The reaction of α- and ω-methylenelactams with nitrones. Influence of electronic and geometric factors on the stereoselectivity of their 1,3-dipolar cycloaddition

Article abstract of DOI:10.1039/b000129p

Various four to six-membered ring α- and ω-methylenelactams 1-5 were reacted with nitrones 6 and afforded good yields of spiroadducts 7-12 through a regiospecific [3 + 2] cycloaddition. Owing to the creation of at least two new asymmetric centres, mixtures of diastereoisomers were usually obtained whose structures were established by two-dimensional NOESY experiments or X-ray crystallography. Among the dipolarophiles, 3-methyleneazetidin-2-ones 1 yielded adducts with a very high stereoselectivity. A similar behaviour was also observed with α-benzoylnitrone 6a as a 1,3-dipole. Electronic and geometric interactions of the reagents 1-5 and 6 in the course of the cycloaddition process were discussed in order to account for the experimental stereochemical outcomes.

Full text of DOI:10.1039/b000129p

View Online / Journal Homepage / Table of Contents for this issue
The reaction of ꢀ- and ꢁ-methylenelactams with nitrones.
Influence of electronic and geometric factors on the
stereoselectivity of their 1,3-dipolar cycloaddition
1
Séverinne Rigolet,a Jean Marie Mélot,*a Joel Vébrel,a Angèle Chiaronib and Claude Richeb
a IUT Département Chimie, 30, Avenue de l’Observatoire, B.P. 1559, 25009 Besançon Cedex,
France
b Institut de Chimie des Substances Naturelles, C.N.R.S., 91198 Gif-sur Yvette, France
Received (in Cambridge, UK) 1st January 2000, Accepted 15th February 2000
Various four to six-membered ring α- and ω-methylenelactams 1–5 were reacted with nitrones 6 and afforded good
yields of spiroadducts 7–12 through a regiospecific [3 ϩ 2] cycloaddition. Owing to the creation of at least two new
asymmetric centres, mixtures of diastereoisomers were usually obtained whose structures were established by two-
dimensional NOESY experiments or X-ray crystallography. Among the dipolarophiles, 3-methyleneazetidin-2-ones 1
yielded adducts with a very high stereoselectivity. A similar behaviour was also observed with α-benzoylnitrone 6a
as a 1,3-dipole. Electronic and geometric interactions of the reagents 1–5 and 6 in the course of the cycloaddition
process were discussed in order to account for the experimental stereochemical outcomes.
Table 1 α-Methylenelactams prepared
Introduction
Heterocyclic spirocompounds are of synthetic interest in
organic chemistry. Indeed, the presence of a spirocarbon atom
induces a relative steric strain and allows the possible thermal
or acid/base promoted rearrangement of these products, yield-
ing new and often unexpected heterocycles.1–7 Some spiro-
compounds and their rearrangement products have recently
shown interesting applications in botany and agronomics.1,2,8
The cycloaddition between dipolarophiles bearing an exocyclic
carbon–carbon double bond and appropriate 1,3-dipoles is
one of the best methods for the synthesis of bicyclic spirocom-
pounds. We have recently reported the reaction of 3-methylene-
isoindolones (methylenephthalimidines) with nitrones.9 Other
examples of 1,3-dipolar cycloadditions between lactams bear-
ing an exo or endo cyclic conjugated double bond have been
described.10–15 In the course of our research, two papers dealing
with the reaction of α-ethylidenelactams with 1,3-dipoles have
been published.16,17 Strauss and Otto reacted α-ethylidene-
azetidin-2-ones with the single N-methyl-α-phenylnitrone and
obtained very unstable adducts under drastic conditions in
modest yields.17 We report here our results for the 1,3-dipolar
cycloaddition of several α- and ω-methylene lactams 1–5
(Table 1) with nitrones 6a–e (Table 2), yielding spirocyclo-
adducts 7–12. Important differences of diastereoselectivity
were encountered in this general reaction, which are discussed
according to the nature of the starting materials and the inter-
actions during the transition state. Some accounts are also
given concerning a comparative reactivity of methylenelactams.
n
R1
R2
1a
1b
1c
2a
2b
3a
3b
0
0
0
1
1
2
2
p-MeO-Ar-
p-Me-Ar-
p-O2N-Ar-
–CH3
–CH3
–CH3
–H
–H
–H
–H
Ph
–H
–H
Ph-CH2–
Table 2 Nitrones 6 prepared
R3
R4
a
b
c
d
e
PhCO–
Ph–
Ph–
Ph–
Ph–
Ph–
Ph–
PhCH2–
But
Me–
1
as shown either by H NMR analysis of the crude reaction
mixtures or TLC.
Results and discussion
Regiochemistry
The [3 ϩ 2] cycloaddition of α-methylenelactams 1–3 and
nitrones 6 gave high yields of stable spiroadducts 7,8a–p
(Schemes 1–3, Table 3). The bislactam 4 afforded bisadducts
9–10 upon reaction with 6a (Scheme 4); cycloaddition of 4
with diphenylnitrone 6b furnished an inextricable mixture of
adducts that we were unable to separate and this reaction was
not further investigated. The 5-methylenepyrrolidin-2-one 5
was very unstable18 and reacted only with the highly reactive
benzoylnitrone 6a (Scheme 5). All methylenelactams 1–5
reacted completely and no secondary product was detected,
All cycloadditions afforded a single regioisomer, identified
as a 5,5-disubstituted isoxazolidine. This regiospecificity was
1
established by H and 13C NMR spectroscopy. First, the 13C
chemical shift of the spiro centre of all adducts was between
80 and 100 ppm (see Experimental section); this value located
the spiro centre near the very electronegative oxygen atom of
the isoxazolidine ring. Secondly, the identification in all adducts
of an ABX system by 1H NMR was in accordance with
the 5-substituted isoxazolidine structure only. These results
supported literature reports.19,20 The theoretical calculations,
DOI: 10.1039/b000129p
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103
This journal is © The Royal Society of Chemistry 2000
1095

View Online
Table 3 Reaction of α-methylenelactams with nitrones; reaction conditions, yields and selectivity
Methylene-
lactam
Yield
(%)
Isomers
Entry
Nitrone
Adducts
T/ЊCa
Time/h
ratio (%)b
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1a
1a
1a
1a
1b
1b
1c
1c
1c
2a
2a
2b
2b
3a
3a
3b
6a
6b
6d
6e
6b
6c
6a
6c
6d
6a
6b
6a
6b
6a
6b
6a
7a
110
110
110
55e
110
110
110
110
110
80
0.25
0.5
6
240
1
3
0.25
3
92c
83c
65c
62c
78c
81
100:0
>95:<5
100:0
>90:<10
>95:<5
90:10
7b:8b
7c
7d:8d
7e:8e
7f:8f
7g
92c
94c
84c
72c
67
100:0
d
7h:8h
7i
7j
6
100:0
100:0
85:15
70:30
54:31:9:6
100:0
80:20
0.5
96
1
7k:8k
7l:7Јl
7m:7Јm:8m:8Јm
7n
7o:8o
7p
50
110
110
80
86
6
88
0.5
96
2.5
80c
73
50
20f
90c
100:0
1
1
a Reaction in toluene unless indicated otherwise. b Evaluated by H NMR in the crude mixture. c Yield of pure 7. d Not evaluated by H NMR.
e Reaction at 110 ЊC caused some formation of by-products. f Ethyl acetate.
Scheme 1
Scheme 3
Scheme 2
performed by the semi-empirical PM3 method included in CS
Chem3D Pro. did not provide any useful information and were
frequently in contradiction with our experimental results. A
strong interaction between the electrophilic sp2 carbon (᎐CH )
᎐
2
of α-methylenelactams 1–3 and the ‘soft’ carbon of the dipolar
linkage was supposed to explain to a great extent the specificity
of the orientation.
Scheme 4
adducts having two new chiral centres, i.e. the quaternary
spiro atom and the carbon bearing the R3 substituent of the
isoxazolidine ring; four asymmetric carbons were created in
the case of bislactam 4. The percentage of stereoisomers in the
crude reaction mixtures has been evaluated by 1H NMR
Stereochemistry
The cycloaddition of non-substituted α-methylenelactams 1–3
or 5-methylenepyrrolidin-2-one 5 with nitrones 6 led to cyclo-
1096
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103

View Online
Fig. 1 ORTEP view of adduct 7a (arbitrary numbering).
Scheme 5
spectroscopy from the relative integration of the N–CHX
proton of the isoxazolidine ring (relative accuracy 5%) or was
achieved only after careful separation of the different adducts
by flash chromatography. Exact structures of spirolactams
7–12 were established by two-dimensional NMR spectro-
scopy (NOESY experiments) and X-ray crystallography. The
formation of diastereoisomeric adducts was never caused by
any interconversion between the Z/E nitrones,21,22 the relative
configuration (Z) of the dipole was always preserved in the
spirocompounds.
Once the structures of cycloadducts were established and in
order to explain the observed stereochemistry, we investigated
the electronic and steric factors between reagents 1–5 and 6
during their cycloadditions.
Cycloaddition of ꢀ-methylenelactams 1–3
The terms ‘endo’ and ‘exo’ refer to the position of the α-methylene-
lactam carbonyl group relative to the dipolar linkage during the
transition state.
Non-substituted α-methylenelactams (R2 = H, Scheme 1)
should normally afford four diastereoisomers distributed into
two pairs of enantiomers 7 and 8.
Fig. 2 The forbidden interaction between α-methylenelactams and
α-benzoylnitrone.
All cycloadditions involving α-benzoylnitrone 6a (Table 3,
entries 1, 7, 10, 14, 16) were stereospecific; the lactam ring size
had no influence on stereospecificity. As a representative experi-
ment,23 the configuration of the single product from the reac-
tion of the substituted azetidinone 1a with 6a was established
by X-ray crystallography: the benzoyl function (R3) and the
lactam carbonyl group were in an anti configuration (Fig. 1).
The structure 7a (rel-4R,7S) was consequently assigned to this
adduct. As shown in Fig. 2, the endo approach which would
lead to the stereoisomers 8 was strongly disadvantaged, since
the two carbonyl groups should be very close together, an
improbable circumstance because of their natural electronic
repulsion.
The cycloaddition of α-methylenelactams with nitrones 6b–e
processed with stereoselectivity. Although minor diastereo-
isomers were difficult to obtain in a pure form, we have isolated
spiroazetidinones 7f and 8f (R4 = Bn) for exact characteris-
ation purposes. Their structures were elucidated from two-
dimensional 1H–1H NOESY spectra and NOE effects reported
in Table 4. The structure 8f rel-(4R,7R) was unambiguously
assigned to the minor isomers by the presence of a very strong
NOE effect between the HX proton and those of the β-lactam
ring. Such an interaction was absent in the case of 7f
rel-(4R,7S). The semi-empirical PM3 method and Dreiding
models procured correct representations of spirolactams 7f
and 8f and both these structures provided confirmation of the
results of the two-dimensional NMR spectroscopy. These
results were unambiguously generalised to other adducts.
Indeed, the formation of 7 was facilitated by the absence of any
steric hindrance in the exo approach of reagents, and thus the
formation of 8 was all the more unfavourable, as the bulk of
the R4 group of nitrones 6 increased and interacted with the
intracyclic methylenic hydrogens (CH -C᎐CH ) of lactams
᎐
2
2
1–3 (Fig. 3 and Table 3, entries 2–6). Indeed, stereospecificity
was reached with N-tert-butylnitrone 6d (Table 3, entries 3
and 9).
We noticed that greater amounts of the minor adducts 8
were obtained with five and six-membered rings lactams 2
and 3. Indeed, whereas methyleneazetidin-2-ones 1 remained
entirely rigid during the nitrone approach, pyrrolidinones and
piperidinones could adopt a conformation (half envelope for 2,
pseudo chair for 3) which minimised the steric effect of the
methylene protons. The endo approach was consequently less
unfavourable.
Upon reaction with nitrones, the 5-substituted methylene-
lactam 2b, which already carried an asymmetric centre, would
normally give rise to eight diastereoisomers, distinguished
in Scheme 3 as four diastereoisomeric pairs of enantio-
mers 7:7Ј:8:8Ј. However, α-benzoylnitrone 6a afforded only
two diastereoisomeric pairs 7l rel-(3R,5R,8R) and 7Јl rel-
(3R,5S, 8S), resulting respectively from the anti (major) and syn
(minor) approaches of the nitrone in relation to the hindering
5-phenyl substituent (Scheme 2). As shown in Fig. 2, the repul-
sive interaction of both carbonyl groups during the approach
of reagents forbade the formation of 8.
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103
1097

View Online
Table 4 Two-dimensional 1H NMR (NOESY experiments, 300 MHz) for adducts 7f, 8fa
7f
HX
HA
HB
HL
CH2
8f
HX
HA
HB
HL
CH2
HX
HA
HB
HL
CH2
/
ϩ
/
ϩ
Ϫ
Ϫ
ϩ
ϩ
/
Ϫ
Ϫ
Ϫ
/
(ϩ)
Ϫ
Ϫ
Ϫ
/
HX
HA
HB
HL
CH2
/
ϩ
ϩ
/
ϩ
Ϫ
Ϫ
ϩ
ϩ
/
(ϩ)
Ϫ
Ϫ
Ϫ
/
ϩ
ϩ
Ϫ
(ϩ)
ϩ
Ϫ
Ϫ
Ϫ
(ϩ)
/
(ϩ)
(ϩ)
Ϫ
Ϫ
Ϫ
a HL = β-lactam protons; CH2 = benzylic protons; “ϩ” = strong NOE effect; “(ϩ)” = weak NOE effect; “Ϫ” = no effect; the grey tint square
indicates the determining 1H–1H-correlation through space for adduct 8.
Table 5 Two-dimensional 1H NMR (NOESY experiments, 300 MHz) for adducts 7m, 7Јm and 8ma
7m
HX
HA
HB
HC
HD
HE
7Јm
HX
HA
HB
HC
HD
HE
8m
HX
HA
HB
HC
HD
HE
HX
HA
HB
HC
HD
HE
/
ϩ
/
ϩ
Ϫ
Ϫ
Ϫ
ϩ
ϩ
/
Ϫ
Ϫ
Ϫ
/
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
ϩ
ϩ
/
HX
HA
HB
HC
HD
HE
/
ϩ
/
ϩ
Ϫ
Ϫ
Ϫ
ϩ
ϩ
/
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
ϩ
/
Ϫ
Ϫ
Ϫ
ϩ
ϩ
/
HX
HA
HB
HC
HD
HE
/
ϩ
/
ϩ
ϩ
ϩ
/
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
ϩ
/
Ϫ
Ϫ
Ϫ
ϩ
ϩ
/
ϩ
ϩ
Ϫ
Ϫ
Ϫ
ϩ
ϩ
Ϫ
Ϫ
Ϫ
ϩ
ϩ
Ϫ
Ϫ
Ϫ
Ϫ
/
ϩ
ϩ
/
ϩ
/
ϩ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
ϩ
Ϫ
ϩ
ϩ
ϩ
ϩ
ϩ
a “ϩ” = positive NOE effect; “Ϫ” = no effect; the grey tint square indicates the determining 1H–1H-correlation through space for each structure.
Fig. 3 Interaction of α-methylenelactams and α-phenylnitrones.
The structures of the diastereoisomers (four enantio-
meric pairs 7m:7Јm:8m:8Јm) stemming from the reaction of
diphenylnitrone 2b with 6b (Scheme 3) were established from
two-dimensional 1H–1H spectroscopy (NOESY experiments)
and the NOE effects are summarised in Table 5. Most of the
signals of the 2D spectrum of the very minor isomer 8Јm
were unfortunately unclear and we could not give satisfactory
analysis. The major isomers 7m rel-(3R,5R,8R) and 7Јm rel-
(3R,5S,8S) were formed (>80%) according to the favoured
interaction of the reagents (Fig. 3, exo approach); the selectivity
proceeding from the anti approach towards the 5-phenyl played
a secondary part only.
Fig. 4 Interactions of methylenepyrrolidin-2-one 5 and benzoyl-
nitrone 6.
Cycloadditions of bislactam 4
With 6a, this molecule gave two pairs of symmetrical adducts 9
and 10 among all those possible. The structure of the major
diastereoisomer (80%) was established by X-ray crystallography
(Fig. 5); the benzoyl group of each isoxazolidine ring and the
neighbouring lactam carbonyl were opposite and this stereo-
chemistry was in accordance with structure 9 rel-(3R,5R,
Cycloaddition of 5-methylenepyrrolidin-2-one 5 (Scheme 5)
This ω-methylenelactam gave rise to the non-separable 2,6-
diazaspiro[4.4]nonanes 11 (RS,SR) and 12 (RR,SS). The 2D
NMR NOESY spectrum of the mixture indicated a strong
NOE effect between the HX proton and the N–Me group of the
minor isomers exclusively. Consequently, structure 12 has been
ascribed to the minor isomer from its representation by CS
Chem3D Pro. (PM3). Furthermore, the schematisation of the
approach leading to 12 clearly showed that this one was dis-
advantaged, because it introduced some hindrance between the
N-phenyl group of 6a and the N-methyl of 5 (Fig. 4). No elec-
tronic feature was denoted here, since the position of the lactam
carbonyl group never interacted with the carbonyl of the
nitrone benzoyl function.
8S,11S). The very strong electronic repulsion between the C᎐O
᎐
groups of 4 and 6a should normally prevent the formation of
10. In fact, as shown in Fig. 6, the carbonyl group and the
exocyclic methylene group of 4 were not in the same plane.
Consequently, the repulsion between the carbonyl groups of 4
and 6a during the disadvantaged approach leading to 10 was
reduced.
Reactivity of methylenelactams
Apart from the stable crystalline 3-methyleneazetidin-2-ones 1,
all other methylenelactams were intrinsically or thermally
1098
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103

View Online
Fig. 5 ORTEP view of bisadduct 9 (arbitrary numbering).
The bislactam 4 did not add two equivalents of nitrone 6a
simultaneously: the reaction of 4 with only one equivalent of
α-benzoylnitrone gave approximately equivalent amounts of
monoadducts and bisadducts 9 and 10 (see Experimental
section). The formation of these latter compounds was rather
slow, probably for steric reasons. Piperazine 4 might be con-
sidered as an α- and ω-methylenelactam, because the exocyclic
double bonds were located in the α-position of both the
carbonyl group and nitrogen: the lone pair of each nitrogen
atom compensated for the electronic deficiency of this exocyclic
double bond. The reactivity of the latter was halfway between
those of α- and ω-methylenelactams.
Conclusions
Fig. 6 View of compound 4 modelled by PM3.
The [3 ϩ 2] cycloaddition of α-methylenelactams 1–3 with
nitrones 6 was a general and high-yielding regiospecific
reaction. Its stereoselectivity was discussed from the possible
electronic or steric interactions of reagents during their
approaches. The major stereoisomers 7 were the result of the
preferential approach locating the lactam carbonyl group in the
exo position relative to the dipolar linkage. The endo approach
was entirely forbidden when α-benzoylnitrone 6a was used as a
dipole because of electronic repulsion between the carbonyl
groups of both reagents. It was also greatly prevented in the
case of other nitrones, whose nitrogen substituent R4 created a
steric hindrance towards the lactam hydrogen atoms. Particu-
larly, the rigidity and strong internal strain of 3-methylene-
azetidin-2-ones 1 explained their exceptionally high reactivity
and the great stereoselectivity of their cycloadditions with
nitrones. These properties should be investigated because of the
importance of the β-lactam ring and possible later chemical
transformations of spirocompounds. The bislactam 4 behaved
as an α- and ω-methylenelactam and gave simple exploitable
results with α-benzoylnitrone 6a only. The ω-methylenelactam 5
was much less reactive, because of its own unstable character
and the deactivation of the exo-methylene group: it reacted only
with 6a. The formation of the major stereoisomers 11 was
explained by the absence of steric hindrance between the N–R4
group of the nitrone and the N–Me moiety of 5.
fragile and, as shown in Table 3 and Schemes 1–5, we were
unable to apply any standard conditions within the scope of
their cycloaddition with nitrones. We have recently shown9 that
differential thermal analysis (DTA) proved to be a very useful
technique for exploring the thermal behaviour of organic
reagents and quickly establishing reaction conditions. DTA was
applied here and conditions stated in Table 3 were regarded as
ideal. Although any comprehensive comparison of the reactiv-
ity of methylenelactams was impossible, some unambiguous
conclusions were drawn from our results.
In particular, all thermal analyses of equimolecular mixtures
of 3-methyleneazetidin-2-ones 1 and nitrones 6 revealed that
the reagents exothermically cycloadded immediately they
melted. The very high reactivity of β-lactams 1 was also immedi-
ately obvious from the experimental results (Table 3, entries
1–9). There was no doubt that the release of the strong intra-
cyclic strain in the course of the cycloaddition explained the very
good dipolarophilic activity of 1. Conversely, piperidin-2-one
3b and 5-methylenepyrrolidin-2-one 5 were not strained rings
and their relative reactivity was compared using similar reac-
tion conditions. When the ω-methylenelactam 5 was reacted
with nitrone 6b (20 ЊC, 30 h), the dipole was quantitatively
recovered, whereas 5 degraded into a gummy polymer. In the
same reaction conditions, the α-methylenelactam 3b gave 35%
of spiroadducts, and cycloaddition was complete within a few
days. The presence of a conjugated withdrawing group is a
well-known factor enhancing the dipolarophilic character of
an alkene.24 The weak reactivity of the unactivated 5-methyl-
enepyrrolidin-2-one 5 was compared with methylenephthal-
imidines, their benzo derivatives, whose more stable character
permitted stronger reaction conditions.9
Experimental
General
Melting points were taken using a Dr Tottoli apparatus and are
uncorrected. IR spectra were measured on a Bruker Spectro-
spin (FT-IR) IFS-45 apparatus. 1H and 13C NMR spectra were
run on a Bruker Spectrospin AC200 spectrometer at 200.13
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103
1099

View Online
MHz (1H) and 50.32 MHz (13C). Two-dimensional 1H–1H
spectroscopy (NOESY experiments) of compounds 7f, 8f, 7m,
7Јm, 8m and 8Јm were obtained from a Bruker Avance 300
spectrometer at 300.13 MHz. All compounds were dissolved in
CDCl3 (except for 11 and 12) with 0.1% TMS as internal refer-
ence. The 13C NMR spectra were obtained from proton-noise
decoupled spectra. Chemical shifts are in ppm on the δ scale
and coupling constants J are given in Hz. In 13C NMR spectra,
carbons are numbered according to the IUPAC nomenclature
rules for spirocompounds. Spectroscopic data of very minor
diastereoisomers are acquired from the crude mixtures of
adducts and only well-resolved signals are reported here. Dif-
ferential thermal analysis was performed using a Mettler
Thermal Mechanical Analyser TMA 40. Samples of 15.0 mg of
reagents 1–6 or equimolar mixtures of 1:6 were placed in an
aluminium crucible and their thermal behaviour automatically
analysed (20–150 ЊC; temperature gradient: ϩ10 ЊC minϪ1). All
flash liquid chromatographies were performed in 2.5 cm inside
diameter Pyrex columns with Kieselgel 60, particle size of
0.063–0.200 mm, from Merck (art. 1.07734.1000). Diethyl ether
and THF (Carlo Erba, 99.5% min) were allowed to stand over
anhydrous calcium chloride for 24 hours and thoroughly dried
on sodium wire before use; toluene was directly dried over
sodium wire. Other reagents or solvents were commercial
products and used as received. Owing to the strong electro-
philic character of most α-methylenelactams, great care should
be taken during their manipulation and protective gloves should be
worn.
100 ЊC for 90 minutes. After cooling to room temperature, the
solid residues were filtered through a pad of Celite. The filtrate
was dried over anhydrous sodium sulfate and most of the xylene
evaporated under reduced pressure. The resulting oil (2a, 3a)
was purified by vacuum distillation and stored at Ϫ20 ЊC. The
bismethylenelactam 4 was recrystallised from a propan-1-ol–
cyclohexane mixture.
1-Methyl-3-methylenepyrrolidin-2-one 2a. 60%, bp0.2 mmHg: 48–
49 ЊC, nD20: 1.5025. δH: 2.80 (m, 2H), 2.98 (s, 3H, CH3), 3.43 (m,
2H), 5.32 (t, 1H, J 2.2), 5.96 (t, 1H, J 2.2).
1-Methyl-3-methylenepiperidin-2-one 3a. 45%, bp0.2 mmHg: 55–
58 ЊC, nD20: 1.5080. δH: 1.89 (m, 2H), 2.56 (dd, 2H, J 1.5 and 6.0),
3.01 (s, 3H, CH3), 3.39 (m, 2H), 5.25 (dd, 1H, J 1.5 and 3.5),
6.16 (dd, 1H, J 1.5 and 3.5).
1,4-Dimethyl-3,6-dimethylenepiperazine-2,5-dione 4. 55%,
mp: 140–142 ЊC (rapid heating, decomp.). δH: 3.30 (s, 6H,
N-Me), 4.99 (m, 2H), 5.88 (m, 2H).
The 5-substituted pyrrolidin-2-one 2b was synthesised from
ethyl bromomethylacrylate32 and N-methylbenzylideneimine,
following the well-described procedure of Villiéras et al.33,34
1-Methyl-3-methylene-5-phenylpyrrolidin-2-one 2b. 65%, mp:
68–70 ЊC. νmax (KBr)/cmϪ1: 1685 (C᎐O), 1660 (C᎐C). δ : 2.62
᎐
᎐
H
(ddt, 1H, J 2.5, 4.4 and 17.0), 2.69 (s, 3H, N-CH3), 3.22 (ddt,
1H, J 2.5, 8.5 and 17.0), 4.50 (dd, 1H, J 4.4 and 8.5), 5.35
(dd, 1H, J 2.3 and 2.7), 6.07 (dd, 1H, J 2.3 and 2.7), 7.2–7.4
(m, 5H).
Preparation of starting materials
The piperidin-2-one 3b was obtained by the Ac2O promoted
rearrangement of N-benzylnipecotic acid (N-benzyl-3-carb-
oxypiperidine).35 The authors used an excessive amount of
acetic anhydride. Dividing this quantity by 10 gave high yields
of 3b and greatly simplified the work-up of the crude mixture;
the methylenepiperidin-2-one was thus obtained in a 75% over-
all yield from the commercial inexpensive ethyl nipecotate.
The 5-methylenepyrrolidin-2-one 5 was prepared by the fol-
lowing modified procedure.18 Aqueous 40% methylamine (3 ml,
35 mmol) was slowly added at 0 ЊC to pent-3-eno-4-lactone
(α-angelicalactone) (3.5 g, 35 mmol). After 2 hours in the cold,
water was distilled off (20 ЊC, 10 mmHg), leaving an oil which
was crystallised from diethyl ether. The crude product (3 g, 23
mmol) was dehydrated at 70–80 ЊC under high vacuum (0.15
mmHg) and 5 was collected in liquid nitrogen. Overall yield:
56%, mp: 58–59 ЊC. This lactam was used immediately; it poly-
merised readily even at low temperature.
The α-benzoyl-N-phenylnitrone 6a [N-(2-phenyl-2-oxoethyl-
idene)aniline N-oxide] was prepared by the following modified
method.25 A solution of commercial nitrosobenzene (5.35 g,
0.05 mol) in acetone (40 cm3) was cooled with an ice–salt bath.
Phenacylpyridinium bromide (13.9 g, 0.05 mol) was dissolved
in the minimum of water (20–30 cm3) and cooled. The two
resultant suspensions were combined and 1 mol dmϪ3 aqueous
sodium hydroxide solution was very slowly added with con-
tinuous pH control over a period of ca. 3–4 h at 0 ЊC under
vigorous stirring. The addition was stopped when the pH
reached and remained at 7 (40–50 ml were required). After 1 h
in the cold bath, the orange–yellow solid was filtered, washed
with ice water, then with diethyl ether and air-dried. Yield: 9.8 g
(87%). Mp: 107–109 ЊC. Addition of the sodium hydroxide solu-
tion without extreme care led to an intractable gum.
Nitrones 6b–d were obtained by the hydrogen peroxide oxid-
ation of the corresponding N-substituted benzylamines in
85–90% yields.26 The N-methyl-α-phenylnitrone 6e [(N-benzyl-
idene)methylamine N-oxide] was synthesised from benzalde-
hyde and methylhydroxylamine hydrochloride in 86% yield.27
The 3-methyleneazetidin-2-ones 1a–c were obtained on a
20–30 g scale from the readily accessible 2-bromomethyl-3-
bromopropanoic acid.28,29
The lactams 2a, 3a and 4 were prepared from N-methyl-
pyrrolidin-2-one, N-methylpiperidin-2-one and 1,4-dimethyl-
piperazine-2,5-dione (“sarcosine anhydride”) respectively,
following substantially modified known procedures.30,31 Sodium
hydride (60% dispersion in mineral oil, 8.2 g, 205 mmol) and
diethyl oxalate (43.2 cm3, 49 g, 315 mmol) were suspended in
anhydrous diethyl ether or THF (150 cm3) and the mixture
heated at 35–40 ЊC under an argon atmosphere. The appropriate
lactam (150 mmol) was added dropwise (or portionwise for the
solid piperazinedione) over a one hour period. Stirring was
continued for 3 days at 35–40 ЊC, and the mixture cooled to
0 ЊC. The grey pasty solid was filtered off, washed with anhy-
drous diethyl ether and dried under reduced pressure (40 ЊC,
20 mmHg). This sodium salt of the 3-ethoxalyllactam and
paraformaldehyde (9.5 g, 315 mmol) were then suspended in
xylene (200 cm3) and the heterogeneous mixture was stirred at
General procedure for the preparation of substituted 1-oxo-5-
oxa-2,6-diazaspiro[3.4]octanes 7,8a–i (Table 3, entries 1–9)
The equimolecular mixture of 3-methyleneazetidin-2-one 1 and
nitrone 6 (2.5 mmol) in anhydrous toluene (5 cm3) was stirred
at 55–110 ЊC under an argon atmosphere until reaction was
complete (TLC, cyclohexane–ethyl acetate, 80:20). The solvent
was then evaporated under reduced pressure and the adducts
isolated by addition of 95% ethanol (3 cm3), cooling at Ϫ10 ЊC,
filtration, washing with ethanol (5 cm3) and air-drying.
Recrystallisation from propan-1-ol gave analytically pure
products 7.
7-Benzoyl-2-(4-methoxyphenyl)-1-oxo-5-oxa-2,6-diazaspiro-
[3.4]octane 7a. Yield: 92%, mp: 155–156 ЊC. δH: 2.75 (dd, 1HB,
J 6.6 and 12.8), 3.08 (d, 1HA, J 12.8), 3.78 (s, 3H, O-CH3), 3.85
(d, 1H, J 6.5), 4.10 (d, 1H, J 6.5), 5.38 (d, 1HX, J 6.6), 6.9–8.1
(m, 14H). δC: 34.0 (C8), 55.3 (O-CH3), 55.5 (C3), 69.3 (C7),
91.3 (C4), 114.3–156.3 (aromatic C), 164.0 (C1), 195.4 (C᎐O,
᎐
benzoyl). νmax(KBr)/cmϪ1: 1740 (C᎐O β-lactam), 1690 (C᎐O
᎐
᎐
benzoyl) (Found: C, 72.3; H, 5.5; N, 6.9. C25H22N2O4 requires
C, 72.45; H, 5.35; N, 6.76%).
The total structure has been published.23
1100
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103

View Online
6,7-Diphenyl-2-(4-methoxyphenyl)-1-oxo-5-oxa-2,6-diaza-
spiro[3.4]octane 7b:8b. Yield 83%. Major diastereoisomer 7b:
mp: 130–131 ЊC. δH: 2.71 (dd, 1HB, J 3.8 and 12.8), 3.13 (dd,
1HA, J 7.1 and 12.8), 3.75 (d, 1H, J 6.1), 3.78 (s, 3H, O-CH3),
3.80 (d, 1H, J 6.1), 4.89 (dd, 1HX, J 3.8 and 7.1), 6.8–7.4 (m,
14H). δC: 42.2 (C8), 55.3 (O-CH3), 55.3 (C3), 70.5 (C7), 89.3 (C4),
112.6–151.5 (aromatic C), 164.1 (C1). νmax(KBr)/cmϪ1: 1745
(Found: C, 74.2; H, 5.8; N, 7.3. C24H22N2O3 requires C, 74.59;
H, 5.74; N, 7.25%). Minor diastereoisomer 8b: selected δH: 4.71
(t, 1HX, J 7.8); selected δC: 44.8 (C8), 54.9 (C3), 69.1 (C7), 89.3
(C4).
1HB, J 6.6 and 12.8), 3.20 (d, 1HA, J 12.8), 3.95 (d, 1H, J 7.1),
4.30 (d, 1H, J 7.1), 5.47 (d, 1HX, J 6.6), 7.3–8.3 (m, 14H).
δC: 34.2 (C8), 55.9 (C3), 69.7 (C7), 92.0 (C4), 115.8–149.5 (aro-
matic C), 165.9 (C1), 195.3 (C᎐O, benzoyl). νmax(KBr)/cmϪ1
:
᎐
1740 (C᎐O β-lactam), 1685 (C᎐O benzoyl) (Found: C, 66.9; H,
᎐
᎐
4.5; N, 9.8. C24H19N3O2 requires C, 67.13; H, 4.46; N, 9.79%).
6-Benzyl-2-(4-nitrophenyl)-1-oxo-7-phenyl-5-oxa-2,6-diaza-
spiro[3.4]octane 7h:8h. Yield: 94%. Major diastereoisomer 7h:
mp: 174–175 ЊC. δH: 2.80 (dd, 1HB, J 7.2 and 12.9), 3.20 (dd,
1HA, J 7.2 and 12.9), 3.80 (d, 1H, J 6.5), 3.95 (d, 1H, J 6.5), 4.02
(d, 1H, J 14.0), 4.09 (d, 1H, J 14.0), 4.32 (t, 1HX, J 7.2), 7.3–8.2
(m, 14H). δC: 43.5 (C8), 55.2 (C3), 61.1 (CH2-Ph), 69.7 (C7), 89.4
2-(4-Methoxyphenyl)-1-oxo-7-phenyl-6-tert-butyl-5-oxa-2,6-
diazaspiro[3.4]octane 7c. Yield: 65%, mp: 155–156 ЊC. δH: 1.12
(s, 9H, C(CH3)3), 2.61 (dd, 1HB, J 6.7 and 12.7), 3.07 (dd, 1HA,
J 7.9 and 12.7), 3.60 (d, 1H, J 5.9), 3.78 (s, 3H, O-CH3),
3.80 (d, 1H, J 5.9), 4.52 (dd, 1HX, J 6.7 and 7.9), 6.8–7.5 (m,
9 H). δC: 26.3 (3 CH3), 46.2 (C8), 55.5 (O-CH3), 55.5 (C3), 59.7
(C(CH3)3), 62.7 (C7), 88.2 (C4), 116.5–142.6 (aromatic C), 164.7
(C1). νmax(KBr)/cmϪ1: 1745 (Found: C, 72.1; H, 7.1; N, 7.6.
C22H26N2O3 requires C, 72.11; H, 7.15; N, 7.64%).
(C4), 116.5–143.5 (aromatic C), 166.3 (C1). νmax(KBr)/cmϪ1
:
1755 (Found: C, 69.5; H, 5.2; N, 10.1. C24H21N3O4 requires C,
69.39; H, 5.10; N, 10.11%). Minor diastereoisomer 8h: selected
δC: 44.8 (C8), 54.8 (C3), 69.2 (C7), 89.5 (C4).
2-(4-Nitrophenyl)-1-oxo-7-phenyl-6-tert-butyl-5-oxa-2,6-
diazaspiro[3.4]octane 7i. Yield: 84%, mp: 205 ЊC. δH: 1.13 (s,
9H, C(CH3)3), 2.65 (dd, 1HB, J 6.6 and 12.8), 3.10 (dd, 1HA,
J 7.8 and 12.8), 3.70 (d, 1H, J 6.3), 3.95 (d, 1H, J 6.3), 4.54 (dd,
1HX, J 6.6 and 7.8), 7.2–8.2 (m, 9H). δC: 26.1 (3 CH3), 46.3 (C8),
55.5 (C3), 59.7 (C(CH3)3), 62.6 (C7), 88.6 (C4), 116.6–143.4
(aromatic C), 166.2 (C1). νmax(KBr)/cmϪ1: 1745 (Found: C,
65.9; H, 6.0; N, 11.0. C21H23N3O4 requires C, 66.13; H, 6.08;
N, 11.02%).
2-(4-Methoxyphenyl)-6-methyl-1-oxo-7-phenyl-5-oxa-2,6-
diazaspiro[3.4]octane 7d. The equimolecular mixture of 1a and
6e (2.5 mmol) in toluene (5 cm3) was stirred at 50–60 ЊC under
argon for 10 days. Cyclohexane (5 cm3) was then added, giving a
white solid. After filtration and air-drying, 7d was recrystallised
from cyclohexane–ethyl acetate (1:1). Yield: 62%, mp: 125–
126 ЊC. δH: 2.65 (dd, 1HB, J 6.5 and 12.5), 3.07 (dd, 1HA, J 7.2
and 12.5), 3.54 (d, 1H, J 5.9), 3.73 (s, 3H, O-CH3), 3.84 (d, 1H,
J 5.9), 3.92 (dd, 1HX, J 6.5 and 7.2), 6.8–7.4 (m, 9H). δC: 42.8
(N-CH3), 43.2 (C8), 54.5 (O-CH3), 54.5 (C3), 71.6 (C7), 87.7
General procedure for the preparation of substituted 1-oxo-6-
oxa-2,7-diazaspiro[4.4]nonanes 7,8j–k and 1-oxo-7-oxa-2,8-
diazaspiro[4.5]decanes 7,8n–o (Table 3, entries 10, 11, 14, 15)
The mixture of nitrone 6a or 6b (5 mmol) and 3-methylene-
pyrrolidin-2-one 2a or piperidin-2-one 3a (5–7 mmol, see
below), was dissolved in dry toluene (5 cm3) and hydroquinone
was added (10–15 mg). The reaction vessel was heated at the
appropriate temperature under argon until TLC (ethyl acetate–
cyclohexane 70:30) indicated complete consumption of the
nitrone. The solvent was evaporated under reduced pressure
and the crude residue analysed. The adducts were purified by
recrystallisation from n-propanol.
(C4), 113.4–155.5 (aromatic C), 163.7 (C1). νmax(KBr)/cmϪ1
:
1750 (Found: C, 70.5; H, 6.3; N, 8.6. C19H20N2O3 requires C,
70.35; H, 6.21; N, 8.64%).
6,7-Diphenyl-2-(4-methylphenyl)-1-oxo-5-oxa-2,6-diazaspiro-
[3.4]octane 7e:8e. Total yield: 78%. Major diastereoisomer 7e:
mp: 131–132 ЊC. δH: 2.30 (s, 3H, CH3), 2.69 (dd, 1HB, J 5.8 and
12.6), 3.22 (dd, 1HA, J 7.2 and 12.6), 3.49 (d, 1H, J 6.1), 3.76
(d, 1H, J 6.1), 4.90 (dd, 1HX, J 5.8 and 7.2), 7.0–7.6 (m,
14H). NOESY experiment for adduct 7e is reported in Table 4.
δC: 20.8 (CH3), 42.4 (C8), 55.2 (C3), 70.6 (C7), 89.3 (C4), 116.7–
150.9 (aromatic C), 164.5 (C1). νmax(KBr)/cmϪ1: 1760 (Found:
C, 77.5; H, 6.0; N, 7.7. C24H22N2O2 requires C, 77.81; H, 5.99;
N, 7.56%). Minor diastereoisomer 8e: selected δH: 3.70 (d, 1H),
3.95 (d, 1H), 4.72 (t, 1HX, J 7.6); selected δC: 44.9 (C8), 54.8
(C3), 69.2 (C7), 89.3 (C4).
8-Benzoyl-2-methyl-1-oxo-7-phenyl-6-oxa-2,7-diazaspiro-
[4.4]nonane 7j. Reaction at 80 ЊC for 30 min with 5 mmol of 2a.
Yield: 72%, mp: 119–121 ЊC. δH: 2.27 (dd, 1H, J 7.4 and 8.3),
2.51 (m, 1H, J 2.4 and 7.4), 2.62 (dd, 1HB, J 5.7 and 12.2),
2.91 (s, 3H, N-CH3), 3.09 (dd, 1HA, J 8.3 and 12.2), 3.25 (m,
1H, J 2.4 and 8.3), 3.55 (m, 1H, J 2.4 and 7.4), 5.39 (dd, 1HX,
J 5.7 and 8.3), 7.0–8.1 (m, 10H). δC: 30.3 (CH3), 31.5 (C4), 40.3
(C9), 45.9 (C3), 70.2 (C8), 84.7 (C5), 114.5–149.8 (aromatic C),
171.2 (C1), 196.1 (C᎐O, benzoyl). νmax(KBr)/cmϪ1: 1710 (C᎐O
6-Benzyl-2-(4-methylphenyl)-1-oxo-7-phenyl-5-oxa-2,6-diaza-
spiro[3.4]octane 7f:8f. For identification purpose, the crude
mixture was dissolved in hot toluene (2 cm3) and chromato-
graphed on silica gel (150 g) with cyclohexane–ethyl acetate
(85:15) as eluent. Total yield: 81%. Major diastereoisomer 7f:
mp: 117–118 ЊC. δH: 2.30 (s, 3H, CH3), 2.70 (dd, 1HB, J 7.3 and
12.9), 3.20 (dd, 1HA, J 7.3 and 12.9), 3.65 (d, 1H, J 5.9), 3.80
(d, 1H, J 5.9), 4.01 (d, 1H, J 14.0), 4.11 (d, 1H, J 14.0), 4.28
(t, 1HX, J 7.3), 7.1–7.5 (m, 14H). δC: 20.8 (CH3), 43.5 (C8), 54.8
(C3), 61.3 (CH2-Ph), 69.8 (C7), 89.0 (C4), 116.7–139.3 (aromatic
C), 165.3 (C1). νmax(KBr)/cmϪ1: 1740 (Found: C, 78.0; H, 6.3; N,
7.4. C25H24N2O2 requires C, 78.10; H, 6.29; N, 7.29%). Minor
diastereoisomer 8f: mp: 177–178 ЊC. δH: 2.24 (s, 3H, CH3), 2.82
(dd, 1HB, J 7.2 and 13.1), 2.92 (dd, 1HA, J 7.2 and 13.1), 3.70 (d,
1H, J 6.2), 3.75 (d, 1H, J 6.2), 3.90 (d, 1H, J 13.6), 4.00 (d, 1H,
J 13.6), 4.03 (t, 1HX, J 7.2), 7.0–7.4 (m, 14H). δC: 21.1 (CH3),
48.5 (C8), 54.5 (C3), 59.2 (CH2-Ph), 69.3 (C7), 89.0 (C4), 116.0–
139.0 (aromatic C), 165.5 (C1). NOESY experiments for 7f and
8f are summarised in Table 4.
᎐
᎐
γ-lactam), 1690 (C᎐O benzoyl) (Found: C, 71.1; H, 5.8; N, 8.3.
᎐
C20H20N2O3 requires C, 71.41; H, 5.99; N, 8.33%).
7,8-Diphenyl-2-methyl-1-oxo-6-oxa-2,7-diazaspiro[4.4]-
nonane 7k:8k. Reaction at 50 ЊC for 96 hours with 7.5 mmol
of 2a. Total yield: 67%. Major diastereoisomer 7k (59%): mp:
127 ЊC. δH: 2.12 (m, 1H, J 7.3 and 13.8), 2.43 (dd, 1HB, J 7.6 and
12.3), 2.52 (dd, 1H, J 3.2 and 13.8), 2.91 (s, 3H, N-CH3), 3.12
(dd, 1HA, J 7.6 and 12.3), 3.24 (dd, 1H, J 3.2 and 8.9), 3.54 (dd,
1H, J 3.2 and 7.3), 4.91 (t, 1HX, J 7.6), 6.9–7.6 (m, 10H).
δC: 30.2 (CH3), 31.7 (C4), 45.8 (C3), 47.2 (C9), 70.3 (C8), 83.8
(C5), 115.0–150.5 (aromatic C), 171.7 (C1). νmax(KBr)/cmϪ1
:
1700 (Found: C, 74.2; H, 6.6; N, 9.1. C19H20N2O2 requires C,
74.00; H, 6.54; N, 9.11%). Minor diastereoisomer 8k: selected
δH: 2.96 (s, 3H, N-CH3), 4.62 (dd, 1HX, J 7.8 and 9.7); selected
δC: 30.3 (CH3), 45.4 (C3), 47.8 (C9), 69.7 (C8).
9-Benzoyl-2-methyl-1-oxo-8-phenyl-7-oxa-2,8-diazaspiro-
[4.5]decane 7n. Reaction at 80 ЊC for 30 min with 5 mmol of 3a.
Yield: 80%, mp: 136–138 ЊC. δH: 1.90 (m, 2H), 2.32 (dd, 1HB,
7-Benzoyl-2-(4-nitrophenyl)-1-oxo-6-phenyl-5-oxa-2,6-diaza-
spiro[3.4]octane 7g. Yield: 92%, mp: 187–189 ЊC. δH: 2.80 (dd,
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103
1101

View Online
J 8.4 and 12.0), 2.33 (m, 2H), 2.90 (s, 3H, N-CH3), 3.35 (m, 2H),
3.41 (dd, 1HA, J 8.4 and 12.0), 5.44 (t, 1HX, J 8.4), 6.9–8.1 (m,
10H). δC: 19.5 (C4), 32.6 (CH3), 35.0 (C5), 43.7 (C9), 49.6 (C3),
71.3 (C8), 81.0 (C6), 112.8–151.5 (aromatic C), 168.4 (C1), 196.4
(C᎐O, benzoyl). νmax(KBr)/cmϪ1: 1655 (C᎐O δ-lactam), 1685
of silica gel (eluent: cyclohexane–ethyl acetate 85:15, then
70:30). Total yield: 88%. Diastereoisomer 7m: 48%; mp:
150 ЊC. δH: 1.98 (dd, 1HD, J 7.1 and 14.2), 2.40 (dd, 1HB, J 7.7
and 12.4), 2.72 (s, 3H, CH3), 2.95 (dd, 1HC, J 7.1 and 14.2),
3.28 (dd, 1HA, J 7.7 and 12.4), 4.71 (t, 1HE, J 7.1), 4.95 (t,
1HX, J 7.7), 6.9–7.5 (m, 15H). δC: 29.1 (CH3), 41.6 (C4), 47.8
(C9), 62.1 (C3), 71.1 (C8), 84.3 (C5), 117.2–151.0 (aromatic C),
᎐
᎐
(C᎐O benzoyl) (Found: C, 71.7; H, 6.4; N, 7.8. C H N O
᎐
20 20
2
3
requires C, 71.98; H, 6.33; N, 7.99%).
173.0 (C᎐O). Diastereoisomer 7Јm: 27%, mp: 143 ЊC. δ : 2.27
᎐
H
8,9-Diphenyl-2-methyl-1-oxo-7-oxa-2,8-diazaspiro[4.5]decane
7o:8o. Reaction at 50 ЊC for 96 hours with 7.5 mmol of 3a.
Yield: 73%. Major diastereoisomer 7o (63%), mp: 143 ЊC. δH:
1.89 (m, 2H), 2.18 (dd, 1HB, J 9.5 and 12.2), 2.45 (m, 2H), 2.92
(s, 3H, N-CH3), 3.32 (dd, 1HA, J 7.5 and 12.2), 3.38 (m, 2H),
4.95 (dd, 1HX, J 7.5 and 9.5), 6.9–7.5 (m, 10H). δC: 19.7 (C4),
33.6 (CH3), 34.9 (C5), 49.7 (C3), 50.3 (C9), 70.9 (C8), 80.5 (C6),
112.5–154.2 (aromatic C), 169.0 (C1). νmax(KBr)/cmϪ1: 1650
(Found: C, 74.4; H, 7.0; N, 8.7. C20H22N2O2 requires C, 74.51;
H, 6.88; N, 8.69%). Minor diastereoisomer 8o: selected δH: 3.01
(s, 3H, CH3), 4.54 (dd, 1HX, J 7.3 and 9.9). δC: 19.5 (C4),
32.8 (CH3), 35.4 (C5), 47.1 (C3), 48.8 (C9), 69.3 (C8), 81.7 (C6),
169.0 (C1).
(dd, 1HD, J 4.8 and 13.9), 2.55 (dd, 1HC, J 8.0 and 13.9), 2.71
(dd, 1HB, J 7.2 and 12.2), 2.80 (s, 3H, CH3), 2.98 (dd, 1HA,
J 10.0 and 12.2), 4.44 (dd, 1HE, J 4.8 and 8.0), 4.60 (dd, 1HX,
J 7.2 and 10.0), 6.7–7.6 (m, 15H). δC: 29.5 (CH3), 42.2 (C9), 49.2
(C4), 61.4 (C3), 70.3 (C8), 84.5 (C5), 115.5–152.0 (aromatic C),
171.9 (C᎐O). Diastereoisomer 8m: 8%, mp: 137–138 ЊC. δ :
᎐
H
1.88 (dd, 1HB, J 6.9 and 13.9), 2.55 (dd, 1HD, J 7.4 and 12.2),
2.70 (dd, 1HA, J 6.9 and 13.9), 2.71 (s, 3H, CH3), 3.00 (dd,
1HC, J 9.7 and 12.2), 4.46 (dd, 1HE, J 7.4 and 9.7), 4.60 (t,
1HX, J 6.9), 6.8–7.5 (m, 15H); selected δC: 28.6 (CH3), 41.8
(C4), 48.2 (C9), 62.3 (C3), 71.0 (C8), 84.1 (C5), 172.8 (C᎐O).
᎐
Diastereoisomer 8Јm: 5%, mp: 209 ЊC. δH: 1.95 (dd, 1HD, J 7.6
and 14.0), 2.28 (dd, 1HC, J 8.0 and 14.0), 2.45 (dd, 1HB, J 6.0
and 13.1), 2.51 (s, 3H, CH3), 2.65 (dd, 1HA, J 10.0 and 13.1),
4.52 (dd, 1HE, J 7.6 and 8.0), 4.90 (dd, 1HX, J 6.0 and 10.0),
6.4–7.4 (m, 15H). νmax(KBr)/cmϪ1: 1700 (Found: C, 78.2; H,
6.4; N, 7.3. C25H24N2O2 requires C, 78.10; H, 6.29; N, 7.29%).
All NOESY experiments for the latter adducts are summarised
in Table 5.
Reaction of 1-benzyl-3-methylenepiperidin-2-one 3b with
ꢀ-benzoylnitrone 6a (Table 3, entry 16)
The equimolecular mixture of 3b and 6a (5 mmol) was dis-
solved in ethyl acetate (5 cm3); hydroquinone (10 mg) was then
added and the solution was stirred for 2.5 hours. The solvent
was evaporated and the residue purified by chromatography on
silica gel (eluent: cyclohexane–ethyl acetate, 70:30), giving 7p as
a very viscous oil which crystallised at or below Ϫ10 ЊC. Yield:
90%. δH: 1.68–2.33 (m, 4H), 2.31 (dd, 1HB, J 8.3 and 11.9), 3.20
(m, 2H), 3.36 (dd, 1HA, J 8.3 and 11.9), 4.15 (d, 1H, J 14.5),
4.71 (d, 1H, J 14.5), 5.48 (t, 1HX, J 8.3), 6.9–8.2 (m, 15H).
δC: 19.7 (C4), 32.7 (C5), 43.8 (CH2, benzyl), 46.9 (C3), 50.4 (C10),
71.4 (C9), 81.5 (C6), 113.2–149.9 (aromatic C), 178.5 (C1), 196.8
(C᎐O, benzoyl). νmax(KBr)/cmϪ1: 1690 (C᎐O benzoyl), 1660
Cycloaddition of the substituted piperazine-2,5-dione 4 with ꢀ-
benzoyl-N-phenylnitrone 6a: preparation of 3,11-dibenzoyl-6,13-
dimethyl-7,14-dioxo-2,10-diphenyl-1,9-dioxa-2,6,10,13-tetra-
azadispiro[4.2.4.2]tetradecane 9:10 (Scheme 4)
A solution of the bislactam 4 (420 mg, 2.53 mmol) and nitrone
6a (1.14 g, 5.06 mmol) in toluene (5 cm3) was stirred at 80 ЊC for
4 hours, the reaction being monitored by TLC (ethyl acetate–
cyclohexane, 20:80). The mixture assumed an orange–red
colour and a solid formed within 15 minutes. When TLC indi-
cated complete conversion of reagents, the solvent was evapor-
ated under reduced pressure, giving a red gum which was stirred
with 95% ethanol for several hours until precipitation of
adducts was complete. Filtration and washing with ethanol
gave a mixture of the two bisadducts 9:10. Fractional crystal-
lisation from propan-1-ol gave pure 9; it was unfortunately
impossible to isolate the second diastereoisomer 10 by column
chromatography, owing to its total insolubility in common
solvents. Total yield: 80%. Major diastereoisomer 9: 46%; mp:
205–206 ЊC. δH: 3.07 (dd, 1HB, J 10.1 and 13.3), 3.35 (s, 3H,
CH3), 3.41 (dd, 1HA, J 7.3 and 13.3), 5.53 (dd, 1HX, J 7.3 and
10.1), 7.0–8.0 (m, 20H). δC: 29.5 (CH3), 43.9 (C4), 71.4 (C3),
90.8 (C5), 119.2–148.6 (aromatic C), 165.1 (C᎐O), 194.3 (C᎐O
᎐
᎐
(C᎐O lactam) (Found: C, 76.2; H, 6.2; N, 6.8. C H N O
᎐
27 26
2
3
requires C, 76.03; H, 6.14; N, 6.57%).
Reaction of 1-methyl-3-methylene-5-phenylpyrrolidin-2-one 2b
with nitrones 6a–b (Table 3, entries 12, 13)
An equimolecular mixture of methylenelactam 2b and nitrone 6
(2.5 mmol) was stirred in boiling toluene (5 cm3) during the
appropriate reaction time. After evaporation of the solvent, the
residue was purified by column chromatography on silica gel.
8-Benzoyl-3,7-diphenyl-2-methyl-1-oxo-6-oxa-2,7-diazaspiro-
[4.4]nonane 7l:7Јl. The crude residue was dissolved in DMF
(2 cm3) and chromatographed on 120 g of silica gel (eluent:
cyclohexane–ethyl acetate 70:30). Total yield: 86%. Major dia-
stereoisomer 7l: 60%; mp: 167–169 ЊC. δH: 2.05 (dd, 1HD, J 7.1
and 14.4), 2.65 (dd, 1HB, J 5.6 and 12.3), 2.70 (s, 3H, CH3), 3.00
(dd, 1HC, J 7.1 and 14.4), 3.20 (dd, 1HA, J 8.2 and 12.3), 4.70
(t, 1HE, J 7.1), 5.45 (dd, 1HX, J 5.6 and 8.2), 7.0–8.1 (m,
15H). δC: 28.6 (CH3), 40.0 (C9), 42.4 (C4), 61.4 (C3), 70.3 (C8),
᎐
᎐
benzoyl). Minor diastereoisomer 10: selected δH: 3.24 (s, 3H,
CH3), 3.60 (m, 1H), 5.61 (m, 1HX). νmax(KBr)/cmϪ1: 1670
(Found: C, 69.8; H, 5.3; N, 8.8. C36H32N4O6 requires C, 70.12;
H, 5.23; N, 9.09%).
Evidence of the formation of monoadducts: A solution of
the bislactam 4 (165 mg, 1.0 mmol) and nitrone 6a (225 mg,
1.0 mmol) in toluene (5 cm3) was stirred at 80 ЊC for 1 hour,
until TLC (cyclohexane–ethyl acetate, 80:20) indicated com-
plete reaction of the dipole. The solvent was evaporated
under reduced pressure (20 ЊC) and the crude red residue
analysed by 1H NMR spectroscopy. The relative molar propor-
tions of unreacted bislactam 4 (40%), monoadducts (30%)
and bisadducts (30%) were determined with an average
accuracy of 5% from the signals at 5.4–5.6 ppm (HX protons
of mono and bisadducts) and 5.8–5.9 ppm (unshielded
methylenic protons of 4 and monoadducts; these latter signals,
located at 4.91 and 5.81 ppm, and the ones of bislactam
4—4.99 and 5.88 ppm—were sufficiently separated to allow a
correct estimation).
84.6 (C5), 116.0–149.6 (aromatic C), 195.7 (C᎐O benzoyl).
᎐
Minor diastereoisomer 7Јl: 26%, mp: 148–149 ЊC. δH: 2.50 (dd,
1HD, J 3.9 and 14.4), 2.70 (dd, 1HB, J 5.5 and 12.2), 2.75 (s, 3H,
CH3), 2.80 (dd, 1HC, J 8.3 and 14.4), 3.12 (dd, 1HB, J 8.2 and
12.2), 4.43 (dd, 1HE, J 3.9 and 8.3), 5.45 (dd, 1HX, J 5.5 and
8.2), 7.0–8.1 (m, 15H). δC: 28.6 (CH3), 40.6 (C9), 41.0 (C4), 60.9
(C3), 69.8 (C8), 84.6 (C5), 115.8–150.6 (aromatic C), 195.7 (C᎐O
᎐
benzoyl). νmax(KBr)/cmϪ1: 1710 (C᎐O γ-lactam), 1690 (C᎐O
᎐
᎐
benzoyl) (Found: C, 75.4; H, 5.6; N, 6.8. C26H24N2O3 requires
C, 75.71; H, 5.86; N, 6.79%).
2-Methyl-1-oxo-3,7,8-triphenyl-6-oxa-2,7-diazaspiro[4.4]-
nonane 7m:7Јm:8m:8Јm. The crude reaction mixture was
dissolved in toluene (2 cm3) and chromatographed on 120 g
1102
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103

View Online
Crystallographic study of bis-adduct 9:†
A colourless
References
monocrystal with dimensions 0.240 × 0.120 × 0.060 mm was
obtained by slow evaporation of a solution of 9 in 95% ethanol
and mounted on an Enraf Nonius CAD4 diffractometer.
Radiation used: graphite filtered Cu-Kα, λ = 1.54180 Å;
µ = 0.745. Data were collected at 293 2 ЊC. All scanning used
the θ/2θ technique with 4.74Њ < θ < 67.89Њ. The structure was
solved by direct methods using SHELXL93.36a Atomic scatter-
ing factors were taken from International Tables of X-Ray
Crystallography.37 Crystallographic data of 9: C36H32N4O6,
M = 616.66 g molϪ1, dcalc = 1.326 g cmϪ3, triclinic system, P-1
space group, a = 8.932(2), b = 9.670(3), c = 10.105(30) Å, α =
100.95(2)Њ, β = 105.61(2)Њ, γ = 106.37(2)Њ, V = 772.4(3) Å3,
Z = 1. Number of independent reflections: 2797 (R(int) =
0.0327); they were 2080 reflections with I > 2σ(I). The refine-
ment36b converged to R1(F) = 0.050, wR2 = 0.1087 with R1 =
Σ (Fobs Ϫ Fcalc)/Σ (Fobs). Residual electron density between
1 K. C. Liu and R. K. Howe, J. Org. Chem., 1983, 48, 4590.
2 R. K. Howe, B. R. Shelton and K. C. Liu, J. Org. Chem., 1985, 50,
903.
3 R. K. Howe and B. R. Shelton, J. Org. Chem., 1990, 55, 4603.
4 M. C. Aversa, G. Cum, G. Stagno d’Alcontres and N. Uccella,
J. Chem. Soc., Perkin Trans. 1, 1972, 222.
5 A. Kerbal, J. Vebrel, M. Roche and B. Laude, Tetrahedron Lett.,
1990, 31, 4145.
6 M. Msaddek, M. Rammah, G. Schmitt and J. Vebrel, Bull. Soc.
Chim. Belg., 1992, 101, 323.
7 M. Msaddek, M. Rammah, K. Ciamala, J. Vebrel and B. Laude,
Synthesis, 1997, 1495.
8 R. K. Howe and K. C. Liu, USP 4 364 768/1980; Chem. Abstr., 1981,
94, 47311h.
9 S. Rigolet, P. Goncalo, J. M. Melot and J. Vebrel, J. Chem. Res. (M),
1998, 2813.
10 N. Langlois, N. Van Bac, N. Dahuron, J. M. Delcroix, A. Deyine,
D. Griffart-Brunet, A. Chiaroni and C. Riche, Tetrahedron, 1995,
51, 3571.
Ϫ0.216 and 0.189 e ÅϪ3
.
11 N. Langlois, D. Griffart-Brunet, N. Van Bac, A. Chiaroni and
C. Riche, C.R. Seances Acad. Sci., 1995, 320, 155.
12 J. H. Bailey, D. T. Cherry, K. H. Crapnell, M. G. Moloney and
S. B. Shim, Tetrahedron, 1997, 53, 11731.
Reaction of 1-methyl-5-methylenepyrrolidin-2-one 5 with nitrone
6a (Scheme 5)
13 P. Mitcuch, L. Fisera, V. Ondrus and P. Ertl, Molecules, 1997, 2,
57.
14 P. Oravec, L. Fisera, I. Goljer and P. Ertl, Monatsh. Chem., 1991,
122, 977.
15 P. Oravec, L. Fisera, P. Ertl and D. Vegh, Monatsh. Chem., 1991,
122, 821.
16 J. P. Bouillon, Z. Janouzek, H. G. Viehe, B. Tinant and J. P.
Declercq, J. Chem. Soc., Perkin Trans. 1, 1996, 1853.
17 A. Strauss and H. H. Otto, Helv. Chim. Acta, 1997, 80, 1823.
18 C. Wedler, B. Costisella and H. Schick, J. Prakt. Chem., 1990, 332,
557.
Freshly prepared 1-methyl-5-methylenepyrrolidin-2-one 5 (2.5
mmol), α-benzoylnitrone 6a (2.5 mmol) and hydroquinone (15
mg) were stirred under argon at room temperature in ethyl
acetate (5 cm3) for 30 hours. TLC (CHCl3) showed complete
reaction of 6a. After evaporation of the solvent (25 ЊC, 15
1
mmHg), the oily residue was analysed by H NMR spectro-
scopy and 2D NOESY experiments. The adducts were not
separable but their admixture was crystallised from 95%
ethanol (3 cm3) and washed with the same solvent (5 cm3).
19 R. Huisgen, H. Hauck, R. Grashey and H. Seidl, Chem. Ber., 1968,
101, 2568.
20 H. Harikawa, T. Nishitani, T. Iwasaki and I. Inoue, Tetrahedron
Lett., 1979, 20, 2193.
21 J. S. Splitter, T. M. Su, H. Ono and M. Calvin, J. Am. Chem. Soc.,
1971, 93, 4075.
22 W. B. Jennings, D. R. Boyd and L. C. Waring, J. Chem. Soc., Perkin
Trans. 1, 1976, 610.
23 A. Chiaroni, C. Riche, S. Rigolet, J. M. Melot and J. Vebrel, Acta
Crystallogr., Sect. C, 2000, in the press.
24 K. B. G. Torssell, Nitrile Oxides, Nitrones and Nitronates in Org-
anic Synthesis, VCH Verlagsgesellschaft, Weinheim, 1988, p. 36 and
pp. 42–47.
25 F. Krönke and E. Börner, Chem. Ber., 1936, 69, 2006.
26 S. I. Murahashi, M. Mitsui, T. Shiota, T. Tsuda and S. Watanabe,
J. Org. Chem., 1990, 55, 1736.
3-Benzoyl-7-oxo-2-phenyl-1-oxa-2,6-diazaspiro[4.4]nonane
11:12. Yield: 50%; Major diastereoisomer 11: δH (C6D6): 1.17
(m, 1H), 1.83 (m, 3H), 2.25 (m, 1H), 2.66 (dd, 1H, J 7.5 and
13.3), 2.93 (s, 3H, CH3), 4.62 (dd, 1HX, J 7.5 and 8.6), 6.8–8.4
(m, 10H). δC: 28.7 (C9), 32.4 (C8), 40.5 (C4), 69.1 (C3), 100.2
(C5), 114.9–149.5 (aromatic C), 174.9 (C᎐O, lactam), 195.8
᎐
(C᎐O, benzoyl). Minor diastereoisomer 12: δ (C6D6): 1.17 (m,
᎐
H
1H), 1.95 (m, 2H), 2.25 (m, 3H), 2.41 (s, 3H, CH3), 4.65 (m,
1HX), 6.8–8.4 (m, 10H). δC: 25.5 (C9), 32.9 (C8), 39.8 (C4), 69.1
(C3), 100.2 (C5), 114.9–149.5 (aromatic C), 174.9 (C᎐O, lactam),
᎐
195.6 (C᎐O, benzoyl). ν
(KBr)/cmϪ1: 1720 (C᎐O lactam),
᎐
᎐
max
1690 (C᎐O benzoyl) (Found: C, 71.5; H, 6.0; N, 8.2. C H -
᎐
20 20
N2O3 requires C, 71.41; H, 5.99; N, 8.33%).
27 K. S. Chan, M. L. Yeung, W. K. Chan, R. J. Wang and T. F. C. Muk,
J. Org. Chem., 1995, 60, 1741.
28 J. M. Cassady, G. A. Howe, J. M. Robinson and I. K. Stamos, Org.
Synth., 1983, 61, 77.
Attempted reaction of 1-methyl-5-methylenepyrrolidin-2-one 5
with nitrone 6b
29 M. Ueda and H. Mori, J. Polym. Sci. Part A: Polym. Chem., 1990,
20, 2597.
30 G. M. Ksander, J. E. McMurry and M. Johnson, J. Org. Chem.,
1977, 42, 1180.
31 M. Ueda, M. Takahashi and T. Suzuki, J. Polym. Sci., Polym. Phys.
Ed., 1983, 20, 1139.
32 J. Villiéras and M. Rambaud, Synthesis, 1982, 924.
33 N. El Alami, These de Doctorat, Université de Nantes, 1987.
34 N. El Alami, C. Belaud and J. Villiéras, Tetrahedron Lett., 1987, 28,
59.
Freshly distilled 1-methyl-5-methylenepyrrolidin-2-one 5 (3
mmol), α,N-diphenylnitrone 6b (2.5 mmol) and hydroquinone
(15 mg) were stirred at room temperature in ethyl acetate (5
cm3) for 30 hours under an argon atmosphere. A reddish oily
residue was obtained after evaporation of the solvent (30 ЊC, 15
mmHg). Its 1H NMR analysis showed complete disappearance
of 5 and the absence of any HX proton corresponding to the
expected spiroadduct. Addition of 95% ethanol (5 cm3) and
cooling to Ϫ10 ЊC for several hours gave a white solid, mp:
114–115 ЊC, which had characteristics identical to 6b (90%
recovery).
35 D. L. Lee, C. J. Morrow and H. Rapoport, J. Org. Chem., 1974, 39,
893.
36 G. M. Sheldrick, a) SHELXL86, Program for Crystal Structure
Solution, 1986; b) SHELXL93, Program for Crystal Structure
Refinement, University of Göttingen, Germany, 1993.
37 International Tables for X-Ray Crystallography, ed., A. C. Wilson,
Kluwer Academic Publishers, Dordrecht, 1994.
Acknowledgements
S. R. and J. M. M. are grateful to Drs G. Crini and G. Moine for
the realisation of 2D 1H–1H NMR spectra and to Mlle P. Banet
for technical assistance.
† CCDC reference number 207/400. See http://www.rsc.org/suppdata/
p1/b0/b000129p for crystallographic files in .cif format.
Paper b000129p
J. Chem. Soc., Perkin Trans. 1, 2000, 1095–1103
1103

View Online