Radical cyclisation of epoxynitrile-2-azetidinones mediated by Cp2TiCl

Article abstract of DOI:10.1016/j.tet.2007.01.049

The reductive radical cyclisation of δ- and ε-epoxynitrile-2-azetidinones has been achieved using titanocene monochloride. The reaction was regioselective and afforded bicyclic β-lactams and tricyclic β-lactams containing an aryl group fused to a seven-me

Full text of DOI:10.1016/j.tet.2007.01.049

Tetrahedron 63 (2007) 3017–3025
Radical cyclisation of epoxynitrile-2-azetidinones
mediated by Cp2TiCl
*
Laura M. Monleon, Manuel Grande and Josefa Anaya
ꢀ
Departamento de Quꢀımica Orgaꢀnica, Facultad de Ciencias Quꢀımicas, Universidad de Salamanca, E-37008 Salamanca, Spain
Received 18 December 2006; revised 23 January 2007; accepted 25 January 2007
Available online 30 January 2007
Abstract—The reductive radical cyclisation of d- and 3-epoxynitrile-2-azetidinones has been achieved using titanocene monochloride. The
reaction was regioselective and afforded bicyclic b-lactams and tricyclic b-lactams containing an aryl group fused to a seven-membered ring.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
N
O
N
The increasing resistance of bacteria to classical b-lactam
antibiotics1 is well documented and this has provoked a
growing interest in the synthesis of new b-lactams able to
supersede the destructive action of b-lactamases.2
O
benzocarbapenem
benzocarbacephem
Figure 1. Structures of b-lactamase inactivators.
An approach to deactivate these enzymes consists in modify-
ing the structure of classical b-lactam antibiotics, trying to
make them insensitive to the b-lactamase attack. An alterna-
tive to avoid the enzymatic destruction of the antibiotics uses
a substance that disables the b-lactamase in synergy with
a b-lactam antibiotic. In this context, benzocarbapenems
and benzocarbacephems have been designed as inactivators
of b-lactamases (Fig. 1).
products after hydrolysis could be bi- and tricyclic hydroxy-
keto-b-lactams. 8
2. Results and discussion
The synthesis of d- and 3-epoxynitrile-2-azetidinones 1–4
required for our study is illustrated in Scheme 1. The starting
precursors, b-lactams 5–7, were prepared by ketene–imine
cycloaddition9 between methoxyacetyl chloride in the pres-
ence of TEA and the imines obtained from trans-cinnamal-
dehyde and 2,2-dimethylaminoethanol or o-cyanoaniline.
The Staudinger reaction carried out with 2,2-dimethyletha-
nolamine afforded the racemic cis-2-azetidinone 5 in 77%
yield. In contrast, the use of o-cyanoaniline as amine af-
forded in 91% yield a 2:3 cis/trans isomeric mixture from
which the pure b-lactams 6 and 7 could be isolated by chro-
matography and crystallisation.10
On the other hand, the interest in free radical reactions ap-
plied to synthetic problems continues to increase and these
reactions have successfully been used for growing a number
of synthetic targets, including the synthesis of five- and six-
membered carbocyclic and heterocyclic compounds.3
In connection to our current research interest in the prep-
aration4 and biological activities5 of b-lactams, we have
recently reported a series of radical cyclisations on epoxy-
olefin- and epoxyaldehyde-2-azetidinones, mediated by
titanocene monochloride,6 to afford chiral bi- and tricyclic
b-lactams.7 In this context, we report here the radical cycli-
sation of d- and 3-epoxynitrile-2-azetidinones as a new route
to the synthesis of new bi- and tricyclic b-lactams. The n-exo
cyclisation process is based on the homolytic cleavage of an
oxiranyl ring with titanocene monochloride followed by
intramolecular addition to the cyano group. The resulting
1
The H NMR coupling constants between H-3 and H-4
shown by these 2-azetidinones clearly established the rela-
tive cis configuration for compounds 5 and 6 (Jꢀ4.4) and
the trans configuration for 7 (Jꢀ1.8).5,11
Compound 5, by desilylation and Swern oxidation gave the
aldehyde 8 in 71% yield, which was transformed into the ho-
mologue aldehyde 9 in 45% yield through a Wittig reaction.
Different pathways were examined to prepare the nitriles 10
* Corresponding author. Tel.: +34 923 294482; fax: +34 923 294574;
e-mail: mgrande@usal.es
0040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.01.049

3018
L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
H
H
N
H
H
N
H
H
N
Ph
Ph
Ph
MeO
O
MeO
MeO
O
H2N
g
a, b, c
d, e
OH
nCHO
nCN
OTBS
O
f
5
10: n= 0
11: n= 1
8: n= 0
9: n= 1
h
O
NH2
Ph
Ph
CN
H
H
H
H
MeO
MeO
CN
5
6 CN
H
H
N
O
3
4
3
4
N
a, c
h
Ph
MeO
O
N
O
O
1
nCN
1a–4a: 5 6α-epoxy
1b–4b: 5 6β-epoxy
3a,b: 3,4-cis
4a,b: 3,4-trans
1a,b: n= 0
2a,b: n= 1
6: 3,4-cis
7: 3,4-trans
˚
˚
Scheme 1. Reagents and conditions: (a) trans-cinnamaldehyde, CH2Cl2, molecular sieve 4 A, reflux, for 5; trans-cinnamaldehyde, toluene, molecular sieve 4 A,
reflux, for 6 and 7; (b) tBu(Me)2SiCl, pyr/DMAP, rt; (c) MeOCH2COCl, TEA, rt; (d) HCl (0.1 M), MeOH, rt; (e) Swern oxidation; (f) Ph3P(Cl)CH2OMe, nBuLi,
THF, ꢁ15 ꢂC; then HClO4 (30%), rt; (g) Me2N-NH2, MeOH, rt; then MMPP, 0 ꢂC; (h) m-CPBA, NaHCO3, CH2Cl2, rt.
and 11, and we found that the best yield (80%) was obtained
by treatment of the aldehydes 8 and 9 with N,N-dimethyl-
hydrazine, followed by oxidation with MMPP.12
1 and 2 and the (5b,6b-epoxy) configuration for the more
polar isomers.
The stereochemistry depicted in Scheme 1 for the 5,6-ep-
oxy-b-lactams 3 and 4 was deduced from the structures
proposed for their cyclisation products 14a and 15a/15b, re-
spectively (Table 2). The stereoselective formation of these
homobenzocarbacephems (see below) suggests that the cyc-
lisation could be a process analogous to those observed in
our previous studies.7 Hence, the epoxides 3a and 4a, pre-
cursors of 14a and 15a (C6-aOH), respectively, should have
the relative configuration 5a,6a and the epoxide 4b, the pre-
cursor of the tricyclic b-lactam 15b (C6-bOH), should have
the relative configuration 5b,6b.
The epoxidation of cyano-2-azetidinones 10 and 11 with
m-CPBA gave diastereomeric mixtures of epoxylactams
1a/1b (1:1) and 2a/2b (3:2) in 74 and 85% yield, respec-
tively. From these mixtures only the epoxide 1a could be
isolated as a pure substance by chromatography on silica
gel.13 Similar results were obtained in the epoxidation of
the cis- and trans-isomers 6 and 7. Treatment of these iso-
mers with m-CPBA gave a 3:2 diastereomeric mixture of
epoxylactams 3a/3b in 85% yield and a 1:2 diastereomeric
mixture of 4a/4b in 90% yield, from which only the epoxide
3a could be isolated as a pure substance.
Two procedures6g were checked to explore the reactivity of
the epoxynitriles 1–4 (Table 2) with titanocene monochlor-
ide. Method A: a green solution of Cp2TiCl in THF, gener-
ated in situ from Cp2TiCl2 and Zn0 at room temperature,
was slowly added to a THF solution of the epoxide. Method
B: the epoxide in THF solution was slowly added to the THF
solution of titanium-III reagent.
The C-5 and C-6 configurations depicted in Scheme 1 for the
epoxynitriles 1a,b and 2a,b were tentatively proposed by
comparison of the respective polarities (Rf) and H NMR
1
data (Table 1) with those of the chiral epoxy-2-azetidinones
Ia and Ib whose absolute configurations have been previ-
ously assigned as (5a,6a-epoxy) and (5b,6b-epoxy), respec-
tively.7b From these data, it emerges that the less polar
isomers of each pair of epoxides 1, 2 and I show that the hy-
drogen atoms H-4 and H-6 are slightly unshielded and the
proton H-5 is slightly shielded in comparison with those of
the more polar isomers. Consequently, we propose the
(5a,6a-epoxy) configuration for the less polar epoxynitriles
We first examined the 6-exo- and 7-exo-radical cyclisation
process in the aliphatic d- and 3-epoxynitrile-2-azetidinones
1 and 2 (Table 2), assuming that the reductive opening of ep-
oxide ring with titanocene monochloride should firstly cause
the formation of the C-6 benzylic radical intermediate. Thus,
the addition of titanocene monochloride solution on the
epoxynitrile-2-azetidinone 1a solution (Method A) followed
by acidic work-up proceeded with total regioselectivity to
give the cyclisation product, the bicyclic b-lactam 12a, in
60% yield (entry 1, Table 2).
Table 1. Selected spectral data of epoxides 1, 2 and I
Compound (Rf)
1H NMR
H-5
H-4
H-6
O
Ph
H
H
Under the same conditions the reaction was carried out in a
4:5 isomeric mixture of 1a/1b to give in 60% yield, a 3:4 mix-
ture of D4,5-carbacephem 12a and the starting epoxide 1b
(entry 2, Table 2). Similar results were obtained when a 3:2
isomeric mixture of the epoxides 2a/2b was used as the start-
ing material. In this case a 1:2:7 mixture of the compounds
13a/2a/2b was obtained in 83% yield (entry 4, Table 2).
MeO
5
6
3
4
N
1a (0.48)a
1b (0.46)a
2a (0.23)b
2b (0.22)b
Ia (0.30)c
Ib (0.25)c
3.72
3.59
3.78
3.58
3.52
3.51
3.20
3.26
3.20
3.22
3.28
3.88
3.95
3.80
3.81
3.80
3.76
3.72
CO2Me
O
O
O
O
O
Ia: 5 6 -epoxy
Ib: 5 6 -epoxy
Under these conditions it seems that 5b,6b-epoxides 1b and
2b do not cyclise, so we decided to repeat the same reactions
in the isomeric mixtures of the epoxides 1 and 2 but
a
Benzene/ethyl acetate 8:2.
Benzene/ethyl acetate 9:1.
Hexanes/ethyl acetate 7:3.
b
c

L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
3019
Table 2. Reaction of epoxides 1–4 with Cp2TiCl
Entry
Epoxide
Method (Time, h)
Products (yield, %)a
Bilactam Alkene
Epoxide (%)b
Tribactam
1
2
3
4
5
6
7
8
9
10
1a
A (3)
A (3)
B (3)
A (1)
B (3)
A (2)
A (3)
B (5)
A (2)
B (1)
—
1b (33)
1b (25)
12a (60)
12a (27)
12a (19)
13a (10)c
13a (8)c
—
—
1a/1b (4:5)
1a/1b (4:5)
2a/2b (3:2)
2a/2b (3:2)
3a
3a/3b (1:1)
3a/3b (1:1)
4a/4b (1:2)
4a/4b (1:2)
—
—
—
—
10 (13)
—
2a (16)c, 2b (57)c
2a (12)c, 2b (55)c
—
—
—
4a (5)
—
11 (12)c
—
14a (56)
6 (25)
6 (26)
—
14a (35)
14a (24)
15a (17)
15a (11)
—
—
15b (61)
15b (34)
7 (22)
OH
OH
OH
Ph
Ph
Ph
Ph
O
6
H
H
N
H
H
N
H
H
H
H
5
5
H
H
MeO
O
MeO
MeO
MeO
O
4
Ph
O
MeO
6
5
7
7
8
8
9
6
1N
7
4
4
O
O
O
9
8
3
2
3
1N
1N
3
2
2
O
O
O
12a
13a
14a
15a
15b
a
Isolated yield after column chromatography.
Recovered material.
b
c
Calculated yield from GC/MS and 1H NMR.
reversing the order of addition of the reagents (Method B).
Unfortunately, the 6-exo- and 7-exo-radical processes were
not observed and lower yields for bicyclic b-lactams 12a
and 13a were obtained in favour of the alkenes 10 and 11
(entries 3 and 5, Table 2).
The hydroxyl IR absorption bands as well as the presence of
four Csp3ꢁH methyne signals in NMR spectra are in agree-
ment with the structures depicted in Table 2 for the tricyclic
b-lactams 14a, 15a and 15b. The stereochemistry of the ben-
zofused tricyclic b-lactams 14a and 15a was deduced from
the coupling constants of H-6 with the vicinal protons H-5
and H-7 (J5,6¼7.5 Hz, J6,7¼9.8 Hz in 14a; J5,6¼8.4 Hz,
J6,7¼4.3 Hz in 15a). These coupling constants are character-
istic of an anti-arrangement between these hydrogen atoms
for compound 14a and an anti/syn-arrangement for com-
pound 15a. The configuration of the tricyclic b-lactam 15b
was deduced from the coupling constant between H-6 and
H-7 (J6,7¼9.4 Hz) as well as from NOE-difference spectrum
data. Irradiation on the signal at d¼4.25 ppm (m, H-5/H-6)
in compound 15b resulted in a 2.9% increment of the signal
at d¼4.06 ppm (dd, H-7) and a 2.1% increment of the signal
at d¼4.66 ppm (d, H-8). These data suggest a relative syn-
relationship between H-6/H-8 and between H-5/H-7 and
are consistent with the trans-arrangement of the H-5, H-6
and H-7 hydrogen atoms attached to the seven-membered
ring. These configurational assignments are also supported
by the dissimilar chemical shifts displayed by H-8 and C-8
The above results prove that a 7-exo-radical cyclisation of
3-epoxynitrile-2-azetidinones promoted by Cp2TiCl is pos-
sible and new tricyclic b-lactams could be synthesised.
Thus, expecting that a planar geometry for the radical accep-
tors could be favourable for these cyclisation processes, we
explored the reactivity of the benzoepoxynitriles 3 and 4.
The reaction of pure cis-isomer 3a with titanocene mono-
chloride by the Method A, proceeded with total regio- and
stereoselectivity to give the desired cyclisation product, the
homobenzocarbacephem 14a in 56% yield (entry 6, Table
2). But the same reaction carried out in a 1:1 mixture of
the cis-isomers 3a/3b afforded in 60% yield, a 5:7 mixture
of elimination and cyclisation products 6/14a (entry 7, Table
2). Different results were obtained when a 1:2 mixture of the
trans-epoxynitriles 4a/4b was used as the starting material.
In this case, the homobenzocarbacephems 15a and 15b
were obtained in 17% and 61% yield, respectively, and a
small amount of the epoxide 4a was also isolated (entry 9,
Table 2). Also in these cases the reverse addition (Method
B) does not improve the results. We found that the isomer
3b does not cyclise and lower yields for homobenzocarb-
acephems 14a, 15a and 15b were obtained in favour of the
elimination products 6 and 7 (entries 8 and 10, Table 2).
in 15a (dH-8, 5.17 ppm; dC-8, 84.0 ppm) and 15b (dH-8
,
4.66 ppm; dC-8, 87.3 ppm). The nearly syn-diaxial arrange-
ment of H-8 and the hydroxyl group in 15a (dihedral angle
8
6
H–C –C –OH¼ꢁ16ꢂ, dH–O¼2.4 A) justifies the unshielding
of H-8 (Dd +0.51 ppm) in this isomer with respect to the
chemical shift in 15b (dihedral angle H–C8–C6–OH¼50ꢂ,
˚
˚
dH–O¼2.7 A, Fig. 2). Also, the g-shielding effect of the
C-6 hydroxyl group on C-8 is more effective in 15a (dihedral
angle C8–C7–C6–OH¼ꢁ23ꢂ) than in 15b (dihedral angle
C8–C7–C6–OH¼55ꢂ), in agreement with the observed
chemical shifts (Dd ꢁ3.3 ppm).14
The reaction products 12–15 were characterised by IR and
NMR spectroscopies and FABHRMS analysis. These com-
pounds show arylketone IR absorption bands while the IR
cyanide group absorption bands and the NMR oxirane pro-
ton signals were absent. The NMR spectra of bicyclic b-lac-
tams 12a and 13a are very similar, as both show signals for
three methyl groups, three methynes (two Csp3 and one Csp2 ),
and a phenyl group. The main difference is the presence of
extra methylene group signal for the carbacephem 13a.
The evolution of the epoxy b-lactams 1–4 on reaction with
Cp2TiCl could be explained as depicted in Scheme 2. The
benzyl radicals cis/trans-I generated by homolytic cleavage
of the oxirane ring can progress through two different path-
ways: (a) reduction to the benzyl anion followed by b-elim-
ination of the titanium-oxo moiety to give the alkenes 6, 7,

3020
L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
In order to evaluate the b-lactamase activity of these new
bicyclic and tricyclic b-lactams, we also have prepared the
unsaturated benzofused tricyclic b-lactams 16 and 17 by de-
hydration of the tricyclic b-lactams 14 and 15, respectively
(Scheme 3).
OH
Ph
Ph
H
H
H
H
N
MeO
O
MeO
O
6
5
7
8
a
4
O
O
9
N
3
2
1
14a: 7,8-cis, 5βPh, 6αOH
15a: 7,8-trans, 5βPh, 6αOH
15b: 7,8-trans, 5αPh, 6βOH
16: 7,8-cis
17: 7,8-trans
Scheme 3. Reagents and conditions: (a) p-MeC6H4SO2Cl, TEA, DMAP,
CH2Cl2, rt.
Figure 2. Energy minimised stereoscopic models for compounds 15a and
15b.15
The reaction of the pure compound 14a with p-toluenesul-
fonyl chloride in CH2Cl2 in the presence of TEA at room
temperature, afforded in 36% yield the cis-homobenzo-
carbacephem 16. The same reaction using a 1:3 mixture of
compounds 15a/15b as starting material, gave in 35% yield
the trans-homobenzocarbacephem 17.
10 or 11 and (b) radical trapping by the cyanide group to
give the intermediate cis/trans-II, which progress after acid
work-up to the desired homocarbacephems 14a, 15a or 15b
and, through subsequent dehydration, to the bicyclic b-lac-
tams 12a or 13a.
The presence of the conjugated ketone in compounds 16 and
17 was evident from their IR and NMR spectral data, and also
the molecular ions observed in their FABHRMS spectra
([M++23]¼328.0926 for 16 and 328.0936 for 17b) support
the structures depicted for these compounds in Scheme 3.
TiIV
O
H
H
N
Ph
MeO
a
a
6, 7, 10, 11
N
b
The present study shows that the d- and 3-epoxynitrile radi-
cal cyclisations mediated by titanocene monochloride are
stereocontrolled processes and can be applied to prepare
new polycyclic b-lactams.
n
O
cis/trans–Ia,b
b
12a, 13a
As far as we know, the 7-exo-radical cyclisation of 3-epoxy-
nitriles mediated by titanocene monochloride has been
applied for the first time to benzonitrile acceptors and the
synthesised products, compounds 14 and 15, are the first ex-
amples of homobenzocarbacephems reported in the litera-
ture. Further studies to optimise the yields and expand this
approach to optically active benzofused tricyclic b-lactams
as well as several biological activity tests are in progress and
will be reported in due course.
TiIV
O
H2O
14a, 15a,b
Ph
H
H
N
MeO
H3O
N
TiIV
n
O
cis/trans–IIa,b
Scheme 2. Proposed pathways explaining the formation of compounds 6, 7
and 10–15.
The product distribution in these reactions seems to be
directed by stereoelectronic effects. The specific formation
of the bi- and tricyclic b-lactams 12–15 could be explained
if we consider that the addition of the benzyl radical to the
nitrile group is quite slow,8 thus the cyclisation process
should go through a latter TS than in the case of alkyl radi-
cals16 so that it should lead to a higher selectivity as ob-
served. The ability of the cis/trans-Ia,b radicals to cyclise
depends on the accessibility of the triple bond by the benzyl
radical and on the correct alignment of the C-6 sꢃ and the
nitrile p* orbitals. An analysis of the molecular models let
us to verify that the cis/trans-Ia,b radicals can come close
3. Experimental section
3.1. General methods
Flash chromatographies were run on silica gel (Merck 60,
230–400 mesh) and thin layer chromatographies (TLC) on
commercial silica gel plates (Merck F254). Mass spectra
(MS) were recorded on a VG TS-250 spectrometer: EI at
70 eV; FAB with xenon as ionisation gas; HRMS with m-ni-
trobenzyl alcohol matrix and 10 keV acceleration potential.
IR spectra were recorded as neat film on a Bomem MB-
100 instrument. 1H and 13C NMR spectra were obtained on
Bruker instruments WP200SY and Advance 400DRX (200
and 400 MHz, respectively) in CDCl3 solutions with tetra-
methylsilane as internal standard. Solvents and reagents were
purified according to standard techniques.
ꢃ
˚
to the cyanide group (<4 A), but the orientation of the s
and p* orbitals seems to be best arranged for cyclisation
in isomers 1a, 2a, 3a and 4b, in agreement with the experi-
mental results.

L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
3021
3.2. Synthesis of mono-b-lactams 5, 6 and 7
13C NMR (100 MHz): d 58.9, 62.8, 85.6, 102.8, 117.2,
123.2, 125.3, 125.6, 126.8, 128.6, 133.8, 134.1, 135.3,
136.6, 137.5, 164.4; FABHRMS calcd for C19H16N2O2Na
(M++23): 327.1104; found: 327.1097.
Method A: a mixture of trans-cinnamaldehyde (10 mmol)
and 2,2-dimethylaminoethanol (10 mmol) in CH2Cl2
ꢂ
˚
(100 mL) with 4 A molecular sieve was heated at 50 C until
the starting material disappeared in TLC. Then, the mixture
was filtered through CeliteÒ and the solvent was removed
under reduced pressure. To a cooled (0 ꢂC) solution of the
imine and catalytic amount of DMAP in anhydrous pyridine
(25.0 mL), tert-butyldimethylsilyl chloride (2.72 g, 20.0 mmol)
was added under argon. The resulting mixture was allowed
to warm to room temperature and stirred overnight. The
crude mixture was poured into a cold solution of ammonium
chloride and extracted with dichloromethane. The organic
layer was washed with brine, dried (anhydrous Na2SO4) and
concentrated under reduced pressure. To a solution of the
crude silylated imine in anhydrous CH2Cl2 (100 mL), dry
TEA (2.8 mL, 20 mmol,) and methoxyacetyl chloride
(0.11 mL, 12.0 mmol) were successively added under argon.
The reaction mixture was stirred at room temperature for 2 h
and then poured into a cold ammonium chloride solution,
neutralised to pH 7 and extracted with dichloromethane.
The organic layer was washed with brine, dried (anhydrous
Na2SO4) and the solvent removed under reduced pressure.
Flash chromatography of the residue with hexanes/ethyl ace-
tate 9:1 mixture as eluent gave compound 5 (3 g, 77%).
3.2.3. trans-1-(o-Cyanophenyl)-4-[(E)-styryl]-3-meth-
oxy-2-azetidinone (7). Colourless solid. Mp 60–61 ꢂC
(pentane/ethyl ether). IR (KBr): 2226, 1770 cmꢁ1; 1H NMR
(200 MHz, CDCl3): d 3.59 (3H, s), 4.60 (1H, d, J¼1.8 Hz),
5.25 (1H, dd, J¼1.8, 8.5 Hz), 6.20 (1H, dd, J¼8.5, 15.9 Hz),
6.90 (1H, d, J¼15.9 Hz), 7.25–7.45 (6H, m), 7.55 (1H, t,
J¼8.2 Hz), 7.60 (1H, d, J¼8.2 Hz), 7.68 (1H, d, J¼8.2 Hz);
13C NMR (100 MHz, CDCl3): d 58.2, 63.9, 89.5, 103.1,
117.2, 123.4, 125.6, 126.7, 128.7, 133.8, 135.3, 136.6, 137.9,
164.4; FABHRMS calcd for C19H16N2O2Na (M++23):
327.1104; found: 327.1108.
3.3. Synthesis of aldehydes 8 and 9
3.3.1. 1-(10,10-Dimethyl-10-formylmethyl)-4-[(E)-styryl]-
3-methoxy-2-azetidinone (8). To a solution of the silylated
b-lactam 5 (5.0 g, 12.8 mmol) in methanol (193 mL), 0.1 M
HCl (116 mL) was added. The resulting solution was stirred
at room temperature for 2 h, then neutralised with solid
NaHCO3 and methanol was evaporated under reduced pres-
sure. The resulting aqueous layer was extracted with ethyl
acetate (3ꢄ15 mL), the organic combined extracts were
washed with brine, dried (anhydrous Na2SO4) and the sol-
vent was removed under reduced pressure. The crude prod-
uct was purified by flash chromatography (hexanes/ethyl
acetate 4:1) and the resulting pure alcohol (2.72 g, 77%)
was used for the next step. To a cooled (ꢁ78 ꢂC) solution
of oxalyl chloride (1.73 mL, 20.0 mmol) in anhydrous
dichloromethane (55.2 mL), a solution of freshly distilled
DMSO (3.9 mL, 54.9 mmol) in anhydrous dichloromethane
(2.6 mL) was added dropwise under argon. The resulting
mixture was stirred for 5 min keeping the temperature at
ꢁ78 ꢂC and then a solution of the above pure alcohol (2.75 g,
10.0 mmol) in anhydrous dichloromethane (13.0 mL) was
added dropwise for a 1 h period. The reaction mixture was
stirred for 30 min at ꢁ78 ꢂC and then TEA (4.2 mL,
30 mmol) was added. The reaction was vigorously stirred
for 40 min at ꢁ78 ꢂC and then allowed to warm to 0 ꢂC,
and stirred for 2 h. The crude mixture was poured in water
and extracted with ethyl acetate. The organic extract was
washed with brine, dried (anhydrous Na2SO4) and concen-
trated under reduced pressure. Aldehyde 8 (2.6 g, 95%)
was obtained as a colourless solid after crystallisation in
hexanes/dichloromethane mixtures. Mp 90–91 ꢂC (hexanes/
dichloromethane). IR (KBr): 2840, 1755, 1717, 1660, 760,
700 cmꢁ1; 1H NMR (200 MHz): d 1.42 (s, 3H), 1.45 (s, 3H),
3.44 (s, 3H), 4.50 (dd, 1H, J¼4.7, 9.3 Hz), 4.62 (d, 1H,
J¼4.7 Hz), 6.30 (dd, 1H, J¼9.3, 16.0 Hz), 6.70 (d, 1H,
J¼16.0 Hz), 7.30–7.50 (m, 5H), 9.68 (s, 1H); 13C NMR
(50 MHz): d 20.7, 21.0, 58.5, 60.6, 63.3, 84.2, 124.7,
126.6, 128.3, 128.6, 135.8, 136.0, 166.8, 198.4; FABHRMS
calcd for C16H20NO3 (M++1): 274.1438; found: 274.1429.
Method B: a mixture of trans-cinnamaldehyde (10 mmol)
and o-cyanoaniline (10 mmol) in toluene (100 mL) was
heated under reflux in a Dean–Stark apparatus until the start-
ing material disappeared in TLC and then the solvent was
removed under reduced pressure. To a solution of the crude
imine in anhydrous CH2Cl2 (100 mL), dry TEA (2.8 mL,
20 mmol) and methoxyacetyl chloride (0.11 mL, 12.0 mmol)
were consecutively added under argon. The reaction mixture
was stirred at room temperature for 4 h and then same pro-
cedure as in Method A was followed. Flash chromatography
of the residue with hexanes/ethyl acetate 9:1 mixture as
eluent gave a 2:3 mixture of 6/7 (2.77 g, 91%). From this
mixture, by crystallisation in pentane/ethyl ether, 890 mg
of crystalline trans-2-azetidinone 7 was obtained together
with 1.8 g of filtrate. A new flash chromatography of the
filtrate with hexanes/ethyl acetate 95:5 mixture as eluent,
afforded 350 mg of pure cis-isomer 6.
3.2.1. 1-(10,10-Dimethyl-20-tert-butyldimethylsilyloxy-
ethyl)-4-[(E)-styryl]-3-methoxy-2-azetidinone (5). Colour-
1
less oil. IR (film): 1751, 1650, 1500, 760, 700 cmꢁ1; H
NMR (200 MHz): d 0.04 (s, 6H), 0.90 (s, 9H), 1.24 (s, 3H),
1.33 (s, 3H), 3.39 (s, 3H), 3.52 (d, 1H, J¼4.4 Hz), 3.67 (d,
1H, J¼4.4 Hz), 4.40–4.50 (m, 2H), 6.30 (dd, 1H, J¼8.3,
16.0 Hz), 6.70 (d, 1H, J¼16.0 Hz), 7.30–7.50 (m, 5H); 13C
NMR (50 MHz): d 18.2, 23.1, 23.4, 25.6, 58.2, 58.3, 60.7,
68.3, 83.9, 126.6, 126.8, 128.1, 128.6, 134.7, 136.4, 168.4;
FABHRMS calcd for C22H33NO3Si (M++1): 390.2459;
found: 390.2463.
3.2.2. cis-1-(o-Cyanophenyl)-4-[(E)-styryl]-3-methoxy-
2-azetidinone (6). Colourless oil. IR (film): 2225, 1766,
1493, 755 cmꢁ1; 1H NMR (400 MHz): d 3.53 (s, 3H), 4.90
(d, 1H, J¼5.0 Hz), 5.60 (dd, 1H, J¼5.0, 9.2 Hz), 6.25 (dd,
1H, J¼9.2, 16.0 Hz), 7.05 (d, 1H, J¼16.0 Hz), 7.25–7.45
(m, 7H), 7.60 (t, 1H, J¼8.2 Hz), 8.10 (d, 1H, J¼8.2 Hz);
3.3.2. 1-(10,10-Dimethyl-20-formylethyl)-4-[(E)-styryl]-3-
methoxy-2-azetidinone (9). To a stirred suspension of
methoxymethylenetriphenylphosphonium chloride (10.3 g,
30.0 mmol) in anhydrous THF (60 mL) at ꢁ15 ꢂC under
argon atmosphere, BuLi (1.6 M in hexanes, 30.0 mmol) was

3022
L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
added dropwise. After the addition was finished, the reaction
mixture was stirred for further 10 min at ꢁ15 ꢂC. Then, a so-
lution of aldehyde 8 (2.73 g, 10.00 mmol) in anhydrous THF
(6.0 mL) was slowly added, and the reaction mixture was
stirred and slowly warmed to room temperature for 2 h.
The crude mixture was poured into a cold ammonium chlo-
ride solution and extracted with ethyl acetate (3ꢄ25.0 mL).
The combined organic extracts were washed successively
with a saturated NaHCO3 solution and brine, dried over an-
hydrous Na2SO4 and concentrated under reduced pressure.
Compound 9 (1.30 g, 45%) was obtained as a colourless oil
after purification by silica gel chromatography (hexanes/
128.8, 136.0, 136.1, 166.6; FABHRMS calcd for
C17H21N2O2 (M++1): 285.1598; found: 285.1577.
3.5. General procedure for the synthesis of epoxynitrile-
b-lactams 1–4
A solution of the specific alkene (1.0 mmol), m-CPBA
(370 mg, 1.5 mmol) and NaHCO3 (126 mg, 1.5 mmol) in dry
dichloromethane (10.0 mL) was stirred under argon atmo-
sphere at room temperature until the disappearance of the
starting material was observed by 1H NMR spectra. The re-
action was then quenched with 10% v/v aqueous Na2S2O3
and the aqueous layer was separated and extracted with ethyl
acetate (3ꢄ15 mL). The combined organic extracts were
washed with a saturated solution of NaHCO3 and brine, dried
(anhydrous Na2SO4) and the solvent was evaporated under
reduced pressure.
ethyl acetate 4:1). IR (film): 2834, 1748, 760, 700 cmꢁ1
;
1H NMR (400 MHz): d 1.42 (s, 3H), 1.43 (s, 3H), 2.74
(dd, 1H, J¼2.0, 16.0 Hz), 3.07 (dd, 1H, J¼2.0, 16.0 Hz),
3.42 (s, 3H), 4.49–4.50 (m, 2H), 6.30 (dd, 1H, J¼9.4,
16.0 Hz), 6.72 (d, 1H, J¼16.0 Hz), 7.25–7.45 (m, 5H),
9.80 (t, 1H, J¼2.0 Hz); 13C NMR (100 MHz): d 25.7,
27.7, 52.8, 54.6, 58.6, 60.7, 83.7, 124.6, 126.6, 128.3,
128.7, 135.6, 136.0, 166.2, 200.2; FABHRMS calcd for
C17H22NO3 (M++1): 288.1594; found: 288.1579.
3.5.1. 1-(10,10-Dimethyl-10-cyanomethyl)-4-[(1a,2a)-
epoxy-2-phenylethyl]-3-methoxy-2-azetidinone (1a) and
1-(10,10-dimethyl-10-cyanomethyl)-4-[(1b,2b)-epoxy-2-
phenylethyl]-3-methoxy-2-azetidinone (1b). Compound
10 (230 mg, 0.9 mmol) was reacted with m-CPBA for 5
days to gave a 1:1 diastereomeric mixture of products 1a
and 1b (180 mg, 74%) after purification by flash chromato-
graphy (hexanes/ethyl acetate, 4:1). The isomer 1a (80 mg)
was isolated from enriched fractions by flash chromato-
graphy (benzene/ethyl acetate 9:1).
3.4. General procedure for the preparation of nitriles
10 and 11
To a stirred solution of the specific aldehyde (1.0 mmol) in
methanol (5.7 mL) under argon atmosphere, N,N-dimethyl-
hydrazine (0.11 mL, 1.43 mmol) was slowly added. After
the completion of addition, the reaction mixture was stirred
at room temperature until the starting material disappeared
in TLC. The reaction mixture was cooled to 0 ꢂC and a solu-
tion of magnesium monoperoxyphthalate (1.9 g, 3.0 mmol)
in methanol (4.34 mL) was added and the resulting mixture
was stirred for 10 min at 0 ꢂC. Methanol was removed under
reduced pressure and the crude residue was diluted with
dichloromethane. The organic solution was washed twice
with brine, dried (anhydrous Na2SO4) and concentrated
under reduced pressure.
3.5.1.1. Isomer 1a. Colourless oil. IR (film): 2240, 1771,
1600, 1500, 750, 700 cmꢁ1; 1H NMR (400 MHz): d 1.68 (s,
3H), 1.88 (s, 3H), 3.20 (dd, 1H, J¼2.0, 8.0 Hz), 3.63 (s, 3H),
3.72 (dd, 1H, J¼4.8, 8.0 Hz), 3.95 (d, 1H, J¼2.0 Hz),
4.64 (d, 1H, J¼4.8 Hz), 7.30–7.45 (m, 5H); 13C NMR
(100 MHz): d 25.8, 27.2, 50.6, 58.1, 58.7, 59.6, 60.3, 83.8,
119.4, 125.6, 128.6, 128.7, 135.4, 165.3; FABHRMS calcd
for C16H18N2O3Na (M++23): 309.1215; found: 309.1122.
1
3.5.1.2. Isomer 1b. From a 4:5 mixture of 1a/1b. H
3.4.1. 1-(10,10-Dimethyl-10-cyanomethyl)-4-[(E)-styryl]-3-
methoxy-2-azetidinone (10). From 750 mg (2.75 mmol) of
aldehyde 8, 595 mg (80%) of nitrile 10 was obtained as a
yellow oil after purification by flash chromatography (hex-
anes/ethyl acetate 4:1). IR (film): 2238, 1771, 1650, 1500,
NMR (400 MHz): d 1.86 (s, 3H), 1.89 (s, 3H), 3.26 (dd, 1H,
J¼2.0, 8.2 Hz), 3.47 (s, 3H), 3.59 (dd, 1H, J¼5.2, 8.2 Hz),
3.80 (d, 1H, J¼2.0 Hz), 4.60 (d, 1H, J¼5.2 Hz), 7.25–7.45
(m, 5H); 13C NMR (100 MHz): d 26.2, 26.6, 49.8, 57.3, 59.3,
60.1, 61.0, 83.1, 119.6, 125.7, 128.6, 128.6, 135.5, 165.7.
1
700 cmꢁ1; H NMR (400 MHz): d 1.74 (s, 3H), 1.81 (s,
3H), 3.45 (s, 3H), 4.52 (dd, 1H, J¼4.8, 9.4 Hz), 4.59 (d, 1H,
J¼4.8 Hz), 6.32 (dd, 1H, J¼9.4, 16.0 Hz), 6.81 (d, 1H,
J¼16.0 Hz), 7.30–7.50 (m, 5H); 13C NMR (100 MHz):
d 26.0, 26.9, 50.2, 58.7, 61.3, 84.5, 119.6, 123.2, 126.8,
128.3, 128.6, 128.7, 135.7, 137.1, 162.6; FABHRMS calcd
for C16H19N2O2 (M++1): 271.1441; found: 271.1439.
3.5.2. 1-(10,10-Dimethyl-10-cyanoethyl)-4-[(1a,2a)-epoxy-
2-phenylethyl]-3-methoxy-2-azetidinone (2a) and 1-(10,
10-dimethyl-10-cyanoethyl)-4-[(1b,2b)-epoxy-2-phenyl-
ethyl]-3-methoxy-2-azetidinone (2b). Epoxidation of com-
pound 11 (250 mg, 0.9 mmol) throughout 6 days gave a 3:2
mixture of the compounds 2a and 2b (225 mg, 85%) after pu-
rificationbyflashchromatography(hexanes/ethyl acetate7:3).
3.4.2. 1-(10,10-Dimethyl-20-cyanoethyl)-4-[(E)-styryl]-3-
methoxy-2-azetidinone (11). From 900 mg (3.14 mmol)
of aldehyde 9, 625 mg (70%) of nitrile 11 was obtained as
a pale yellow oil after purification by flash chromatography
(hexanes/ethyl acetate 7:3). IR (film): 2253, 1755, 1655,
1500, 755, 700 cmꢁ1; 1H NMR (200 MHz): d 1.47 (s, 6H),
2.74 (d, 1H, J¼16.6 Hz), 3.05 (d, 1H, J¼16.6 Hz), 3.42 (s,
3H), 4.55 (dd, 1H, J¼4.4, 8.6 Hz), 4.57 (d, 1H, J¼4.4 Hz),
6.35 (dd, 1H, J¼8.6, 16.0 Hz), 6.75 (d, 1H, J¼16.0 Hz),
7.30–7.40 (m, 5H); 13C NMR (50 MHz): d 25.4, 26.7,
29.5, 54.8, 58.7, 61.2, 84.1, 117.0, 125.0, 126.8, 128.6,
3.5.2.1. Isomer 2a. From a 3:2 mixture of 2a/2b. 1H NMR
(400 MHz): d 1.45 (s, 3H), 1.48 (s, 3H), 2.72 (d, 1H,
J¼16.7 Hz), 3.11 (d, 1H, J¼16.7 Hz), 3.20 (dd, 1H, J¼1.9,
7.7 Hz), 3.60 (s, 3H), 3.78 (dd, 1H, J¼4.6, 7.7 Hz), 3.81 (d,
1H, J¼1.9 Hz), 4.61 (d, 1H, J¼4.6 Hz), 7.40–7.50 (m, 5H);
13C NMR (100 MHz): d 25.3, 26.8, 29.1, 54.2, 57.6, 59.3, 59.6,
60.7, 83.3, 116.9, 125.5, 125.7, 128.6, 128.7, 135.4, 166.4.
1
3.5.2.2. Isomer 2b. From a 3:2 mixture of 2a/2b. H
NMR (400 MHz): d 1.57 (s, 3H), 1.61 (s, 3H), 2.87 (d, 1H,

L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
3023
J¼16.7 Hz), 3.06 (d, 1H, J¼16.7 Hz), 3.22 (dd, 1H, J¼2.0,
8.3 Hz), 3.45 (s, 3H), 3.58 (dd, 1H, J¼5.0, 8.3 Hz), 3.80 (d,
1H, J¼2.0 Hz), 4.58 (d, 1H, J¼5.0 Hz), 7.30–7.40 (m, 5H);
13C NMR (100 MHz): d 25.6, 26.5, 29.3, 54.3, 57.6, 58.2,
59.3, 60.8, 82.5, 117.0, 125.5, 125.7, 128.6, 128.7, 135.5, 166.2.
(4.0 mmol) of activated zinc granules was added. The result-
ing red mixture was then vigorously stirred under argon with
rigorous exclusion of oxygen until a green colour was ob-
served (about 20 min).
3.7. General procedure for the radical cyclisation
reaction. Synthesis of bi- and tricyclic b-lactams 12–15
3.5.3. cis-1-(o-Cyanophenyl)-4-[(1a,2a)-epoxy-2-phenyl-
ethyl]-3-methoxy-2-azetidinone (3a) and cis-1-(o-cyano-
phenyl)-4-[(1b,2b)-epoxy-2-phenylethyl]-3-methoxy-2-
azetidinone (3b). The epoxidation of compound 6 (260 mg,
0.9 mmol) during 5 days afforded a 3:2 diastereomeric mix-
ture of compounds 3a/3b (230 mg, 85%). After purification
of the crude product by flash chromatography (hexanes/ethyl
acetate 9:1), 50 mg of the pure isomer 3a (18%) and 180 mg
of a 1:1 mixture of 3a/3b (66%) were isolated.
Method A: a THF green suspension of Cp2TiCl, generated as
reported above, was added dropwise through cannula to the
corresponding epoxide (1.0 mmol) in THF (17.0 mL). The
reaction mixture was stirred at room temperature until a
colour change from green to orange was observed and then
the reaction is quenched with 10% v/v aqueous KH2PO4
(30.0 mL). The aqueous phase was extracted with ethyl ace-
tate and the combined organic extracts were filtered through
CeliteÒ, dried (anhydrous Na2SO4) and concentrated in
vacuo.
3.5.3.1. Isomer 3a. Colourless solid. Mp 138–139 ꢂC (hex-
1
anes/CH2Cl2). IR (KBr) 2228, 1777, 1160 cmꢁ1; H NMR
(400 MHz): d 3.37 (dd, 1H, J¼1.0, 7.6 Hz), 3.58 (s, 3H), 3.86
(d, 1H, J¼1.0 Hz), 4.80 (dd, 1H, J¼5.2, 7.6 Hz), 4.87 (d, 1H,
J¼5.2 Hz), 7.20–7.35 (m, 6H), 7.61 (t, 1H, J¼8.2 Hz), 7.65
(t, 1H, J¼8.2 Hz), 7.96 (d, 1H, J¼8.2 Hz); 13C NMR
(100 MHz): d 56.2, 59.5, 59.8, 61.1, 84.0, 103.9, 117.1, 123.7,
125.7, 126.0, 128.5, 133.9, 135.6, 138.5, 165.2; FABHRMS
calcd for C19H17N2O3 (M++1): 321.1239, found: 321.1201.
Method B: a solution of the corresponding epoxide
(1.0 mmol) in THF (17.0 mL) was added dropwise through
cannula to a green suspension of Cp2TiCl, generated in situ
from Cp2TiCl2 and Zn0 as described above. The resulting
reaction mixture was stirred until a colour change was
observed and then worked up as reported in Method A.
1
3.5.3.2. Isomer 3b. From a 1:1 mixture of 3a/3b. H
3.7.1. Bicyclic b-lactam 12a. Method A. From 80 mg
(0.3 mmol) of epoxynitrile 1a, 46 mg (60%) of compound
12a was obtained as a colourless solid after purification
by flash chromatography (hexanes/ethyl acetate 7:3). Like-
wise, from a 4:5 mixture of compounds 1a/1b (180 mg,
0.63 mmol) a 4:3 mixture of compounds 1b/12a (106 mg,
60%) was obtained. Method B. From a 4:5 mixture of com-
pounds 1a/1b (100 mg, 0.36 mmol) a 2:4:3 mixture of com-
pounds 1b/10/12a (55 mg, 57%) was obtained. Compound
12a: mp 176–179 ꢂC (hexanes/dichloromethane). IR:
NMR (400 MHz): d 3.35 (dd, 1H, J¼1.9, 4.5 Hz), 3.64 (s,
3H), 3.74 (d, 1H, J¼1.9 Hz), 4.75–4.80 (m, 2H), 7.15–7.35
(m, 6H), 7.62 (t, 1H, J¼6.9 Hz), 7.68 (t, 1H, J¼6.9 Hz), 7.85
(d, 1H, J¼6.9 Hz); 13C NMR (100 MHz): d 57.0, 58.3, 60.5,
61.0, 85.3, 103.2, 117.2, 122.8, 125.7, 125.8, 128.6, 128.7,
134.1, 135.2, 138.1, 164.2.
3.5.4. trans-1-(o-Cyanophenyl)-4-[(1a,2a)-epoxy-2-phen-
ylethyl]-3-methoxy-2-azetidinone (4a) and trans-1-(o-
cyanophenyl)-4-[(1b,2b)-epoxy-2-phenylethyl]-3-meth-
oxy-2-azetidinone (4b). The epoxidation of compound 7
(850 mg, 2.8 mmol) for 2 days afforded a 1:2 diastereomeric
mixture of compounds 4a/4b (805 mg, 90%) that was
impossible to resolve chromatographically.
1
1755, 1694, 1500, 745, 700 cmꢁ1; H NMR (400 MHz):
d 1.36 (s, 3H), 1.83 (s, 3H), 3.59 (s, 3H), 4.50 (dd, 1H,
J¼1.9, 4.9 Hz), 4.73 (d, 1H, J¼4.9 Hz), 6.96 (d, 1H,
J¼1.9 Hz), 7.30–7.40 (m, 5H); 13C NMR (100 MHz): d
21.2, 24.1, 50.7, 59.1, 62.0, 86.1, 128.0, 128.3, 128.8,
135.6, 139.2, 140.0, 167.4, 196.3; FABHRMS calcd for
C16H18NO3 (M++1): 272.1281; found: 272.1244.
1
3.5.4.1. Isomer 4a. From a 1:2 mixture of 4a/4b. H
NMR (400 MHz): d 3.36 (dd, 1H, J¼2.0, 4.5 Hz), 3.64 (s,
3H), 3.75 (d, 1H, J¼2.0 Hz), 4.74–4.82 (m, 2H), 7.20–
7.35 (m, 6H), 7.63 (t, 1H, J¼7.0 Hz), 7.65 (t, 1H,
J¼7.0 Hz), 7.82 (d, 1H, J¼7.0 Hz); 13C NMR (100 MHz):
d 56.9, 58.3, 60.5, 60.9, 85.3, 103.1, 117.2, 122.8, 125.5,
125.7, 128.6, 128.7, 134.0, 135.2, 138.0, 164.2.
3.7.2. Bicyclic b-lactam 13a. Method A. From a 3:2 mixture
of compounds 2a/2b (100 mg, 0.33 mmol) a 2:7:1 mixture
of compounds 2a/2b/13a (82 mg, 83%) was obtained.
Method B. From a 3:2 mixture of compounds 2a/2b
(100 mg, 0.33 mmol) a 3:14:3:2 mixture of compounds 2a/
2b/11/13a (86 mg, 87%) was obtained. Compound 13a: 1H
NMR (400 MHz): d 1.33 (s, 3H), 1.67 (s, 3H), 2.58 (d, 1H,
J¼11.4 Hz), 2.75 (d, 1H, J¼11.4 Hz), 3.58 (s, 3H), 4.63
(d, 1H, J¼5.1 Hz), 4.78 (dd, 1H, J¼2.0, 5.1 Hz), 6.80 (d,
1H, J¼2.0 Hz), 7.25–7.45 (m, 5H); 13C NMR (100 MHz):
d 24.8, 27.5, 29.1, 54.0, 59.0, 59.2, 83.5, 125.7, 128.1,
128.7, 135.5, 139.5, 141.0, 166.2, 200.0.
1
3.5.4.2. Isomer 4b. From a 1:2 mixture of 4a/4b. H
NMR (400 MHz): d 3.31 (dd, 1H, J¼1.9, 4.2 Hz), 3.62 (s,
3H), 4.04 (d, 1H, J¼1.9 Hz), 4.66 (d, 1H, J¼2.0 Hz), 4.87
(dd, 1H, J¼2.0, 4.2 Hz), 7.20–7.35 (m, 6H), 7.60 (dt, 1H,
J¼1.5, 6.9 Hz), 7.62 (dt, 1H, J¼1.5, 6.9 Hz), 7.90 (d, 1H,
J¼6.9 Hz); 13C NMR (100 MHz): d 57.8, 58.3, 59.3, 61.4,
84.7, 102.2, 117.2, 123.2, 125.6, 125.5, 128.6, 128.7,
134.1, 135.2, 138.0, 164.2.
3.7.3. Tricyclic b-lactam 14a. Method A. From 50 mg
(0.16 mmol) of epoxynitrile 3a, 28 mg (56%) of compound
14a was obtained as a colourless solid after purification
by flash chromatography (hexanes/ethyl acetate 8:2). Like-
wise, from a 1:1 mixture of compounds 3a/3b (100 mg,
0.31 mmol), 24 mg (25%) of compound 6 and 35 mg
3.6. In situ generation of Cp2TiCl
To 548 mg (2.2 mmol) of titanocene dichloride in anhydrous
and strictly deoxygenated THF (12.5 mL), 262 mg

3024
L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
(35%) of compound 14a were isolated by flash chromato-
graphy (hexanes/ethyl acetate 4:1). Method B. From a 1:1
mixture of compounds 3a/3b (80 mg, 0.25 mmol), 20 mg
(26%) of compound 6 and 19 mg (24%) of compound 14a
were isolated by flash chromatography (hexaꢂnes/ethyl ace-
tate 4:1). Compound 14a: Mp 178–180 C (hexanes/
3.8.1. Tricyclic b-lactam 16. From 14a (30 mg,
0.09 mmol), 10 mg (36%) of 16 was isolated after flash chro-
matography (hexanes/ethyl acetate 85:15). IR (film): 2928,
1
2846, 1767, 1657, 764 cmꢁ1; H NMR (200 MHz): d 3.72
(s, 3H), 4.80 (d, 1H, J¼4.9 Hz), 4.92 (dd, 1H, J¼3.8,
4.9 Hz), 6.64 (d, 1H, J¼3.8 Hz), 7.25–7.35 (m, 6H), 7.57
(t, 1H, J¼7.5 Hz), 7.93 (d, 1H, J¼7.5 Hz), 8.06 (d, 1H,
J¼7.5 Hz); 13C NMR (50 MHz): d 55.0, 59.5, 83.7, 120.5,
125.0, 125.2, 127.1, 127.7, 128.5, 130.8, 134.5, 136.4, 137.5,
148.6, 163.2, 195.6; FABHRMS calcd for C19H15NO3Na
(M++23): 328.0950; found: 328.0926.
CH2Cl2). IR (KBr): 3480, 1760, 1669 cmꢁ1 1H NMR
;
(400 MHz): d 3.73 (s, 3H), 4.28 (d, 1H, J¼7.5 Hz), 4.29
(dd, 1H, J¼5.0, 9.8 Hz), 4.75 (dd, 1H, J¼7.5, 9.8 Hz),
4.82 (d, 1H, J¼5.0 Hz), 7.11 (dt, 1H, J¼0.7, 7.6 Hz), 7.19
(d, 1H, J¼7.6 Hz), 7.22–7.30 (m, 2H), 7.49 (ddd, 1H,
J¼1.5, 7.6, 8.2 Hz), 8.10 (dd, 1H, J¼0.7, 8.2 Hz); 13C
NMR (100 MHz): d 59.6, 60.1, 69.9, 72.4, 84.1, 119.3,
124.8, 127.8, 128.2, 128.9, 131.2, 133.3, 135.3, 137.0,
163.7, 199.9; FABHRMS calcd for C19H18NO4 (M++1):
324.1230; found: 324.1224.
3.8.2. Tricyclic b-lactam 17. From a 1:3 mixture of com-
pounds 15a/15b (90 mg, 0.28 mmol), 30 mg (35%) of 17
was isolated after flash chromatography (hexanes/ethyl ace-
tate 85:15). IR (film): 3066, 2935, 2839, 1760, 1650,
1
771 cmꢁ1; H NMR (200 MHz): d 3.64 (s, 3H), 4.77 (d,
3.7.4. Tricyclic b-lactams 15a and 15b. Method A. From
a 1:2 mixture of compounds 4a/4b (380 mg, 1.19 mmol),
19 mg (5%) of epoxide 4a, 65 mg (17%) of compound 15a
and 235 mg of 15b (61%) were isolated by flash chromato-
graphy (hexanes/ethyl acetate 4:1). Method B. From a 1:2
mixture of compounds 4a/4b (400 mg, 1.25 mmol), 84 mg
(22%) of 7, 44 mg (11%) of 15a and 137 mg (34%) of 15b
were isolated by flash chromatography (hexanes/ethyl ace-
tate 4:1).
1H, J¼1.9 Hz), 4.86 (dd, 1H, J¼1.9, 3.4 Hz), 6.59 (d, 1H,
J¼3.4 Hz), 7.24 (t, 1H, J¼7.8 Hz), 7.30–7.35 (m, 5H),
7.57 (t, 1H, J¼7.8 Hz), 7.92 (d, 1H, J¼7.8 Hz), 8.17 (d,
1H, J¼7.8 Hz); 13C NMR (50 MHz): d 58.0, 59.1, 88.9,
120.4, 124.9, 127.1, 127.7, 128.3, 128.5, 130.9, 134.4,
136.5, 137.5, 148.3, 162.8, 194.1; FABHRMS calcd for
C19H15NO3Na (M++23): 328.0950, found: 328.0936.
Acknowledgements
3.7.4.1. Isomer 15a. IR (film): 3449, 1761,
1691 cmꢁ1 1H NMR (400 MHz): d 3.57 (s, 3H), 4.06
;
Financial support for this work from the Ministerio de Edu-
ꢀ
cacion y Ciencia of Spain (CTQ2005-05026/BQU) and the
Junta de Castilla y Leon (SA070/03) is gratefully acknowl-
(dd, 1H, J¼2.2, 4.3 Hz), 4.34 (d, 1H, J¼8.4 Hz), 4.58 (dd,
1H, J¼4.3, 8.4 Hz), 5.17 (d, 1H, J¼2.2 Hz), 7.18 (t, 1H,
J¼7.4 Hz), 7.36–7.41 (m, 4H), 7.55 (ddd, 1H, J¼0.9, 7.4,
8.2 Hz), 8.33 (d, 1H, J¼8.2 Hz); 13C NMR (100 MHz):
d 58.4, 64.4, 65.8, 74.6, 84.0, 119.0, 124.4, 128.2, 128.4,
128.8, 129.9, 133.1, 134.6, 139.3, 165.8, 196.6; FABHRMS
calcd for C19H17NO4Na (M++23): 323.1050; found:
323.1047.
ꢀ
edged. We would also like to thank a referee for his sugges-
tions on the cyclisation mechanism and the Ministerio de
ꢀ
Educacion y Ciencia of Spain for a grant to L.M.M.
References and notes
1. See, for example: (a) Fisher, J. F.; Meroueh, S. O.; Mobashery,
S. Chem. Rev. 2005, 105, 395; (b) Ritter, T. K.; Wong, C.-H.
3.7.4.2. Isomer 15b. IR (film): 3449, 1762, 1671 cmꢁ1
;
1H NMR (400 MHz): d 3.53 (s, 3H), 4.10 (dd, 1H, J¼1.8,
9.4 Hz), 4.23–7.27 (m, 2H), 4.67 (d, 1H, J¼1.8 Hz), 7.13
(t, 1H, J¼7.2 Hz), 7.17 (d, 1H, J¼7.9 Hz), 7.30–7.35 (m,
2H), 7.37 (dd, 1H, J¼1.3, 7.2 Hz), 7.50 (ddd, 1H, J¼1.3,
7.2, 8.2 Hz), 8.14 (dd, 1H, J¼0.6, 8.2 Hz); 13C NMR
(100 MHz): d 58.5, 65.3, 69.6, 71.8, 87.4, 120.0, 124.9,
128.1, 128.7, 128.9, 129.3, 131.1, 133.4, 135.4, 135.8, 163.7,
199.3; FABHRMS calcd for C19H17NO4Na (M++23):
323.1050; found: 323.1035.
Angew. Chem., Int. Ed. 2001, 40, 3508; (c) Dıaz, N.; Suarez,
D.; Merz, K. M. M., Jr. J. Am. Chem. Soc. 2000, 122, 4197;
(d) Page, M. I. Chem. Commun. 1998, 1609; (e) Spratt, B. G.
Science 1994, 264, 388.
ꢀ
ꢀ
2. For selected reviews, see: (a) Alcaide, B.; Almendros, P. Curr.
ꢀ
Org. Chem. 2002, 6, 245; (b) Gomez-Gallego, M.; Manchedo,
M. J.; Sierra, M. A. Tetrahedron 2000, 56, 5743; (c) Setti, E. L.;
Micetich, R. G. Curr. Med. Chem. 1998, 5, 101.
3. For selected references, see: (a) Zard, S. Z. Radical Reactions
in Organic Synthesis; Oxford University Press: New York, NY,
2003; (b) Radicals in Organic Synthesis; Renaud, P., Sibi,
M. P., Eds.; Wiley-VCH: Weinheim, 2002; (c) Curran, D. P.;
Porter, N. A.; Giese, B. Stereochemistry of Radical
Reactions; VCH: New York, NY, 1996; (d) Giese, B.
Radicals in Organic Synthesis: Formation of Carbon–Carbon
Bonds; Pergamon: Oxford, 1986.
3.8. Preparation of unsaturated tricyclic b-lactams
16 and 17
To a solution of the 10-hydroxy-tricyclic compound
(1.0 mmol) in anhydrous dichloromethane (10 mL), dimeth-
ylaminopyridine, TEA (0.17 mL, 1.20 mmol) and tosyl
chloride (229 mg, 1.20 mmol) were successively added un-
der argon atmosphere. The resulting mixture was stirred at
room temperature until the total disappearance of the start-
ing material was observed in TLC. Then, the reaction mix-
ture was poured into a cold saturated solution of NH4Cl
and extracted with dichloromethane (10 mLꢄ3). The com-
bined organic extracts were dried (anhydrous Na2SO4) and
concentrated under reduced pressure.
4. (a) Ruano, G.; Anaya, J.; Grande, M. Synlett 1999, 1441; (b)
Hernando, J. I. M.; Laso, M. N.; Anaya, J.; Gero, S. D.;
Grande, M. Synlett 1996, 281; (c) Anaya, J.; Barton,
ꢀ
D. H. R.; Gero, S. D.; Grande, M.; Martın, N.; Tachdjian, C.
Angew. Chem., Int. Ed. Engl. 1993, 32, 867; (d) Barton,
ꢀ
D. H. R.; Gateau-Olesker, A.; Anaya-Mateos, J.; Cleophax,
J.; Gero, S. D.; Chiaroni, A.; Riche, C. J. Chem. Soc., Perkin
Trans. 1 1990, 3211.

L. M. Monleoꢀn et al. / Tetrahedron 63 (2007) 3017–3025
3025
5. Anaya, J.; Gero, S. D.; Grande, M.; Hernando, J. I. M.; Laso,
N. M. Bioorg. Med. Chem. 1999, 7, 837.
9. (a) Staudinger, M. Liebigs Ann. Chem. 1907, 356, 51; For
reviews on [2+2] cycloaddition of imines and ketenes, see: (b)
Palomo, C.; Aizpurua, J. M.; Ganboa, I.; Oiarbide, M. Curr.
Med. Chem. 2004, 11, 1837; (c) Palomo, C.; Aizpurua, J. M.;
Ganboa, I.; Oiarbide, M. Eur. J. Org. Chem. 1999, 3223; (d)
Georg, G. I.; Ravikumar, V. T. The Organic Chemistry of
b-Lactams; Georg, G. I., Ed.; VCH: Weinheim, 1993;
Chapter 3, p 295.
ꢀ
6. (a) Fernandez-Mateos, A.; Mateos-Buron, L.; Martın de la
Nava, E. M.; Rabanedo Clemente, R.; Rubio Gonzalez, R.;
ꢀ
ꢀ
ꢀ
ꢀ
Sanz Gonzalez, F. Synlett 2004, 2553; (b) Gansauer, A.;
Lauterbach, T.; Narayan, S. Angew. Chem., Int. Ed. 2003, 42,
€
€
3687; (c) Li, J. J. Tetrahedron 2001, 57, 1; (d) Gansauer, A.;
Bluhm, H. Chem. Rev. 2000, 100, 2771; (e) Fernandez-
ꢀ
ꢀ
Mateos, A.; Martın de la Nava, E.; Pascual Coca, G.; Ramos
Silvo, A.; Rubio Gonzalez, R. Org. Lett. 1999, 1, 607; (f)
10. All these compounds are racemic mixtures but only one stereo-
isomer is depicted for simplicity.
ꢀ
€
ꢀ
11. For other examples, see: (a) Alcaide, B.; Aly, M. F.; Rodrıguez,
Gansauer, A.; Bluhm, H.; Pierobon, M. J. Am. Chem. Soc.
ꢀ
1998, 120, 12849; (g) Nugent, W. A.; RajanBabu, T. V.
J. Am. Chem. Soc. 1994, 116, 986.
C.; Rodrıguez-Vicente, A. J. Org. Chem. 2000, 65, 3453; (b)
Ren, X. F.; Turos, E.; Lake, C. H.; Churchill, M. R. J. Org.
Chem. 1995, 60, 6468; (c) Zamboni, R.; Just, G. Can. J.
Chem. 1979, 57, 1945.
ꢀ
7. (a) Anaya, J.; Fernandez-Mateos, A.; Grande, M.; Martian˜ez,
J.; Ruano, G.; Rubio-Gonzalez, M . R. Tetrahedron 2003, 59,
ꢀ
a
ꢀ
ꢀ˜
ꢀ ꢀ
12. Fernandez, R.; Consolacion, G.; Lassaretta, J.-M.; Llera, J.-M.;
241; (b) Ruano, G.; Martianez, J.; Grande, M.; Anaya, J.
ꢀ
J. Org. Chem. 2003, 68, 2024; (c) Ruano, G.; Grande, M.;
Anaya, J. J. Org. Chem. 2002, 67, 8243.
8. The epoxynitrile cyclisation has been recently reported as a
novel and powerful route to mono- and bicyclic b-hydroxyke-
Vazquez, J. Tetrahedron Lett. 1993, 34, 141.
13. It is necessary to emphasise the enormous difficulties found
in the purification process of epoxylactams 1 and 2, due to
their nearly identical behaviour in TLC to nitriles 10 and 11,
respectively.
ꢀ
tones (Scheme 4) by Fernandez-Mateos, A.; Mateos-Buron, L.;
Rabanedo Clemente, R.; Ramos Silvo, A.; Rubio Gonzalez, R.
ꢀ
ꢀ
€
14. Pretsch, E.; Buhlmann, P.; Affolter, C. Structure Determination
Synlett 2004, 1011.
of Organic Compounds, 3rd ed.; Springer: Berlin, Germany,
2000.
15. Conformational minima derived from analysis with CS
Chem3D Pro 5.0 software; CambridgeSoft: Cambridge, MA,
1999; using MM2 calculations.
OH
O
O
1) Cp2TiCl
2) H3O
CN
n
n
16. Daasbjerg, K.; Svith, H.; Grimme, S.; Gerenkamp, M.; M€uck-
€
Lichtenfeld, C.; Gansauer, A.; Barchuk, A.; Keller, F. Angew.
Chem., Int. Ed. 2006, 2041.
Scheme 4. Radical epoxynitrile cyclisations mediated by Ti(III) reagent.
.