (6 H, m),
2.47–2.49 (1 H, m), 2.65–2.67 (1 H, m), 3.13 (2 H, t, J 7.9 Hz),
3.83–3.86 (4 H, m), 6.87 (4 H, t, J 7.2 Hz), 7.02 (8 H, t, J 7.4
Hz), 7.29–7.33 (8 H, m), 7.46 (1 H, d, J 7.3 Hz), 7.54 (1 H, t,
J 7.3 Hz), 7.77–7.80 (2 H, m), 8.68 (1 H, s); δC (100 MHz) 18.23,
22.05, 24.31, 24.95, 26.60, 32.25, 37.70, 37.91, 40.59, 42.89,
47.97, 65.62, 121.46, 124.14, 125.32, 128.03, 128.05, 133.30,
135.76, 136.49, 137.90, 163.54, 165.77; m/z 268.20650; calcd for
C19H26N: 268.20651.
1-Phenylcyclohexene oxide12,44–46
Treatment of 1-phenylcyclohex-1-ene 4 (0.50 g, 3.15 mmol)
according to the general procedure afforded the epoxide as a
colourless oil (0.36 g, 68%). νmax/cmϪ1 (neat) 3084, 1602, 1495,
1446, 1359, 1249, 1173, 1132, 1079, 1030, 993, 974; δH (CDCl3,
250 MHz) 1.22–1.35 (1 H, m), 1.53–1.64 (3 H, m), 1.99–2.06
(2 H, m), 2.16–2.18 (1 H, m), 2.26–2.32 (1 H, m), 3.10 (1 H, t,
J 2.0 Hz), 7.28–7.44 (5 H, m); δC (62.50 MHz) 19.78, 20.09,
24.69, 28.15, 60.13, 61.84, 125.26, 127.12, 128.20, 142.80.
(؉)-2-[(2S,3S)-2-Phenylhexahydropyridin-3-yl]-3,4-dihydro-
isoquinolinium tetraphenylborate
(2S,3S)-cis-3-Amino-2-phenylpiperidine 7 (1.00 g, 5.68 mmol)
was treated with crude 2-(2-bromoethyl)benzaldehyde (2.20 g,
10.3 mmol) according to the general procedure to give the salt
as a yellow solid (1.38 g, 40%), mp 172–174 ЊC (dec.). [α]D20
ϩ141.6 (c = 1.11, CH3CN); νmax/cmϪ1 (Nujol) 3320, 1634, 1603,
1576; δH (CD3CN, 400 MHz) 1.91–1.98 (1 H, m), 2.24–2.33
(4 H, m), 2.47 (1 H, br s), 2.66–2.70 (1 H, m), 2.88–2.97 (2 H,
m), 3.24 (1 H, m), 3.39 (1 H, m), 4.10 (1 H, m), 4.22 (1 H, m),
4.37 (1 H, m), 5.47 (1 H, br s), 6.88 (4 H, t, J 7.0 Hz), 7.02 (8 H,
t, J 7.2 Hz), 7.26–7.36 (12 H, m), 7.40–7.42 (2 H, m), 7.49 (1 H,
t, J 7.2 Hz), 7.72–7.79 (2 H, m), 9.86 (1 H, s); δC (100 MHz)
18.62, 24.40, 27.29, 45.71, 51.28, 61.52, 68.09, 121.51, 124.02,
1-Phenyl-3,4-dihydronaphthalene oxide42,47
Treatment of 1-phenyl-3,4-dihydronaphthalene 21 (1.00 g, 4.85
mmol) according to the general procedure afforded the epoxide
as a pale yellow solid (0.78 g, 72%), mp 104–106 ЊC (lit. 94–
97 ЊC). νmax/cmϪ1 (Nujol) 1602, 1486, 1307, 1155, 1074, 1042,
953; δH (CDCl3, 250 MHz) 2.10 (1 H, td, J 13.7, 5.8 Hz), 2.49–
2.60 (1 H, m), 2.77 (1 H, dd, J 15.5, 5.6 Hz), 2.98–3.06
(1 H, m), 3.71 (1 H, d, J 3.1 Hz), 7.11–7.31 (4 H, m), 7.45–7.61
(5 H, m); δC (62.50 MHz) 22.14, 25.43, 60.89, 62.98, 125.95,
127.68, 127.87, 128.07, 128.18, 128.58, 129.82, 134.99, 137.45,
138.82.
3332
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334

View Article Online
trans-Stilbene oxide28,42,48,49
6 C. Bolm, D. Kadereit and M. Valacchi, Synlett, 1997, 687.
7 Y. Naruta, F. Tani and K. Maruyama, Chem. Lett., 1989, 1269;
Y. Naruta, F. Tani, N. Ishihara and K. Murayama, J. Am. Chem.
Soc., 1991, 113, 6865; H. Yoon and C. J. Burrows, J. Am. Chem.
Soc., 1988, 110, 4087; Z. Gross and S. Ini, J. Org. Chem., 1997, 62,
5514; A. Berkessel and M. Frauenkron, J. Chem. Soc., Perkin Trans.
1, 1997, 2265; R. L. Halterman, S.-T. Jan, H. L. Nimmons, D. J.
Standlee and M. A. Khan, Tetrahedron, 1997, 53, 11257.
8 E. Cesarotti, A. Pasini and R. J. Ugo, J. Chem. Soc., Dalton Trans.,
1981, 2147; K. Nakajima, M. Kojima and J. Fujita, Chem. Lett.,
1986, 1483.
Treatment of trans-stilbene 17 (1.00 g, 5.33 mmol) according to
the general procedure afforded the epoxide as a colourless solid
(0.82 g, 76%), mp 66–67 ЊC (lit. 61–63 ЊC). νmax/cmϪ1 (Nujol)
1601, 1492, 1284, 1176, 1157, 1094, 1072, 1025; δH (CDCl3, 400
MHz) 3.84 (2 H, s), 7.28–7.37 (10 H, m); δC (100 MHz) 63.28,
125.98, 128.62, 129.31, 137.60.
trans-ꢀ-Methylstilbene oxide18
9 S. Chang, N. H. Lee and E. N. Jacobsen, J. Org. Chem., 1993, 58,
6939, and references therein.
Treatment of trans-α-methylstilbene 18 (1.00 g, 5.15 mmol)
according to the general procedure afforded the epoxide as a
colourless oil (0.64 g, 55%). νmax/cmϪ1 (neat) 3061, 1602, 1495,
1449, 1381, 1279, 1157, 1118, 1065, 1027, 980; δH (CDCl3, 400
MHz) 1.46 (3 H, s), 3.96 (1 H, s), 7.30–7.46 (10 H, m); δC (100
MHz) 17.14, 63.48, 67.52, 125.57, 126.92, 127.70, 127.93,
128.60, 129.21, 136.36, 142.75.
10 B. D. Brandes and E. N. Jacobsen, Tetrahedron Lett., 1995, 36,
5123.
11 H. Sakaki, R. Irie and T. Katsuki, Synlett, 1993, 300.
12 B. D. Brandes and E. N. Jacobsen, J. Org. Chem., 1994, 59, 4378.
13 S. Chang, J. M. Galvin and E. N. Jacobsen, J Am. Chem. Soc., 1994,
116, 6937.
14 (a) H. B. Henbest, Chem. Soc. Spec. Publ., 1965, 19, 83; R. C. Ewins,
H. B Henbest and M. A. McKervey, J. Chem. Soc., Chem.
Commun., 1967, 1085; W. H. Pirkle and P. L. Rinaldi, J. Org. Chem.,
1977, 42, 2080; F. J. Montanari, J. Chem. Soc., Chem. Commun.,
1969, 135; (b) R. M. Bowman, J. F. Collins and M. F. Grundon,
J. Chem. Soc., Perkin Trans. 1, 1973, 626; R. M. Bowman, J. F.
Collins and M. F. Grundon, J. Chem. Soc., Chem. Commun., 1967,
1131; J. D. Morrison and H. S. Mosher, Asymmetric Organic
Reactions, American Chemical Society, Washington, DC, 1971,
p. 336.
15 B. Hassine, M. Gorsane, F. Geerts-Evrard, J. Pecher, R. H. Martin
and D. Castelet, Bull. Soc. Chim. Belg., 1986, 95, 547; B. Hassine, M.
Gorsane, J. Pecher and R. H. Martin, Bull. Soc. Chim. Belg., 1986,
95, 557.
16 S. Juliá, J. Masana and J. C. Vega, Angew. Chem., Int. Ed. Engl.,
1980, 19, 929; S. Juliá, J. Guixer, J. Masana, J. Rocas, S. Colonna,
R. Annunziata and H. Molinari, J. Chem. Soc., Perkin Trans. 1, 1982,
1317; S. Colonna, H. Molinari, S. Banfi, S. Juliá, J. Masana and
A. Alverez, Tetrahedron, 1983, 39, 1635; S. Banfi, S. Colonna,
H. Molinari, S. Juliá and J. Guixer, Tetrahedron, 1984, 40, 5207;
B. C. B. Bezuidenhoudt, A. Swanepoel, J. A. N. Augustyn and
D. Ferreira, Tetrahedron Lett., 1987, 28, 4857; S. Itsuno,
M. Sakakura and K. Ito, J. Org. Chem., 1990, 55, 6047; P. W. Baures,
D. S. Eggleston, J. R. Flisak, K. J. Gombatz, I. Lantos, W.
Mendelson and J. J. Remich, Tetrahedron Lett., 1990, 31, 6501;
J. R. Fliasak, K. J. Gombatz, M. M. Holmes, A. A. Jarmas,
I. Lantos, W. G. Mendelson, V. J. Novack, J. J. Remich and
L. Snyder, J. Org. Chem., 1993, 58, 6247; P. A. Bentley, S. Bergeron,
M. W. Cappi, D. E. Hibbs, M. B. Hursthouse, T. C. Nugent,
R. Pulido, S. M. Roberts and L. E. Wu, Chem. Commun., 1997, 739.
17 S. E. Denmark and Z. Wu, J. Org. Chem., 1998, 63, 2810; S. E.
Denmark and Z. Wu, Synlett, 1999, 847; S. E. Denmark, D. C.
Forbes, D. S. Hays, J. S. De Pue and R. G. Wilde, J. Org. Chem.,
1995, 60, 1391.
Indene oxide50–52
Treatment of indene 20 (0.50 g, 4.31 mmol) according to the
general procedure afforded the epoxide as a colourless oil (0.30
g, 52%). νmax/cmϪ1 (neat) 3027, 2917, 1482, 1464, 1390, 1372,
1232, 1183, 1142, 829, 758, 745, 723; δH (CDCl3, 200 MHz) 2.97
(1 H, dd, J 18.1, 2.7 Hz), 3.21 (1 H, d, J 17.6 Hz), 4.13 (1 H, t,
J 3.0 Hz), 4.26 (1 H, dd, J 2.8, 1.1 Hz), 7.14–7.29 (3 H, m), 7.49
(1 H, dd, J 6.6, 1.7 Hz); δC (100 MHz) 34.62, 57.64, 59.09,
125.22, 126.12, 126.28, 128.59, 140.99, 143.64.
2,2,3-Triphenylethylene oxide21,43,53
Treatment of triphenylethylene 19 (1.00 g, 3.90 mmol) accord-
ing to the general procedure afforded the epoxide as a colour-
less oil which slowly solidified (0.44 g, 42%), mp 66–67 ЊC (lit.
75 ЊC). νmax/cmϪ1 (neat) 3062, 3030, 2957, 2925, 2856, 1605,
1596, 1499, 1471, 1448, 1262, 1221, 741, 698, 621; δH (CDCl3,
250 MHz) 4.40 (1 H, m), 7.10–7.47 (15 H, m); δC (62.50 MHz)
68.03, 68.34, 126.33, 126.75, 127.50, 127.55, 127.64, 127.70,
127.78, 127.83, 127.95, 128.20, 128.62, 135.42, 135.88, 141.12.
ꢀ-Methylstyrene oxide15,54
Treatment of α-methylstyrene 16 (1.00 g, 8.47 mmol) according
to the general procedure afforded the epoxide as a colourless oil
(0.72 g, 64%). νmax/cmϪ1 (neat) 3034, 2958, 2929, 2872, 1604,
1496, 1447, 1381, 1343, 1061, 1027, 860, 759, 699; δH (CDCl3,
250 MHz) 0.86 (3 H, d, J 6.6 Hz), 2.79 (1 H, dd, J 5.4, 0.8 Hz),
2.96 (1 H, d, J 5.4 Hz), 7.24–7.38 (5 H, m); δC (62.50 MHz)
21.74, 56.66, 56.93, 125.23, 127.37, 128.25, 129.00.
18 R. Curci, M. Fiorentino and M. R. Serio, J. Chem. Soc., Chem.
Commun., 1984, 155; D. S. Brown, B. A. Marples, P. Smith and
L. Walton, Tetrahedron, 1995, 51, 3587.
19 D. Yang, Y.-C. Yip, M.-W. Tang, M.-K. Wong, J.-H. Zheng and
K.-K. Cheung, J. Am. Chem. Soc., 1996, 118, 491; D. Yang, X. C.
Wang, M.-K. Wong, Y.-C. Yip and M.-W. Tang, J. Am. Chem. Soc.,
1996, 118, 11311; D. Yang, M. K. Wong, Y. P. Yip, X. C. Wang,
M. W. Tang, J. H. Zheng and K. K. Cheung, J. Am. Chem. Soc.,
1998, 120, 5943.
Acknowledgements
This investigation has enjoyed the support of Glaxo Wellcome
Research & Development and of Loughborough University.
20 A. Armstrong and B. R. Hayter, Chem. Commun., 1998, 621.
21 Z.-X. Wang, Y. Tu, M. Frohn and Y. Shi, J. Org. Chem., 1997,
62, 2328; Z.-X. Wang and Y. Shi, J. Org. Chem., 1997, 62, 8622;
Z.-X. Wang and Y. Shi, J. Org. Chem., 1998, 63, 3099; Y. Zhu, Y. Tu,
H. Yu and Y. Shi, Tetrahedron Lett., 1998, 39, 7819; G. A. Cao,
Z.-X. Wang, Y. Tu and Y. Shi, Tetrahedron Lett., 1998, 39, 4425;
M. Frohn, Z.-X. Wang and Y. Shi, J. Org. Chem., 1998, 63, 6425;
M. Frohn, M. Dalkiewicz, Y. Tu, Z.-X. Wang and Y. Shi, J. Org.
Chem., 1998, 63, 2948.
22 F. A. Davis, M. E. Harakal and S. B. Awad, J. Am. Chem. Soc., 1983,
105, 3123; F. A. Davis and M. S. Hague, J. Org. Chem., 1986, 51,
4083; F. A. Davis and S. Chattopadhyay, Tetrahedron Lett., 1986, 27,
5079; F. A. Davis, J. C. Towson, M. C. Weismiller, S. Lal and
P. J. Caroll, J. Am. Chem. Soc., 1988, 110, 8477; F. A. Davis
and S. G. Lal, J. Org. Chem., 1988, 53, 5004; F. A. Davis and
A. C. Sheppard, Tetrahedron, 1989, 45, 5703; F. A. Davis, R.
ThimmaReddy and M. C. Weismiller, J. Am. Chem. Soc., 1989, 111,
5964; J. C. Towson, M. C. Weismiller, S. G. Lal, A. C. Sheppard and
F. A. Davis, Org. Synth., 1990, 69, 158; F. A. Davis, R. Thimma-
Reddy, J. P. McCauley, Jr., R. M. Przeslawski, M. E. Harakal and
References
1 K. B. Sharpless, Aldrichimica Acta, 1983, 16, 67; J. Gorzynski Smith,
Synthesis, 1984, 629; C. H. Behrens and T. Katsuki, Coord. Chem.
Rev., 1995, 140, 189.
2 See for example: L. A. Paquette, Z. Gao, Z. Ni and G. F. Smith,
J. Am. Chem. Soc., 1998, 120, 2543; S. Amano, N. Ogawa,
M. Ohtsuka and N. Chida, Tetrahedron, 1999, 55, 2205.
3 K. Julienne and P. Metzner, J. Org. Chem., 1998, 63, 4532.
4 T. Katsuki and K. B. Sharpless, J. Am. Chem. Soc., 1980, 102, 5974;
R. M. Hanson and K. B. Sharpless, J. Org. Chem., 1986, 51, 1922;
Y. Gao, R. M. Hanson, J. M. Klunder, S. Y. Ko, H. Masamune and
K. B. Sharpless, J. Am. Chem. Soc., 1987, 109, 5765.
5 J. T. Groves, T. E. Nemo and R. S. Myers, J. Am. Chem. Soc., 1979,
101, 1032; J. T. Groves and R. S. Myers, J. Am. Chem. Soc., 1983,
105, 5791; Y. Naruta, F. Tani and K. Maruyama, Chem. Lett., 1989,
1269; Y. Naruta, F. Tani, N. Ishihara and K. Murayama, J. Am.
Chem. Soc., 1991, 113, 6865; K. Srinivasan, P. Michaud and
J. K. Kochi, J. Am. Chem. Soc., 1986, 108, 2309; E. G. Samsel,
K. Srinivasan and J. K. Kochi, J. Am. Chem. Soc., 1985, 107, 7606.
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334
3333

View Article Online
P. J. Carroll, J. Org. Chem., 1991, 56, 809; F. A. Davis, A. Kumar
and B.-C. Chen, J. Org. Chem., 1991, 56, 1143; F. A. Davis, M. C.
Weismiller, C. K. Murphy, R. ThimmaReddy and B.-C. Chen,
J. Org. Chem., 1992, 57, 7274; F. A. Davis, R. ThimmaReddy,
W. Han and P. J. Carroll, J. Am. Chem. Soc., 1992, 114, 1428;
F. A. Davis, R. ThimmaReddy, W. Han and R. E. Reddy, Pure Appl.
Chem., 1993, 65, 633; B.-C. Chen, C. K. Murphy, A. Kumar,
R. ThimmaReddy, C. Clark, P. Zhou, B. M. Lewis, D. Gala,
I. Mergelsberg, D. Scherer, J. Buckley, D. DiBenedetto and F. A.
Davis, Org. Synth., 1995, 73, 159; F. A. Davis, R. E. Reddy, P. V. N.
Kasu, P. S. Portonovo and P. J. Carroll, J. Org. Chem., 1997, 62, 3625.
23 P. C. B. Page, J. P. Heer, D. Bethell, E. W. Collington and D. M.
Andrews, Tetrahedron: Asymmetry, 1995, 6, 2911; P. C. B. Page,
J. P. Heer, D. Bethell, E. W. Collington and D. M. Andrews, Synlett,
1995, 773; J. Rebek, Jr. and R. McCready, J. Am. Chem. Soc., 1980,
102, 5602; K. Nanjo, K. Susuki and M. Sekiya, Chem. Lett., 1978,
1143; J. Rebek, Jr., S. Wolf and A. Mossman, J. Org. Chem., 1978,
43, 180; P. C. B. Page, J. P. Heer, D. Bethell, E. W. Collington and
D. M. Andrews, Tetrahedron Lett., 1994, 35, 9629.
34 R. D. Bach, J. L. Andres, M.-D. Su and J. J. McDouall, J. Am.
Chem. Soc., 1993, 115, 5768; K. N. Houk, J. Liu, N. C. DeMello and
K. R. Condroski, J. Am. Chem. Soc., 1997, 119, 10147.
35 K. Houk and Y. Washington, J. Am. Chem. Soc., 2000, 122, 2948.
36 W. Adam, R. T. Fell, C. R. Saha-Möller and C.-G. Zhao,
Tetrahedron: Asymmetry, 1998, 9, 397.
37 P. Ballinger and F. A. Long, J. Am. Chem. Soc., 1960, 82, 795.
38 J. O. Edwards, R. H. Pater, R. Curci and F. DiFuria, Photochem.
Photobiol., 1979, 30, 63; L. D. Ball and J. O. Edwards, J. Am. Chem.
Soc., 1956, 78, 1125.
39 Amines for entries c and j were kindly provided by Glaxo-Wellcome.
40 Enantiomeric excesses were determined by HPLC analysis using
Chiralcel OD or OJ columns [hexane–propan-2-ol (90:10 or 95:5)],
at a flow rate of 0.5 mL minϪ1 (equivalent to ca. 80–135 psi),
wavelength set at 254 nm. Phenyl cyclohexene and indene oxides are
separated better on the OJ column. All other epoxides gave better
separations on the OD column under the same conditions, or by
NMR spectroscopy with (Ϫ)-Eu(hfc)3 as the chiral shift reagent (3–8
mol% of (Ϫ)-Eu(hfc)3).
24 P. Milliet, A. Picot and X. Lusinchi, Tetrahedron Lett., 1976, 1573;
P. Milliet, A. Picot and X. Lusinchi, Tetrahedron Lett., 1976,
1577.
41 Absolute configuration of major enantiomer of product. Absolute
configurations were assigned by comparison of the sign of optical
rotation with literature values.
25 G. Hanquet, X. Lusinchi and P. Milliet, Tetrahedron Lett., 1987, 28,
42 This figure differs from that reported in our communication (see
ref. 31).
6061.
26 G. Hanquet, X. Lusinchi and P. Milliet, Tetrahedron Lett., 1988, 29,
3941.
27 G. Hanquet, X. Lusinchi and P. Milliet, C. R. Hebd. Seances Acad.
Sci., 1991, 313, S11, 625; X. Lusinchi and G. Hanquet, Tetrahedron,
1997, 53, 13727.
28 L. Bohé, G. Hanquet, M. Lusinchi and X. Lusinchi, Tetrahedron
Lett., 1993, 34, 7271; L. Bohé, M. Lusinchi and X. Lusinchi,
Tetrahedron, 1999, 55, 141.
43 In contrast to our earlier report (see ref. 31) the epoxide is in this
case formed in the reaction mixture with 20% ee; a fractional
crystallization occurs when using our standard work-up, resulting in
isolation of the epoxide with up to 63% ee.
44 Y. Tu, Z.-X. Wang and Y. Shi, J. Am. Chem. Soc., 1996, 118, 9806.
45 G. Belluci, J Chem. Soc., Perkin Trans. 2, 1973, 292.
46 G. Berti, B. Macchia, F. Macchia and L. Monti, J. Org. Chem., 1968,
33, 4045.
29 V. K. Aggarwal and M. F. Wang, Chem. Commun., 1996, 191.
30 A. Armstrong, G. Ahmed, I. Garnett and K. Gioacolou, Synlett,
1997, 1075; A. Armstrong and A. G. Draffan, Synlett, 1998, 646;
A. Armstrong, G. Ahmed, I. Garnett, K. Gioacolou and J. S. Wailes,
Tetrahedron, 1999, 55, 2341.
31 P. C. B. Page, G. A. Rassias, D. Bethell and M. B. Schilling, J. Org.
Chem., 1998, 63, 2774.
32 M. Yamato, K. Hashigaki, S. Ishikawa and N. Qais, Tetrahedron
Lett., 1988, 29, 6949; M. Yamato, K. Hashigaki, S. Ishikawa and
N. Qais, Tetrahedron, 1990, 46, 5909; W. Schneider and B. Müller,
Arch. Pharm. Chemi., 1961, 645.
47 A. Padwa and D. Owens, J. Org. Chem., 1977, 42, 3076.
48 A. Solladié-Cavallo, A. Diep-Vohuule, V. Sunjic and V. Vinkovic,
Tetrahedron: Asymmetry, 1996, 7, 1783.
49 H.-T. Chang and K. B. Sharpless, J. Org. Chem., 1996, 61, 6456.
50 H. Sasaki, R. Irie, T. Hamada, K. Suzuki and T. Katsuki,
Tetrahedron, 1994, 50, 11827.
51 D. R. Boyd, N. D. Sharma, N. I. Bowers, P. A. Goodrich and
M. R. Groocock, Tetrahedron: Asymmetry, 1996, 7, 1559.
52 L. Sola, A. Vidal-Ferran, A. Moyano, M. Pericas and A. Riera,
Tetrahedron: Asymmetry, 1997, 8, 1559.
53 R. Fleiser, D. Galle and M. Braun, Liebigs Ann., 1997, 1189.
54 S. S. Taj and R. Soman, Tetrahedron: Asymmetry, 1994, 5, 1513.
33 A. Rieche and E. Schmitz, Chem. Ber., 1956, 89, 1257.
3334
J. Chem. Soc., Perkin Trans. 1, 2000, 3325–3334