Aziridination of β-substituted styrene derivatives with 3-acetoxyaminoquinazolin-4(3H)-ones: Probing transition state geometry from changes in diastereoselectivity

Article abstract of DOI:10.1039/b107327n

Reaction of (S)-3-acetoxyamino-2-[1-(t-butyldimethylsilyloxy)ethyl]quinazolin-4(3H)-one 5 (Q¹NHOAc) with styrene and R(β)-substituted E-styrenes (R = SiMe₃, Me, CH₂Cl, CHCl₂) gives the corresponding aziridines d

Full text of DOI:10.1039/b107327n

View Article Online / Journal Homepage / Table of Contents for this issue
Aziridination of ꢀ-substituted styrene derivatives with
3-acetoxyaminoquinazolin-4(3H)-ones: probing transition
state geometry from changes in diastereoselectivity
2
Robert S. Atkinson,*a John Fawcett,a Ian S. T. Lochrie,a Sabri Ulukanlia
and Thomas A. Claxtonb
a Department of Chemistry, Leicester University, Leicester, UK LE1 7RH
b Department of Chemistry, University of York, York, UK YO10 5DD
Received (in Cambridge, UK) 13th August 2001, Accepted 1st February 2002
First published as an Advance Article on the web 12th March 2002
Reaction of (S)-3-acetoxyamino-2-[1-(t-butyldimethylsilyloxy)ethyl]quinazolin-4(3H )-one 5 (Q1NHOAc)
with styrene and R(β)-substituted E-styrenes (R = SiMe3, Me, CH2Cl, CHCl2) gives the corresponding
aziridines diastereoselectively. The diastereoselectivity increases in the same sense from 5 : 1
20 : 1 as the
electron-withdrawing character of R increases [H(Me), CH2Cl, CHCl2] but is accompanied by a decrease in
yield of aziridine. A similar increase in diastereoselectivity is found in the reaction of 3-acetoxyamino-
2-(2,3,3-trimethylpropyl)quinazolin-4(3H )-one 38 (Q4NHOAc) with the same β-substituted styrenes.
An explanation for these observations is offered based on a tighter, more symmetrical transition state
for the aziridination of styrenes bearing the more electron-withdrawing β-substituents and is supported by
SCF calculations.
(TS#s) for aziridination the planes containing alkene and the Q
ring are approximately parallel with an attractive (endo) overlap
Introduction
3-Acetoxyaminoquinazolinones 1 (QNHOAc) are ambiphilic
aziridinating agents for alkenes: they react in comparable yields
with alkenes of grossly different electron availability e.g. styrene
and α,β-unsaturated esters (Scheme 1).1 In the transition states
of the (Q)C᎐O with the π-system of the alkene substituent (2, 3).
᎐
However, the mechanism of reaction in TS# 2 is thought to
differ from that in TS# 3: in aziridination of styrene, TS# 2, Cβ–N
bond formation is believed to run ahead of N–Cα bond form-
ation2 whereas in aziridination of acrylates 3, N–Cβ bond form-
ation runs ahead of Cα–N bond formation i.e. the TS# takes
on some Michael addition character.3 In neither case is any
intermediate involved i.e. the reactions are concerted but asyn-
chronous. The configurations at the sp3-hybridised4 NHOAc
chiral centres in 2 and 3 are opposite because attack on this
nitrogen is assumed to be SN2 in character. Thus the ambiphilic-
ity of QNHOAc arises from its ability to respond to the differ-
ent electron availability of the alkene by a change in mechanism.
Aziridination with QNHOAc 1 is analogous to epoxidation
with peroxyacetic acid but QNHOAc 1 is the more versatile
reagent because peroxyacids do not react well with electron-
deficient alkenes, i.e. acrylates.
The location of R, the Q-2 substituent, in TS#s 2 and 3 means
that with R = R*, a chiral group, diastereoselective aziridin-
ation of prochiral alkenes is feasible.5a,b The work described
in this paper was initiated as a result of the observation that
aziridination of (E)-β-trimethylsilylstyrene with Q1NHOAc 5,
prepared in situ by N-acetoxylation of Q1NH2 4 with lead
tetraacetate (LTA) was more diastereoselective [dr (diastereo-
isomer ratio) 11 : 1]6 than styrene itself (dr 5 : 1) (Scheme 2).7
It seemed that the effect of the trimethylsilyl group on the
diastereoselectivity must be electronic in origin because in the
presumed TS# 8 for aziridination of this alkene, little, if any,
steric interaction of the trimethylsilyl group with the (Q1)-2-
substituent would be expected. To understand this electronic
effect and hence to extend its application to increase diastereo-
selectivity in aziridination more generally, we have studied the
aziridinations of a variety of β-substituted styrenes with
Q1NHOAc and the results are summarised in Scheme 3. The
only other homogeneous product isolated from these reactions
Scheme 1
DOI: 10.1039/b107327n
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828
This journal is © The Royal Society of Chemistry 2002
819

View Article Online
Examination of the crude reaction mixture shows that the
diastereoisomer ratio for aziridine 9 is 5 : 1. The enhancement
in dr from 5 : 1 13 : 1 above is important because it allows a
distinction between signals in the NMR spectrum from both
N-invertomers of the same diastereoisomer (a ratio which is not
affected by chromatography because interconversion is fast on
the real timescale in solution) and signals from different
diastereoisomers.
Aziridination of (E)-β-chloromethylstyrene using Q1NHOAc
5 and purification of the crude product 10 using a Chrom-
atotron gave the major diastereoisomer as a crystalline solid
(N-invertomer ratio 3.5 : 1 favouring Q and Ph cis) and the
minor diastereoisomer as an oil (N-invertomer ratio 2 : 1
favouring Q and Ph cis). Re-examination of the crude reaction
product showed that the initial diastereoisomer ratio was 10 : 1.
For the (E)-β-dichloromethylstyrene-derived aziridine 11,
measurement of the diastereoisomer ratio (20 : 1) and the azir-
idine N-invertomer ratio (7 : 1) in the major diastereoisomer
was again facilitated when the latter was obtained as a crystal-
line solid after silica chromatography. An NMR spectrum of
the mother liquor from this crystallisation after evapor-
ation showed it was enriched in the minor diastereoisomer of
aziridine 11.
Scheme 2 Reagents: i, Pb(OAc)4, CH2Cl2; ii, (E)-PhCH᎐CHSiMe3;
᎐
iii, styrene.
Aziridination of (E)-stilbene gave a 3 : 1 ratio of diastereo-
isomers of 12. A small sample of the minor diastereoisomer
was separated by column chromatography: the major diastereo-
isomer was obtained as a crystalline solid. Likewise, methyl
cinnamate gave a 3 : 1 ratio of diastereoisomers of aziridine 13
with the major one separable by chromatography. In neither of
these cases 12 or 13 was the absolute configuration at the two
chiral centres determined.
Stereostructures of aziridines 6, 7 and 9–11
The stereostructure of the major diastereoisomer of aziridine 6
was previously shown to be that having the 2S,3S configuration
by chemical correlation (Scheme 4) together with an X-ray crys-
Scheme 3
was the 3H-quinazolinone 14, a known decomposition product
of Q1NHOAc.
In the reaction of Q1NHOAc 5 with (E)-β-methylstyrene,
both diastereoisomers of aziridine 9 are present as mixtures of
N-invertomers by NMR spectroscopy in deuteriochloroform of
the crude reaction product. Unravelling of the diastereoisomer
ratio present was facilitated after separating the aziridine from
unchanged β-methylstyrene by chromatography and then by
partial separation of the diastereoisomers (13 : 1 ratio after
Kieselgel chromatography). The major diastereoisomer was
present as a 1.3 : 1 (Q/Ph cis : trans) ratio of N-invertomers and
the minor diastereoisomer as a 2.2 : 1 ratio favouring Q/Ph
trans. This conclusion was arrived at from observation of the
chemical shifts of the CHPh aziridine ring protons (Fig. 1):
protons cis to the quinazolinone ring have been previously
shown to be deshielded relative to those which are trans.8
Scheme 4 Reagents: i, KCN, CsF, CH3CN.
tal stucture determination6 of the minor diastereoisomer of
the related aziridine 15: both of the major diastereoisomers of
aziridines 6 and 15 gave the same product 16 by the aziridine–
azirine–aziridine sequence in Scheme 4.9
X-Ray crystal structures7 of the major diastereoisomers of
aziridines 10 and 11 revealed that these also had the same abso-
lute configuration at the 2-position as 6. Both crystal structures
showed the Q-group cis to the phenyl ring and this is presum-
ably the major N-invertomer in solution since the chemical
shifts of the PhCH aziridine ring protons in each case are at
higher field than those of the corresponding protons in the
minor N-invertomers (see above).
For determination of the absolute configuration of the major
diastereoisomers of styrene- and (E)-β-methylstyrene-derived
aziridines 7 and 9, both were converted to the corresponding
(Q)-2-(1-hydroxyethyl) compounds 17 and 18 by desilylation
using tetrabutylammonium fluoride (TBAF) (Scheme 5).
Fig. 1 Invertomer ratios for diastereoisomeric aziridines 9.
820
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828

View Article Online
In this work aziridination of (E)-β-methylstyrene with
Q2NHOAc 22 was also carried out in the presence of TTB and
aziridine 18 was isolated as a 5 : 1 ratio of diastereoisomers
(Scheme 7) with the major one identical to the crystalline major
Scheme 5
Aziridine 17 was obtained from aziridine 7 (dr 5 : 1) as a 5 : 1
ratio of diastereoisomers which were separated by Chrom-
atotron chromatography. Similarly, aziridine 9 (dr 5 : 1) was
converted into aziridine 18 (dr 5 : 1) without any apparent
interconversion of diastereoisomers: the major diastereoisomer
was separated as a crystalline solid.
Previously we had shown that 3-acetoxyamino-2-(1-hydroxy-
alkyl)quinazolinones 22–24, prepared in situ by acetoxylation
of the corresponding 3-aminoquinazolinones 19–21, react with
styrene diastereoselectively in the presence of titanium() tert-
butoxide (TTB) to give aziridines 17, 25 and 26 respectively
(Scheme 6).5a The major diastereoisomer of aziridine 17 pro-
Scheme 7 Reagents: i, (E)-RCH᎐CHPh2, Ti(OBut)4; ii, silica chrom-
᎐
atography.
diastereoisomer isolated by desilylation of aziridine 9 (Scheme
5). The diastereoisomer of 18 shown is that which would be
formed using our previous TS# model for aziridination of
styrene (see below). Aziridination of (E)-β-methylstyrene with
Q3NHOAc 24 in the presence of TTB gave aziridine 27 as a
single diastereoisomer from NMR examination of the crude
reaction mixture: some aziridine ring-opening to alcohol 29
took place on silica chromatography.
When aziridination of (E)-β-methylstyrene with Q3NHOAc
24 was carried out in the absence of TTB, the diastereoselectiv-
ity was, as expected, much reduced (dr 2 : 1). After separation
of the minor diastereoisomer by chromatography, it was shown
to be absent from the crude reaction product 27 by NMR
spectroscopy.
Although the identity in configuration at C-2 in aziridines 18
and 27 could not be confirmed by the 1H NMR correlation used
for aziridines 17 and 26 above, it seems very likely from the
correspondence in behaviour of β-methylstyrene and styrene in
the aziridinations in Schemes 5–7 that the β-methyl substituent
has virtually no effect upon the stereochemistry of these azirid-
inations and that the configuration at the aziridine ring chiral
centres in 9 is acccordingly 2S,3S.
The complete diastereoselectivity in aziridination of (E)-β-
methylstyrene with Q3NHOAc 24 in the presence of TTB to
give aziridine 27 deserves some comment when taken in con-
junction with our previous observation that aziridination of
α-methylstyrene under the same conditions gave a 5 : 1 ratio of
diastereoisomers.5a Our TS# model for aziridination of (E)-β-
methylstyrene that leads to diastereoisomer 27 has the phenyl
endo and cis to the Q-ring as in 30. However, the sense of dia-
stereoselectivity would remain unchanged even if some com-
petitive aziridination took place via TS# 31 since the same face
of the alkene is being attacked. Attack via TS# 31 seems less
likely than via 30 in any case because of steric interaction
between the N-acetoxy and tert-butyl groups. Attack via anal-
ogous TS#s 32 and 33 on α-methylstyrene, however, would lead
to diastereoisomeric aziridines because opposite faces of the
alkene would be involved: the N-acetoxy and tert-butyl groups
would be anti in either case (Scheme 8).
Scheme 6 Reagents: i, LTA, CH2Cl2, Ϫ20 ЊC; ii, Ti(OBut)4, styrene.
duced was identical to the major one obtained by desilylation
of aziridine 7 in Scheme 5.
Evidence for the same preferred sense of diastereoselectivity
in formation of aziridines 17, 25 and 26 came from the excellent
correlation in their NMR spectra for the chemical shifts of
aziridine ring and RCHOH protons in major and minor dia-
stereoisomers. The absolute configuration for aziridine 26 at the
3-membered ring carbon was shown to be (S) from X-ray crys-
tallography by relating it to the [(S)-tert-leucine-derived]
inducing centre in the Q2-substituent. From the correlations
above, it follows that the configuration at the phenyl-substituted
3-membered ring carbon in the major diastereoisomer of azirid-
ine 17 and hence aziridine 7 is also (S).5a
Aziridination of β-chloromethylstyrene with Q3NHOAc 24
was also completely diastereoselective giving 28 (Scheme 7).
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828
821

View Article Online
Scheme 8
Prior to obtaining confirmation of the absolute configur-
ation of aziridine 10 by X-ray crystallography, we had
attempted to do this by chemical correlation. Desilylation of
the major diastereoisomer by TBAF gave two products in a
1.6 : 1 ratio identified as the aziridine 34a and the imine 35
(Scheme 9a).
Fig. 2 The molecular structure of 40 showing the atom label scheme
and 30% displacement probability ellipsoids. H atoms are omitted for
clarity.
Previously we had prepared a diastereoisomer 34b of aziridine
34a by intramolecular aziridination of the 3-acetoxyamino-
quinazolinone 36 (Scheme 9b).4
Discussion
Aziridination with Q1NHOAc 5 of styrene and β-substituted
styrenes to give 6, 7, 9–11 all proceed in the same diastereosense
to give the same absolute configuration at the phenyl-bearing
aziridine C-2 position. As the β-substituent changes from
The NMR spectra of these two aziridines 34a and 34b were
consistent with their assigned diastereoisomeric relationship
particularly in the similarity of chemical shifts and coupling
constants for the OCH2CH unit but in the dissimilarities for the
CH3CH protons arising from their different environments.
Since both aziridines 34a and 34b have the same absolute con-
figuration for the chiral centre adjacent to the 2-position of the
Q-ring they must be epimeric at both aziridine ring chiral
centres (aziridination of double bonds by QNHOAc is always
stereospecific).
H(Me)
CH2Cl
CHCl2, i.e. becomes more electron-
withdrawing, there is an increase in diastereoselectivity but a
decrease in yield. Aziridination with Q4NHOAc 38 shows the
same trend of increased diastereoselectivity but reduced yield
with the same β-substituent changes if it is assumed that azir-
idination takes place in the same diastereosense in each case.
Although the changes in activation energy bringing about the
changes in diastereoselectivity in these aziridinations are small,
it is the trend of increasing dr with substituent electronegativity
which is significant. Our explanation for this trend is based on
the assumption that the effect of increasing the electron-
withdrawing character of the β-substituent reduces the extent
by which Cβ–N bond formation runs ahead of N–Cα in the TS#
for aziridination of these electron-available alkenes (2 in
Scheme 1).
Aziridination using Q4NHOAc 38
We also briefly examined aziridination of some of the β-substi-
tuted styrene derivatives used above with (racemic) Q4NHOAc
38 and the results are summarised in Scheme 10. The X-ray
crystal structure of the major diastereoisomer of aziridine 40
(Fig. 2) was obtained. The major diastereoisomer of aziridine
41 was obtained as a crystalline solid.
Scheme 9
822
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828

View Article Online
Table 1 Energy levels (hartrees) and calculated orbital coefficients of Cα and Cβ for some of the β-substituted styrenes used in Scheme 3
HOMO
Cα
LUMO
Cα
R
H
E
Cβ
E
Cβ
0.1794
0.1581
0.2579
0.2344
0.1786
0.2993
Ϫ0.2415
Ϫ0.4325
Ϫ0.3016
0.1119
0.2058
0.1823
0.2636
0.2421
0.1693
0.2910
Ϫ0.2569
Ϫ0.4225
Me
Ϫ0.2923
Ϫ0.3100
Ϫ0.3170
Ϫ0.3064
0.1152
0.1004
0.0940
0.0913
0.1625
0.1403
0.2177
0.1937
0.1823
0.3038
Ϫ0.2118
Ϫ0.3318
CH2Cl
CHCl2
SiH3
0.1518
0.1255
0.2158
0.1986
0.1943
0.3237
Ϫ0.2098
Ϫ0.3260
0.1727
0.1490
0.2587
0.2388
0.2238
0.3590
Ϫ0.2214
Ϫ0.3616
Fig. 3
In addition, as the electron-withdrawing character of the
β-substituent increases, there is in the LUMOalkene a progressive
increase in the magnitude of the Cα coefficient and a corre-
sponding decrease in the Cβ-coefficient that will increasingly
favour bond making at the Cα-position. The same effects, more-
over, are brought about by β-substitution of SiH3 in styrene
(as a model for SiMe3) (Table).
Overall, the effect of favouring increased N–Cα bond form-
ation in the TS# will be to make it more symmetrical and tighter
with formation of the two 3-membered ring bonds more but
not completely synchronous. An increase in diastereoselectivity
is expected, i.e. in formation of aziridine 11, because of
the increased site selectivity of the three substituents on the
inducing Q2-chiral centre. Thus using our previous TS# model
for this aziridination,2 42 will become favoured over 43 because
of the augmented interaction between Hβ and CHMeOSi in the
latter.
Scheme 10 Reagents: i, LTA, CH2Cl; ii, (E)-RCH᎐CHPh.
᎐
Using ab initio SCF calculations and a double zeta basis set
(6-319*), the orbital coefficients at Cα and Cβ and HOMO and
LUMO energy levels for some of the β-substituted styrenes
used in Scheme 3 were calculated (Table 1).
According to our mechanism for electron-available alk-
enes2 the aziridination involves overlap of HOMOalkene and
1
LUMOQ NHOAc (the N–OAc σ* orbital) and LUMOalkene
1
with HOMOQ NHOAc (the NHOAc lone pair) with the order-
ing of HOMO and LUMO levels resembling that in Fig. 3 for
styrene.
Since the aziridination is primarily electrophilic attack by
Q1NHOAc, we have chosen LUMO (Q1NHOAc) and HOMO
(alkene) for the reaction scheme, a choice illustrated by the
1
fact that LUMOQ NHOAc–HOMOalkene overlap is larger than
1
LUMOalkene–HOMOQ NHOAc overlap, in Fig. 3.
It is clear from the Table that the effect of increasing the
electron-withdrawing character of the β-substituent from H
CH2Cl
CHCl2 is to lower both the HOMO and LUMO
There is in TS# 42, a stereoelectronic effect which favours the
conformation around the Q2–C bond shown with the C–O
bond at the chiral centre almost in the plane of the Q ring and
levels of the alkene which increasingly favours LUMOalkene
HOMOQ NHOAc overlap.
–
1
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828
823

View Article Online
the O–Si bond orthogonal to this plane:10 the two methyl
groups on silicon then direct the methyl and hydrogen on the
chiral centre to the sites shown. It is this stereoelectronic effect
which is responsible for the superior levels of diastereoselectiv-
ity using Q1NHOAc 5 over Q4NHOAc 38 in spite of the better
size differentiation in the substituents on the inducing chiral
centre in Q4NHOAc 38.
The progressive decrease in yield with increasing electron-
withdrawal by the β-substituent in Schemes 3 and 10 is a reflec-
tion of the inherent electrophilic character of the QNHOAc
species: only in TS# 3 with back donation by the oxygen of the
α,β-unsaturated ester carbonyl is the nucleophilic reactivity of
the QNHOAc brought to the fore in Michael addition to the
double bond.
Aziridination of styrene with Q1NHOAc 5
General procedure 2 was followed using Q1NH2 49 (0.2 g,
0.63 mmol), LTA (0.31 g, 0.69 mmol), HMDS (0.20 g, 1.25
mmol) and styrene (0.13 g, 1.25 mmol) in dry dichlorometh-
ane (2 cm3). After work-up, examination of the residue by
NMR spectroscopy indicated two diastereoisomers of aziridine
7 were present in a 5 : 1 ratio from integral comparison of
signals at δ 2.71 and 2.65 ppm respectively and comparison with
previously reported spectra.4
Aziridination of ꢀ-methylstyrene with Q1NHOAc 5
General procedure 2 was followed using Q1NH2 49 (1 g,
3.13 mmol), LTA (1.53 g, 3.45 mmol), HMDS (1 g, 6.28 mmol)
and β-methylstyrene (0.74 g, 6.28 mmol) in dry dichlorometh-
ane (15 cm3). The crude product was chromatographed over
silica with light petroleum–ethyl acetate (10 : 1) as eluent to give
a mixture of aziridine 9 diastereoisomers (Rf 0.30) as a colour-
less oil (0.82 g, 60%). (Found: Mϩ 435.234. C25H33N3O2Si
requires M 435.234) νmax/cmϪ1 1675s and 1600s; m/z (%) 435
(Mϩ, 6), 247 (72), 132 (100) and 73 (30); δH (400 MHz)
(d5-pyridine) (mixture of N-invertomers) major N-invertomer
(Ph and Q1 cis) 0.00 (3H, s, CH3SiCH3), 0.06 (3H, s, CH3-
SiCH3), 0.68 (9H, s, But), 0.98 (3H, d, J 5.0, azir. CH3), 1.52
(3H, d, J 6.2, CH3CHO), 2.87 (1H, dq, J 6.0 and 5.0, azir. H-3),
3.51 (1H, d, J 6.0, azir. H-2), 5.18 (1H, q, J 6.2, CH3CHO),
6.99–7.67 [8H, m, 6-H, 7-H, 8-H(Q) and CHPh] and 8.20 [1H,
dd, J 8.0 and 1, 5-H(Q)]; minor N-invertomer (observable
signals) δ 1.01 (3H, d, J 5.0, azir. CH3), 1.64 (3H, d, J 6.2,
CH3CHO), 3.80 (1H, dq, J 5.6 and 5.0, azir. H-3), 4.00 (1H, d,
J 5.6, azir. H-2), 5.47 (1H, q, J 6.2, CH3CHO) and 8.28 [1H, d,
J 8.0, 5-H(Q)]. From integral comparison of signals at δ 5.18
and 5.47, the N-invertomer ratio was 1.3 : 1; minor diastereo-
isomer (mixture of N-invertomers) major N-invertomer (phenyl
and Q1 trans); (observable signals) δ 2.92 (1H, dq, J 5.0 and 6.0,
azir. H-3), 4.06 (1H, br d, J 5, azir. H-2) and 5.41 (1H, q, J 6.2,
CH3CHO); minor N-invertomer (observable signal) δ 3.47 (1H,
d, J 5.6, azir. H-2). From integral comparison of signals at
δ 4.06 and 3.47, the N-invertomer ratio was ∼2 : 1. The ratio
of diastereoisomers present was 5 : 1 from integral comparison
of signals at δ 5.18 ϩ 5.47 : 4.06 ϩ 3.47 in the crude reaction
product.
Experimental
For general experimental details, see ref. 11. Unless other-
wise indicated, NMR spectra were run at 250 MHz in deu-
teriochloroform solutions with tetramethylsilane as internal
standard. (E)-β-Dichloromethylstyrene was prepared by the
literature method12 and routinely crystallised from light petrol-
eum before use. All calculations were carried out using the
standard facilities provided by Gaussian 941.13
General procedure 1 for aziridination
The 3-aminoquinazolinone QNH2 (1 eq.) and acetic acid-free
lead tetraacetate (LTA) (1.1 eq.) were added alternately and
continuously in very small portions over 15 min. to a vigorously
stirred solution of dry dichloromethane (1 cm3 per 100 mg
of QNH2) cooled with a dry ice–acetone bath held at Ϫ20 to
Ϫ25 ЊC. After stirring for a further 5 min to give a solution of
the 3-acetoxyaminoquinazolinone, a solution of the alkene
(1.5–3.5 eq.) in dichloromethane (500 mg QNH2 per 1 cm3) was
added and then the temperature of the solution was allowed to
rise to ambient stirring throughout. Lead diacetate was separ-
ated and (on a 500 mg QNH2 scale) washed with dichloro-
methane (15 cm3) the organic solution washed successively with
saturated aqueous sodium hydrogen carbonate (15 cm3) and
water (15 cm3), dried with magnesium sulfate and the solvent
removed by evaporation under reduced pressure.
Kieselgel chromatography of the 5 : 1 ratio of diastereo-
isomers with light petroleum–ethyl acetate as eluent gave a
fraction containing the two diastereoisomers in a 13 : 1 ratio
from integral comparison of signals inter alia at δ 5.18 ϩ
5.47 : 4.06 ϩ 3.47.
General procedure 2 for aziridination in the presence of HMDS
A solution of 3-acetoxyaminoquinazolinone at Ϫ20 ЊC was
prepared as described above, separated from lead di-acetate at
this temperature (on a small scale a Pasteur pipette can be used)
and the cold solution added dropwise over 2 min. to a stirred
solution of the alkene (1.5–3.5 eq.) in dichloromethane (1 cm3/
500 mg) containing HMDS (2–3 eq.) maintained at Ϫ20 ЊC
(bath temperature). The temperature of the stirred solution was
then allowed to rise to ambient and the reaction worked-up as
described above.
Aziridination of (E )-cinnamyl chloride with Q1NHOAc 5
General procedure 2 was followed using Q1NH2 49 (1 g, 3.13
mmol), LTA (1.53 g, 3.45 mmol), HMDS (1.52 g, 9.40 mmol)
and cinnamyl chloride (1.43 g, 9.40 mmol) in dry dichloro-
methane (15 cm3). Chromatography of the product over silica
with light petroleum–ethyl acetate (15 : 1) as eluent gave the
major aziridine diastereoisomer 10 (Rf 0.29) as an oil which
solidified after trituration with light petroleum and crystallised
as a colourless solid (0.75 g, 51%) mp 119–121 ЊC (from
ethanol). [α]D = Ϫ378 (c = 0.5, CHCl3) (Found: C, 63.5; H, 6.9;
N, 8.9. C25H32ClN3O2Si requires C, 63.85; H, 6.85; N, 8.95%);
δH (400 MHz) (mixture of N-invertomers) major N-invertomer
(phenyl and Q1 cis): 0.09 (3H, s, CH3SiCH3), 0.11 (3H, s,
CH3SiCH3), 0.70 (9H, s, But), 0.88 (3H, d, J 6.1, CH3CHO),
3.76 (1H, dd, J 11.5 and 7.1, CHHCl), 3.78 (1H, d, J 3.9, azir.
H-2), 4.13 (1H, m, azir. H-3), 4.48 (1H, dd, J 11.5 and 3.9,
CHHCl), 5.05 (1H, q, J 6.1, CH3CHO), 6.80–7.62 [8H, m, 6-H,
7-H, 8-H(Q) and CH(Ph)] and 8.24 [1H, dd, J 8.5 and 0.6,
5-H(Q)]; minor N-invertomer (observable signals) δ 1.43 (3H, d,
J 6.4, CH3CHO), 3.24 (1H, m, azir. H-3), 3.61 (1H, dd, J 11.9
and 7.5 CHHCl), 4.54 (1H, d, J 4.6, azir. H-2), 5.43 (1H, q,
General procedure 3 for aziridination in the presence of
titanium(IV) tert-butoxide (TTB)
A cold (Ϫ20 ЊC) solution of 3-acetoxyaminoquinazolinone
freed from lead diacetate prepared as described above was
added dropwise over 2 min to a stirred solution of the alkene
(1.5–3.5 eq.) in dichloromethane (1 cm3 per 500 mg QNH2)
containing TTB (2 eq.) maintained at Ϫ20 ЊC (bath tem-
perature). After 5 min the cooling bath was removed and
the temperature of the stirred solution was allowed to rise to
ambient. The reaction mixture was then poured into saturated
hydrogen carbonate solution, the gelatinous precipitate pro-
duced filtered off (Celite), the organic layer separated, washed
with brine, dried and solvent evaporated under reduced
pressure.
824
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828

View Article Online
J 6.4, CH3CHO) and 8.18 [1H, d, J 7.9, 5-H(Q)]. From integral
comparison of the signals at δ 4.48 and 4.54, the ratio of major :
minor N-invertomers was 3.5 : 1; νmax/cmϪ1 2950m, 2860m,
1675s and 1600m; m/z (%) 469 (Mϩ, 1), 247 (68), 166 (44), 130
(51), 117 (32), 103 (48), 91 (30), 77 (55), 75 (55), 73 (100) and 57
(23). A crystal of aziridine 10 suitable for X-ray crystallography
was obtained by crystallisation from ethanol.
Repetition of reaction above and chromatography of the
crude product using a Chromatotron with light petroleum–
ethyl acetate (15 : 1) as eluent, gave a pure sample of minor
0 ЊC and then at room temp. for 15 min. The reaction mixture
was poured into ether (10 cm3), the solution washed with water
(10 cm3), brine (10 cm3), dried and then evaporated under
reduced pressure to give a yellow oil whose NMR spectrum
showed the presence of two aziridine 17 diastereoisomers in
a ratio 5 : 1 from integral comparison of signals at δ 2.75 and
2.65 ppm respectively. This mixture was separated using a
Chromatotron and eluting with light petroleum–ethyl acetate
(5 : 1) to give the major aziridine 17 diastereoisomer (Rf 0.18) as
a colourless oil (0.1 g, 69%). [α]D = 320 (c = 0.8, CHCl3) (Found:
Mϩ 307.132. C18H17N3O2 requires M 307.132); δH 1.51 (3H, d,
J 6.3, CH3CHOH), 2.75 (1H, dd, J 5.0 and 1.9, azir. H-3 trans
to Q1), 3.70 (1H, dd, J 7.9 and 1.9, azir. H-3 cis to Q1), 4.03 (1H,
dd, J 7.9 and 5.0, azir. H-2), 4.49 (1H, d, J 7.6, CH3CHOH ),
5.15 (1H, dq, J 7.6 and 6.3, CH3CHOH), 7.30–7.78 [8H, m,
6-H, 7-H, 8-H(Q) and CH(Ph)] and 8.22 [1H, dd, J 7.9 and 1,
5-H(Q)]; νmax/cmϪ1 1670s and 1600s; m/z (%) 307 (Mϩ, 31), 264
(40), 262 (64), 216 (63), 201 (27), 189 (60), 175 (20), 173 (100),
172 (81), 147 (26), 130 (27), 119 (33), 118 (74), 117 (66), 103
(61), 91 (77), 90 (42), 84 (20) and 76 (34).
aziridine 10 diastereoisomer (Rf 0.29) as a colourless oil. [α]D
=
54 (c = 1, CHCl3) (Found: Mϩ 469.195. C25H32ClN3O2Si
requires M 469.195); δH (mixture of N-invertomers) major
N-invertomer (phenyl and Q cis) Ϫ0.03 (3H, s, CH3SiCH3),
Ϫ0.01 (3H, s, CH3SiCH3), 0.83 (9H, s, But), 1.60 (3H, d, J 6.1,
CH3CHO), 3.69 (1H, d, J 3.9, azir. H-2), 3.77 (1H, dd, J 11.5
and 7.5, CHHCl), 4.33 (1H, dd, J 11.5 and 3.9, CHHCl), 4.40
(1H, m, azir. H-3), 4.95 (1H, q, J 6.1, CH3CHO), 6.99–7.66 [8H,
m, 6-H, 7-H, 8-H(Q) and CH(Ph)] and 8.22 [1H, d, J 8.5,
5-H(Q)]; minor N-invertomer (observable signals) δ 0.00 (3H, s,
CH3SiCH3), 0.04 (3H, s, CH3SiCH3), 0.85 (9H, s, But), 1.63
(3H, d, J 6.1, CH3CHO), 3.25 (1H, m, azir. H-3), 3.61 (1H, dd,
J 11.5 and 7.5, CHHCl), 4.65 (1H, d, J 3.9, azir. H-2), 5.38 (1H,
q, J 6.1, CH3CHO) and 8.19 [1H, d, J 8.5, 5-H(Q)]. From
integral comparison of the signals at δ 4.33 and 4.65, the ratio
of major : minor N-invertomers was 2 : 1; an NMR spectrum
(400 MHz) of the crude product above showed the ratio of
major : minor aziridine diastereoisomers of 10 was 10 : 1 from
integral comparison of signals at δ 4.48 ϩ 4.54 : 4.33 ϩ 4.65
ppm; m/z (%) 469 (Mϩ, 0.7), 412 (34), 247 (100), 166 (25), 130
(20), 129 (24), 117 (22) and 73 (27).
Further elution with light petroleum–ethyl acetate (5 : 1) gave
the minor aziridine 17 diastereoisomer also as a colourless oil
(0.03 g, 20%). [α]D = Ϫ94 (c = 0.5, CHCl3) (Found: Mϩ 307.132.
C18H17N3O2 requires M 307.132); δH 1.50 (3H, d, J 6.3,
CH3CHOH), 2.65 (1H, dd, J 5.0 and 1.9, azir. H-3 trans to Q),
3.49 (1H, dd, J 7.9 and 1.9, azir. H-3 cis to Q), 4.46 (1H, dd,
J 7.9 and 5.0, azir. H-2 and d, J 7.6, CH3CHOH ), 5.35 (1H, dq,
J 7.6 and 6.3, CH3CHOH), 7.38–7.78 [8H, m, 6-H, 7-H, 8-H(Q)
and CH(Ph)] and 8.21 [1H, dd, J 7.9 and 1, 5-H(Q)]; m/z (%)
307 (Mϩ, 22), 264 (25), 262 (38), 216 (77), 201 (29), 189 (39), 174
(21), 173 (73), 172 (58), 134 (72), 130 (21), 119 (56), 118 (49),
115 (36), 105 (55), 104 (26), 103 (56) 92 (71), 91 (100), 90 (31),
78 (37), 77 (42) and 76 (25).
Aziridination of (E )-ꢀ-dichloromethylstyrene with Q1NHOAc 5
General procedure 2 was followed using Q1NH2 49 (0.4 g,
1.25 mmol), LTA (0.61 g, 1.38 mmol), HMDS (0.61 g, 3.76
mmol) and β-dichloromethylstyrene (0.68 g, 3.76 mmol) in dry
dichloromethane (5 cm3). Flash chromatography of the product
over silica with light petroleum–ethyl acetate (15 : 1) as eluent
gave the major aziridine 11 diastereoisomer (Rf 0.29) as a
colourless solid (0.26 g, 41%) mp 130–132 ЊC (from ethanol),
[α]D = Ϫ366 (c = 1, CHCl3) (Found: C, 59.5; H, 6.25; N, 8.4.
C25H31Cl2N3O2Si requires C, 59.5; H, 6.2; N, 8.35%); δH (mix-
ture of N-invertomers) major N-invertomer (phenyl and Q cis)
0.00 (3H, s, CH3SiCH3), 0.02 (3H, s, CH3SiCH3), 0.70 (9H, s,
But), 0.83 (3H, d, J 6.3, CH3CHO), 3.96 (1H, d, J 5.4, azir.
H-2), 4.48 (1H, dd, J 5.4 and 3.8, azir. H-3), 4.98 (1H, q, J 6.3,
CH3CHO), 6.39 (1H, d, J 3.8, CHCl2), 6.90–7.70 [8H, m,
6-H, 7-H, 8-H(Q) and CH(Ph)] and 8.13 [1H, dd, J 7.8 and 1,
5-H(Q)]; νmax/cmϪ1 2950m, 3060m, 2860m, 1670s and 1600s;
m/z (%) 503 (Mϩ, 0.2), 446 (26), 247 (100), 164 (26), 130 (24), 75
(34) and 73 (65); minor N-invertomer (observable signals) δ 1.58
(3H, d, J 6.3, CH3CHO), 3.25 (1H, dd, J 5.4 and 3.8, azir. H-3),
5.63 (1H, q, J 6.3, CH3CHO), 5.81 (1H, d, J 3.8, azir. H-2)
and 6.53 (1H, d, J 5.4, CHCl2). From integral comparison
of the signals at δ 4.98 and 5.63, the ratio of major : minor
N-invertomers was 7 : 1; minor diastereoisomer (observable
signal) δ 3.86 (1H, d, J 5.4, azir. H-2). Examination of the
NMR spectrum of the crude product showed the ratio of
major : minor diastereoisomers of aziridine 11 was 20 : 1 from
comparison of the signals at δ 3.96 and 3.86 ppm. A crystal
of aziridine 11 suitable for X-ray crystallography was grown by
crystallisation from ethanol.
Aziridination of styrene with Q2NHOAc–TTB
General procedure 3 was followed using Q2NH2 195b (0.2 g,
0.98 mmol), LTA (0.48 g, 1.07 mmol) and added to a stirred
solution of TTB (0.66 g, 1.95 mmol) and styrene (0.20 g, 1.95
mmol) in dry dichloromethane (2 cm3) maintained at Ϫ20 ЊC
(bath temperature). After work-up, examination of the residue
by NMR spectroscopy showed the two aziridine 17 diastereo-
isomers (0.09 g, 31%) were present in a 6.5 : 1 ratio from
integral comparison of signals at δ 2.75 and 2.65 ppm respect-
ively in its NMR spectrum. These major and minor aziridine
diastereoisomers were shown to be identical to the major and
minor diastereoisomers isolated in the previous experiment by
comparison of their NMR spectra.
Identification of the major diastereoisomer of aziridine 9:
desilylation of aziridine 9
To a stirred solution of the aziridine 7 (dr 5 : 1) (0.25 g,
0.57 mmol) in THF (2 cm3) at 0 ЊC was added a solution of
TBAF in THF (1 M, 0.18 g, 0.68 mmol) and the mixture stirred
for 30 min at 0 ЊC and then at room temp. for 15 min. The
reaction mixture was poured into ether (10 cm3), the solution
washed with water (10 cm3), then brine (10 cm3), dried and
concentrated under reduced pressure to give a colourless oil.
Flash chromatography of the oil over silica with light
petroleum–ethyl acetate (10 : 1) as eluent gave a mixture of
aziridine 18 (Rf 0.34) as an oil (0.11 g, 59%) together with
unchanged aziridine 9 (0.09 g, 38%). Trituration of the oil with
light petroleum gave the major aziridine diastereoisomer 18 as
a colourless solid mp 127–128 ЊC (from ethanol). [α]D = Ϫ59 (c =
0.5, CHCl3) (Found: C, 71.0; H, 5.95; N, 13.1. C19H19N3O2
requires C, 71.0; H, 6.0; N, 13.15%); δH (mixture of N-
invertomers) major N-invertomer (phenyl and Q trans) 1.42
(3H, d, J 6.0, CH3CHOH), 1.45 (3H, d, J 6.6, azir. CH3), 3.08
(1H, dq, J 6.6 and 5.0, azir. H-3), 3.97 (1H, d, J 5.0, azir. H-2),
4.63 (1H, d, J 6.3, CH3CHOH ), 5.15 (1H, dq, J 6.3 and 6.0,
Identification of the major diastereoisomer of aziridine 7:
desilylation of aziridine 7
To a solution of the crude aziridine 7 prepared as described
above (0.20 g, 0.48 mmol) in THF (3 cm3) at 0 ЊC was added a
solution of tetrabutylammonium fluoride (TBAF) in THF
(1 M, 0.25 g, 0.96 mmol) and the mixture stirred for 30 min at
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828
825

View Article Online
CH3CHOH), 6.95–7.78 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)]
and 8.23 [1H, dd, J 7.9 and 1, 5-H(Q)]; minor N-invertomer
(observable signals) δ 1.31 (3H, d, J 6.6, azir. CH3), 1.70 (3H, d,
J 6.0, CH3CHOH), 3.60 (1H, d, J 6.0, CH3CHOH ), 4.28 (1H,
m, azir. H-3), 4.92 (1H, m, CH3CHOH) and 8.12 [1H, d, J 7.9,
5-H(Q)]. From integral comparison of signals at δ 4.63 and
3.60, the N-invertomer ratio of the major aziridine diastereo-
isomer was 2.7 : 1; νmax/cmϪ1 1690s and 1600s; m/z (%) 321 (Mϩ,
5), 132 (100) and 118 (20). Observable signals from the minor
diastereoisomer in the crude reaction product were present for
major N-invertomer (phenyl and Q trans) at δ 4.15 (1H, d, J 5.0,
azir. H-2) and for the minor N-invertomer at 3.50 (1H, d, J 5.0,
azir. H-2) respectively with the ratio of N-invertomers 1.9 : 1
from comparison of these two signals. The NMR spectrum
of the crude product showed the ratio of major : minor dia-
stereoisomers was 5 : 1 from integral comparison of signals at
δ 4.63 ϩ 3.60 : 4.15 ϩ 3.50 ppm.
of the crude product with light petroleum–ethyl acetate as
eluent (5 : 1) gave the minor diastereoisomer of aziridine 27 (Rf
0.42) as a colourless oil (0.07 g, 23%). (Found: Mϩ 363.194.
C22H25N3O2 requires M 363.194); δH major invertomer (phenyl
and Q trans): 1.05 (9H, s, But), 1.33 (3H, d, J 5.4, azir. CH3),
2.94 (1H, dq, J 5.4 and 5.4, azir. H-3), 3.80 (1H, d, J 10.4,
ButCHOH ), 3.88 (1H, d, J 5.4, azir. H-2), 4.95 (1H, d, J 10.4,
ButCHOH), 6.91–7.76 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)]
and 8.23 [1H, d, J 8.5, 5-H(Q)]; m/z (%) 363 (Mϩ, 0.7), 132
(100), 105 (29) and 83 (40); signals from the minor invertomer
were observable at δ 0.93 (9H, s, But), 1.68 (3H, d, J 5.4, azir.
CH3), 2.78 (1H, d, J 8.8, ButCHOH ), 3.55 (1H, d, J 6.0,
azir. H-2) and 4.63 (1H, d, J 8.8, ButCHOH). The ratio of
N-invertomers was 2.7 : 1 from comparison of signals at δ 4.95
and 4.63. Signals from this aziridine were absent from the spec-
trum of the crude product in the aziridination as above carried
out in the presence of TTB.
Further elution with light petroleum–ethyl acetate (5 : 1) gave
the major aziridine diastereoisomer of 27 (Rf 0.31) as a colour-
less solid (0.09 g, 31%) identical with that isolated previously.
Aziridination of ꢀ-methylstyrene with Q2NHOAc 22–TTB
General procedure 3 was followed using Q2NH2 1914 (0.4 g, 1.95
mmol), LTA (0.95 g, 2.15 mmol), TTB (1.33 g, 3.9 mmol) and
β-methylstyrene (0.46 g, 3.9 mmol) in dry dichloromethane
(6 cm3). An NMR spectrum of the crude product showed two
aziridine diastereoisomers present in a 5 : 1 ratio which were
identical to the major and minor aziridines 18 respectively in
the crude reaction product from the previous experiment.
Aziridination of (E )-cinnamyl chloride with Q3NHOAc 24–TTB
General procedure 3 was followed using Q3NH2 215a (0.3 g,
1.21 mmol), LTA (0.59 g, 1.33 mmol), TTB (0.82 g, 2.42 mmol)
and cinnamyl chloride (0.37 g, 2.42 mmol) in dry dichloro-
methane (5 cm3). Crystallisation of the product by addition of
ethanol gave aziridine 28 (Rf 0.29) as a colourless solid (0.22 g,
46%) mp 149–151 ЊC (from ethanol). [α]D = ϩ190 (c =
0.5, CH2Cl2) (Found: C, 66.15; H, 6.1; N, 10.35. C25H25ClN3O2
requires C, 66.4; H, 6.1; N, 10.55%); δH (mixture of N-invert-
omers) major N-invertomer (phenyl and Q cis): 0.75 (9H, s, But),
3.29 (1H, dd, J 11.3 and 9.4, CHHCl), 3.56 (1H, m, azir. H-3),
3.57 (1H, J 10.4, ButCHOH ), 3.66 (1H, d, J 5.0, azir. H-2), 4.30
(1H, dd, J 11.3 and 4.7, CHHCl), 4.81 (1H, d, J 10.4,
ButCHOH ), 7.0–7.80 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)]
and 8.25 [1H, dd, J 7.9 and 1, 5-H(Q)]; νmax/cmϪ1 1660s and
1590s; m/z (%) 398 (Mϩ, 1.1), 348 (73), 312 (61), 214 (22), 189
(33), 176 (24), 175 (61), 168 (24), 166 (67), 147 (20), 132 (28),
130 (100), 117 (65), 116 (30), 115 (20), 104 (44), 103 (83), 77
(32), 76 (25) and 57 (33); minor N-invertomer (observable
signals): δ 1.00 (9H, s, But), 3.98 (1H, d, J 5.0, azir. H-2), 4.50
(1H, s, br, ButCHOH ) and 8.02 [1H, dd, J 7.9 and 1, 5-H(Q)];
from integration comparison of the signals at δ 8.25 : 8.02 the
ratio of N-invertomers was 4 : 1.
Aziridination of ꢀ-methylstyrene with Q3NHOAc 24–TTB
General procedure 3 was followed using Q3NH2 215a (0.2 g,
0.81 mmol), LTA (0.4 g, 0.90 mmol), TTB (0.55 g, 1.62 mmol)
and β-methylstyrene (0.19 g, 1.62 mmol) in dry dichloro-
methane (3 cm3). An NMR spectrum of the crude product
indicated only one aziridine diastereoisomer was present and,
after trituration with ethanol, some of this aziridine crystallised
out; a further quantity was obtained by Chromatotron chrom-
atography with light petroleum–ethyl acetate as eluent (5 : 1) to
give aziridine 27 (Rf 0.27) as a colourless solid (total 0.14 g,
49%) mp 157–158 ЊC (from ethanol). [α]D = 60 (c = 0.5, CHCl3)
(Found: C, 72.35; H, 6.95; N, 11.4. C22H25N3O2 requires C, 72.7;
H, 6.95; N, 11.55%); δH (mixture of N-invertomers) major
N-invertomer (phenyl and Q trans): 0.68 (9H, s, But), 1.32 (3H,
d, J 5.4, azir. CH3), 3.21 (2H, m, azir. H-3 and H-2), 3.64 (1H,
d, J 10.4, ButCHOH ), 4.70 (1H, d, J 10.4, ButCHOH), 7.26–
7.70 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)] and 8.20 [1H, dd,
J 8.5 and 1.3, 5-H(Q)]; νmax/cmϪ1 1765w, 1680s and 1590s; m/z
(%) 364 (MHϩ, 63), 307 (20), 154 (100), 137 (52), 136 (71) and
132 (40); signals from the minor invertomer were observable at
δ 0.93 (9H, s, But), 1.65 (3H, d, J 5.4, azir. CH3). The ratio of
N-invertomers was 3.6 : 1 from comparison of signals at δ 0.68
and 0.93.
Further elution with light petroleum–ethyl acetate (5 : 1)
(Chromatotron) gave an oily ring-opened alcohol 29 (0.035 g,
11%). (Found: Mϩ 382.213. C22H28N3O3 requires M 382.213);
δH 0.86 (3H, d, J 6.6, NHCHCH3), 1.03 (9H, s, But), 2.96 (1H,
d, J 6.6, CHOHPh), 3.62 (1H, d, J 10.1, ButCHOH ), 3.71
(1H, m, CHCH3), 4.60 (1H, dd, J 6.6 and 6.0, CHOHPh), 5.28
(1H, d, J 10.1, ButCHOH), 6.16 (1H, d, J 2.5, NH ), 7.33–7.82
[8H, m, 6-H, 7-H, 8-H (Q) and CH(Ph)] and 8.28 [1H, dd, J 7.9
and 1.6, 5-H(Q)]; νmax/cmϪ1 3500m, 1680s, 1610w, 1595s and
1565s; m/z (%) 382 (Mϩ, 0.4), 274 (100), 256 (56), 215 (68), 176
(29), 175 (31) and 147 (22).
Desilylation of aziridine 10
To a stirred solution of aziridine 10 (0.09 g, 0.19 mmol) in THF
(1 cm3) at 0 ЊC was added a solution of TBAF in THF (1 M,
0.08 g, 0.29 mmol) and the mixture stirred for 30 min at 0 ЊC
and then at room temp. for 15 min. The reaction mixture was
poured into ether (5 cm3), the solution washed with water
(5 cm3), then brine (5 cm3), dried and concentrated under
reduced pressure to give a colourless oil whose NMR spectrum
showed the presence of aziridine ether 34a and imine 35 in a
1.6 : 1 ratio from integration comparison of signals at δ 8.11
and 8.37 ppm respectively (see below). Flash chromatography
of the product over silica with light petroleum–ethyl acetate
(2 : 1) as eluent gave an oil (Rf 0.53) (0.03 g, 54%). Trituration of
this oil with light petroleum gave the cyclic aziridine ether 34a as
a colourless solid, mp 171–173 ЊC (from ethanol). [α]D = Ϫ104
(c = 0.8, CHCl3) (Found: C, 71.2; H, 5.5; N, 13.1. C19H17N3O2
requires C, 71.45; H, 5.35; N, 13.15%); δH 1.84 (3H, d, J 7.2,
CH3CHO), 3.11 (1H, ddd, J 10.7, 5.3 and 3.5, azir. H-3), 3.27
(1H, d, J 5.3, azir. H-2), 3.58 (1H, dd, J 13.5 and 10.7, CHH),
4.42 (1H, dd, J 13.5 and 3.5, CHH ), 5.10 (1H, q, J 7.2,
CH3CHO), 7.22–7.62 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)]
and 8.11 [1H, ddd, J 8.2, 1.6 and 0.6, 5-H(Q)]; νmax/cmϪ1 1670s
and 1590s; m/z (%) 319 (Mϩ, 23), 216 (89), 215 (100), 214 (27),
188 (21), 187 (27), 173 (93), 156 (20), 130 (21), 117 (60), 115
The above experiment was repeated, but in the absence
of TTB. General procedure 1 was followed using Q3NH2 215a
(0.2 g, 0.81 mmol), LTA (0.4 g, 0.90 mmol) and β-methylstyrene
(0.19 g, 1.62 mmol) in dry dichloromethane (3 cm3). Examin-
ation of the NMR spectrum of the crude product showed two
aziridine diastereoisomers were present in a 1.6 : 1 ratio from
integration of signals at δ 4.70 : 4.95 ϩ 4.63 with the major one
identical to that isolated above. Chromatotron chromatography
826
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828

View Article Online
(34), 104 (20), 103 (21), 102 (25), 90 (20), 84 (25) and 77 (24).
For comparison, signals of the other diastereoisomer 34b are
found at δ 1.75 (3H, d, J 7.0, CH3CHO), 3.25 (1H, ddd, J 13.1,
5.3 and 3.6, azir. H-3), 3.4 (1H, d, J 5.3, azir. H-2), 3.6 (1H, dd,
J 13.5 and ca. 13, CHH), 4.5 (1H, dd, J 13.1 and 3.6, CHH ), 5.6
(1H, q, J ca. 7, CH3CHO), 7.4–7.75 [8H, m, 6-H, 7-H, 8-H(Q)
and CH(Ph)] and 8.25 [1H, ddd, J 7.7 and ca. 1, 5-H(Q)].4
Further elution of the column using the same solvent mix-
ture gave imine 35 as a colourless oil (Rf 0.37) (0.015 g, 23%).
(Found: Mϩ 319.132. C19H17N3O2 requires M 319.132); δH: 1.67
(3H, d, J 6.6, CH3CHO), 2.85 (2H, m, OCH2), 3.05 (2H, m,
CH2CN), 5.06 (1H, q, J 6.6, CH3CHO), 7.22–7.77 [8H, m, 6-H,
7-H, 8-H(Q) and CH(Ph)] and 8.37 [1H, dd, J 7.9 and 1.3, 5-
H(Q)]; δC 17.12 (CH3), 30.9 (CH2CH2), 33.4 (CH2CH2), 64.1
(CH3CHO), 121.3 [CCO(Q)], 125.9–133.3 [4 × CH(Q) and 5 ×
NMR spectrum of the crude product showed the ratio of major
: minor diastereoisomers was 3 : 1 from integration of signals at
δ 4.00 and 3.96 ppm.
Aziridination of (E )-cinnamyl chloride with Q4NHOAc 38
General procedure 2 was followed using Q4NH2 375b (0.2 g,
0.82 mmol), LTA (0.4 g, 0.90 mmol), HMDS (0.26 g, 1.63
mmol) and cinnamyl chloride (0.25 g, 1.63 mmol) in dry
dichloromethane (3 cm3). Flash chromatography of the product
over silica with light petroleum–ethyl acetate (12 : 1) as eluent
gave a mixture of the aziridine 40 as a mixture of diastereo-
isomers (Rf 0.53) as a colourless oil (0.21 g, 64%) which crystal-
lised after standing in the freezer (4–5 days); the major aziridine
40 diastereoisomer was obtained on addition of ethanol as a
colourless solid, mp 94–96 ЊC (from ethanol). (Found: C, 69.6;
H, 6.75; N, 10.45. C25H26ClN3O requires C, 69.75; H, 6.6;
N, 10.6%); δH 0.36 (3H, d, CH3CHBut), 0.88 (9H, s, But), 3.02
(1H, q, J 6.9, CH3CHBut), 3.73 (1H, dd, J 11.3 and 3.7,
CHHCl), 3.78 (1H, d, J 5.6, azir. H-2), 3.96 (1H, ddd, J 7.5, 5.6
and 3.7, azir. H-3), 4.52 (1H, dd, J 11.3 and 7.5, CHHCl), 6.85–
7.62 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)] and 8.25 [1H, dd,
J 7.9 and 1.3, 5-H(Q)]; νmax/cmϪ1 1665s and 1590s; m/z (%) 395
(Mϩ, 9), 347 (25), 346 (100), 339 (21), 175 (42), 174 (46), 173
(23), 166 (41), 131 (34), 130 (66), 121 (26), 119 (40), 117 (62),
116 (28), 103 (29), 82 (26), 77 (26), 57 (33) and 51 (37). Observ-
able signals from the minor diastereoisomer of aziridine 40 in
the NMR spectrum of the crude reaction product were present
at δ 0.73 (9H, s, But), 1.03 (3H, d, J 6.9 CH3CHBut), 4.33 (1H,
dd, J 11.3 and 3.7, CHHCl) and 8.21 [1H, d, J 7.9, 5-H(Q)]. The
NMR spectrum of the crude product showed the ratio of major
: minor diastereoisomers present was 2 : 1 from integration of
signals at δ 4.52 and 4.33 ppm. An X-ray structure determin-
ation was carried out for the major diastereoisomer of aziridine
40 on a crystal obtained from ethanol (see Fig. 2).
CH(Ph)], 138.8 [C(Ph)], 144.8 [C᎐N(Q)], 145.2, 155.6, 157.4
᎐
[C–N(Q), N᎐C(Ph) and CO(Q)]; m/z (%) 319 (Mϩ, 68), 215 (26),
᎐
186 (20), 173 (27), 131 (41), 104 (43), 85 (91), 77 (44), 69 (41)
and 57 (100).
Aziridination of stilbene with Q1NHOAc 5
General procedure 2 was followed using Q1NH2 49 (0.2 g, 0.63
mmol), LTA (0.31 g, 0.69 mmol), HMDS (0.25 g, 1.54 mmol)
and styrene (0.14 g, 0.75 mmol) in dry dichloromethane (3 cm3).
Flash chromatography of the crude product over silica with
light petroleum–ethyl acetate (5 : 1) as eluent gave a pure sample
of the minor aziridine 12 diastereoisomer (Rf 0.56) as a colour-
less oil (0.007 g, 2%). δH Ϫ0.20 (3H, s, CH3SiCH3), 0.00 (3H, s,
CH3SiCH3), 0.95 (9H, s, But), 1.84 (3H, d, J 6.3, CH3CHO),
3.95 (1H, d, J 6, azir. H-2 trans to Q), 4.88 (1H, d, J 6, azir. H-3
cis to Q), 4.95 (1H, q, J 6.3, CH3CHO), 7.20–7.84 [13H, m, 6-H,
7-H, 8-H(Q) and CH(Ph)] and 8.16 [1H, d, J 7.8, 5-H(Q)].
Further elution with the same solvent mixture gave major
aziridine 12 diastereoisomer (Rf 0.48) as an oil which, on tritur-
ation with light petroleum, gave a colourless crystalline product
(0.053 g, 34%). [α]D = Ϫ599 (c = 0.6, CHCl3) (Found: Mϩ
497.250. C30H35N3O2Si requires M 497.250); δH 0.00 (3H, s,
CH3SiCH3), 0.02 (3H, s, CH3SiCH3), 0.69 (9H, s, But), 0.81
(3H, d, J 6.3, CH3CHO), 3.69 (1H, d, J 6.0, azir. H-3 trans to
Q), 4.61 (1H, d, J 6.0, azir. H-2 cis to Q), 5.07 (1H, q, J 6.3,
CH3CHO), 6.86–7.55 [13H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)]
and 8.14 [1H, d, J 7.8, 5-H(Q)]. From the NMR spectrum of
the crude product the ratio of major : minor diastereoisomers
was 3 : 1 from integration of signals at δ 5.07 and 4.95 ppm.
νmax/cmϪ1 1660s and 1600s; m/z (%) 497 (Mϩ, 0.7), 247 (47), 195
(15) and 194 (100).
Aziridination of (E )-ꢀ-dichloromethylstyrene with Q4NHOAc 38
General procedure 2 was followed using Q4NH2 375b (0.2 g,
0.82 mmol), LTA (0.4 g, 0.90 mmol), HMDS (0.26 g, 1.63
mmol) and β-dichloromethylstyrene (0.18 g, 0.98 mmol) in dry
dichloromethane (3 cm3). Trituration of the oily product with
light petroleum gave major aziridine 41 diastereoisomer as a
colourless solid (0.1 g, 29%) mp 163–165 ЊC (from ethanol).
(Found: C, 64.0; H, 5.9; N, 9.8. C23H25Cl2N3O requires C, 64.2;
H, 5.85; N, 9.75%); δH 0.28 (3H, d, J 6.6, CH3CHBut), 0.72 (9H,
s, But), 2.90 (1H, q, J 6.6, CH3CHBut), 3.91 (1H, d, J 5.3, azir.
H-2), 4.28 (1H, dd, J 5.3 and 3.5, azir. H-3), 6.39 (1H, d, J 3.5,
CHCl2), 6.83–7.50 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)] and
8.07 [1H, dd, J 8.2 and 1.6, 5-H(Q)]; νmax/cmϪ1 1665s and 1590s;
m/z (%) 429 (Mϩ, 2), 347 (38), 346 (100), 174 (22), 117 (21) and
116 (23). Observable signals from the minor diastereoisomer of
aziridine 41 were present in the NMR spectrum of the crude
reaction mixture at δ 2.62 (1H, q, J 6.6, CH3CHBut), 3.81 (1H,
d, J 5.3, azir. H-2) and 4.68 (1H, m, azir. H-3). From the NMR
spectrum of the crude product, the ratio of major : minor dia-
stereoisomers of aziridine 41 is 3.5 : 1 from integration of
signals at δ 3.91 and 3.81 ppm.
Aziridination of methyl cinnamate with Q1NHOAc 5
General procedure 1 was followed using Q1NH2 49 (0.3 g,
0.94 mmol), LTA (0.46 g, 1.03 mmol) and methyl cinnamate
(0.31 g, 1.88 mmol) in dry dichloromethane (4 cm3). The NMR
spectrum of the crude product revealed the presence of dia-
stereoisomers of aziridine 13 (0.15 g, 33%) together with
3H-quinazolinone 14 in a 3 : 1 ratio, from integration of signals
at δ 4.00 ϩ 3.96 ppm and 9.28 respectively. Separation by flash
chromatography over silica with light petroleum–ethyl acetate
(7 : 1) as eluent gave major aziridine 13 diastereoisomer (Rf
0.32) as a colourless oil. (Found: Mϩ 479.234. C26H33N3O4Si
requires M 479.224); δH Ϫ0.11 (3H, s, CH3SiCH3), 0.00 (3H, s,
CH3SiCH3), 0.70 (9H, s, But), 1.24 (3H, d, J 6.3, CH3CHO),
3.51 (3H, s, CO2CH3), 3.58 (1H, d, J 4.7, azir. H-3 or 2), 4.00
(1H, d, J 4.7, azir. H-2 or 3), 5.03 (1H, q, J 6.3, CH3CHO),
7.25–7.55 [8H, m, 6-H, 7-H, 8-H(Q) and CH(Ph)] and 7.98
[1H, d, J 7.9, 5-H(Q)]; νmax/cmϪ1 1760s, 1710s and 1625s; m/z
(%) 479 (Mϩ, 1), 422 (29), 248 (27), 247 (100) and 75 (22);
observable signals from the minor diastereoisomer were present
in the crude reaction product at δ Ϫ0.18 (3H, s, CH3Si-
CH3), Ϫ0.12 (3H, s, CH3SiCH3), 0.55 (9H, s, But), 1.62 (3H, d,
J 6.3, CH3CHO) and 3.96 (1H, d, J 4.7, azir. H-3 or 2). The
Crystal structure determination
Crystal data for 40.† C23H26ClN3O, M = 395.92, monoclinic,
space group P21/c, a = 9.750(9), b = 12.881(14), c = 16.715(9) Å,
β = 95.70(6)Њ, V = 2089(3) Å3, T = 200 K, Z = 4, µ(Mo-Kα) =
0.201 mmϪ1 colourless block, crystal dimensions 0.56 × 0.53 ×
0.43 mm. Full matrix least squares based on F 2 gave R1 = 0.056
for 2156 observed data (F > 4σ(F)) and wR2 = 0.237 for all 2863
data, GOF = 1.072 for 257 parameters.
† CCDC reference number 169151. See http://www.rsc.org/suppdata/
p2/b1/b107327n/ for crystallographic files in .cif or other electronic
format.
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828
827

View Article Online
6 R. S. Atkinson, M. P. Coogan and I. S. T. Lochrie, Tetrahedron Lett.,
1996, 37, 5179.
7 Preliminary communication: R. S. Atkinson, T. A. Claxton, I. S. T.
Lochrie and S. Ulukanli, Tetrahedron Lett., 1998, 39, 5113.
8 D. J. Anderson, D. C. Horwell and R. S. Atkinson, J. Chem. Soc.,
Sect. C, 1971, 624.
9 R. S. Atkinson, M. P. Coogan and I. S. T. Lochrie, J. Chem. Soc.,
Perkin Trans. 1, 1997, 897.
10 B. Gung, J. P. Melnick, M. A. Wolf and A. King, J. Org. Chem.,
1995, 60, 1947; B. Gung and M. B. Francis, Tetrahedron Lett., 1995,
36, 2579.
11 R. S. Atkinson and S. Ulukanli, J. Chem. Soc., Perkin Trans. 2, 1999,
771.
Data were measured on a Siemens P4 diffractometer with
graphite monochromated Mo-Kα radiation (λ = 0.7107 Å) using
an ω scan technique. Three standard reflections monitored
every 100 scans showed no significant variation in intensity, the
reflections were corrected for Lorentz and polarisation effects.
The structure was solved by direct methods and refined by
full-matrix least squares on F 2 using the program SHELXTL.15
All hydrogen atoms were included in calculated positions (C–H
= 0.96 Å) using a riding model. All atoms were refined using
anisotropic displacement parameters.
12 M. S. Newman and P. K. Sujeeth, J. Org. Chem., 1978, 43, 4367.
13 Gaussian 94 (Revision B.2), M. Frisch, G. W. Trunks, H. B. Schlegel,
P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. A.
Keith, G. A. Peterson, J. A. Montgomery, K. Paghavachari, M. A.
Al-Laham, V. G. Zakrewski, J. V. Ortiz, J. R. Foresman, J.
Cioslowski, B. B. Stefanov, A. Nanayakara, M. Challacombe, C. Y.
Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, E. S. Repogle,
R. Gomperts, J. S. Binkley, C. Gonzales, R. L. Martin, D. J. Fox, D.
J. Defrees, J. Baker, M. Head-Gordon, J. P. Stewart and J. A. Pople,
Gaussin Inc., Pittsburgh, PA, 1993.
References
1 R. S. Atkinson, Tetrahedron, 1999, 55, 1519.
2 R. S. Atkinson, S. Ulukanli and P. J. Williams, J. Chem. Soc., Perkin
Trans. 1, 1999, 2121.
3 R. S. Atkinson and P. J. Williams, J. Chem. Soc., Perkin Trans. 1,
1996, 1951.
4 R. S. Atkinson and P. J. Williams, J. Chem. Soc., Perkin Trans. 1,
1996, 205.
5 (a) R. S. Atkinson, A. P. Ayscough, W. T. Gattrell and T. M.
Raynham, J. Chem. Soc., Perkin Trans. 1, 1998, 2783; (b) R. S.
Atkinson and G. Tughan, J. Chem. Soc., Perkin Trans. 1, 1987,
2803.
14 R. S. Atkinson, B. J. Kelly and J. Williams, Tetrahedron, 1992, 48,
7713.
15 G. M. Sheldrick, SHELXTL ver. 6.10, Bruker AXS Inc., Madison,
WI, USA, 2000.
828
J. Chem. Soc., Perkin Trans. 2, 2002, 819–828