The formation of silylated β-lactams from silylketenes through Lewis acid promoted [2+2] cycloaddition: A combined theoretical and experimental study

Article abstract of DOI:10.1002/ejoc.200400907

The stereoselective formation of silylated cis-β-lactams from (trimethylsilyl)ketene and an α-imino ester by Lewis acid catalysis is described. Theoretical results suggest that the reaction between (trimethylsilyl)ketene and trans- (methoxycarbonyl)-N-methylformaldimine would proceed most favourably with the BF₃ catalyst coordinated to the ketene. Moreover, the calculated energy barriers account for the cis:trans ratio found experimentally. ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2005).

Full text of DOI:10.1002/ejoc.200400907

FULL PAPER
The Formation of Silylated β-Lactams from Silylketenes through Lewis Acid
Promoted [2+2] Cycloaddition: A Combined Theoretical and Experimental
Study
Béatrice Pelotier,[a] Michel Rajzmann,[a] Jean-Marc Pons,[a] Pablo Campomanes,[b]
Ramón López,[b] and Tomás L. Sordo*[b]
Keywords: Silylketenes / Lactams / Lewis acids / Reaction mechanisms / Density functional calculations
The stereoselective formation of silylated cis-β-lactams from
ably with the BF3 catalyst coordinated to the ketene. More-
(trimethylsilyl)ketene and an α-imino ester by Lewis acid ca- over, the calculated energy barriers account for the cis:trans
talysis is described. Theoretical results suggest that the reac-
tion between (trimethylsilyl)ketene and trans-(methoxycar-
bonyl)-N-methylformaldimine would proceed most favour-
ratio found experimentally.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2005)
Introduction
give the cis-β-lactam (Scheme 1). The mechanism is based-
both on experimental evidence of the occurrence of a zwit-
terionic intermediate[3] and on theoretical studies.[4]
Examples of the use of silylketenes[5] in β-lactam syn-
thesis are rather scarce and always involve an electron-
poor imine.[6,7] A typical example is provided by Zaitse-
The Staudinger [2+2] cycloaddition reaction between an
imine and a ketene has been known for nearly a century[1]
and is still widely used to prepare β-lactams.[2] The mecha-
nism of the reaction involves stereoselective attack by the
lone pair of the nitrogen atom of the imine on the central
carbon atom of the ketene and subsequent conrotatory elec- vaand co-workers[7] who prepared a trans-silylated β-lac-
trocyclisation of the resultant zwitterionic intermediate to tam from (trimethylsilyl)ketene (1)[8] and (methylsulfo-
Scheme 1.
Scheme 2.
nyl)chloraldimine (see Scheme 2).[7] No Lewis acid was
used in any of these examples.
Our experience of silylketene chemistry led us to pro-
pose that these compounds could in certain circumstances
behave as nucleophiles, for example, in the Lewis acid cat-
[a] Laboratoire SYMBIO (UMR CNRS 6178), Equipe RéSO,
Université Paul Cézanne, Faculté des Sciences et Techniques,
Service D12, 13397 Marseille Cedex 20, France
E-mail: jean-marc.pons@univ.u-3mrs.fr
[b] Departamento de Química Física y Analítica, Universidad de
Oviedo,
Julián Clavería 8, 33006, Oviedo, Principado de Asturias, Spain alysed formation of β-lactones from aldehydes. This con-
clusion, based on experimental[9] and theoretical work,[10]
E-mail: tsordo@uniovi.es
Supporting information for this article is available on the
WWW under http://www.eurjic.com or from the author.
and the above-mentioned reactivity of silylketenes
Eur. J. Org. Chem. 2005, 2599–2606
DOI: 10.1002/ejoc.200400907
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2599

B. Pelotier, M. Rajzmann, J.-M. Pons, P. Campomanes, R. López, and Tomás L. Sordo,
FULL PAPER
towards imines[6,7] led us to study, both experimentally NMR experiments (100.3 MHz, CDCl3); the main observa-
and theoretically, the formation of β-lactams from silylke- tions are as follows:
tenes and imines under Lewis acid catalysis conditions.
i) No complexation between silylketene 1 and BF3–Et2O
The publication of a study of a catalysed Staudinger reac- can be observed between –40 and 20 °C (i.e. no shift in the
tion[11] prompted us to publish our results.[12]
signals corresponding to silylketene 1 and complexed-Et2O
carbon atoms occurred).
ii) Imine 3–BF3 complexes can be observed at –40 °C (see
Expt. Sect.).
iii) β-Lactam 4 is formed in the NMR tube at –40 °C.
Since no intermediate, except the imine–BF3 complex,
was observed directly, we decided to investigate the mecha-
nism of the reaction by theoretical means.
Results and Discussion
Experimental Study
The study was carried out with two different imines: a
standard one, n-hexanaldimine 2, and an electron-poor one,
n-butyl glyoxylate imine 3 (see Scheme 3). The latter was
chosen since it is a fairly stable compound that can be
stored and because it gives reproducible results.[13] (trimeth-
ylsilyl)ketene (1) and BF3–Et2O were the other reagents
used in the reaction which was carried out in Et2O.
Theoretical Study
It is well known that the Staudinger reaction between
ketenes and imines involves low energy barriers allowing
It appears from Scheme 3 that the desired reaction oc- the quick formation of β-lactams even at reduced tempera-
curs only when the glyoxylate imine 3 and BF3–Et2O react tures.[14] However, the presence of a silyl substituent on the
with (trimethylsilyl)ketene 1. Indeed, β-lactam 4 is formed ketene slows the process considerably and therefore, in ge-
in fair yield as a 60:15:15:10 (cis:cis:trans:trans) mixture of neral, a catalyst is required. Indeed, our calculations of the
1
four diastereoisomers (determined by 400 MHz H NMR Staudinger reaction at the B3LYP/6-311+G(d,p)/PCM//
spectroscopy). In the other three cases, at temperatures at B3LYP/6-311+G(d,p) level of theory reveal that the rate-
or below 0 °C only the starting materials were recovered. determining barrier for the uncatalysed reaction between
When the temperature was allowed to rise to room tempera- methylketene and formaldimine, which corresponds to the
ture, a complex mixture of products was formed (resulting electrocyclic conrotatory closure of the zwitterionic inter-
from the degradation of the starting materials) but no β- mediate, has a value of 17.1 kcalmol–1 for the anti approach
lactams (silylated or not) were observed.
and 23.2 kcalmol–1 for the syn approach, whereas for the
The yields of β-lactam 4 were disappointing when vari- reaction with silylketene, the corresponding values are 26.2
ous other Lewis acids were used (SnCl4: 0%; AlCl3: 20%; and 25.2 kcalmol–1, respectively (see Figure 1). These calcu-
EtAlCl2 and TMSOTf: 35% yields). With MgBr2–Et2O, the lations are in agreement with the fact that silylketenes are
β-lactam ring was formed in 84% yield but it was not pos- much more stable than alkylketenes.
sible to prevent desilylation: β-lactam 4 was obtained along
In order to investigate the mechanism of the Staudinger
with the corresponding desilylated molecule (up to 50% of reactions studied experimentally, we calculated the energy
the reaction yield). With BF3–OEt2 as the Lewis acid, the profiles for the reactions of silylketene (H3SiCH=C=O)
use of other solvents did not lead to better yields of β-lac- with trans-(methoxycarbonyl)-N-methylformaldimine (R2 =
tam 4 (CH2Cl2: 30%; toluene: 39%) and had very little ef- CO2Me) and trans-N-methyl-methylformaldimine (R2
=
fect on the cis:trans ratio (CH2Cl2: 78:22; toluene: 74:26).
Me) catalyzed by BF3 as Lewis acid. These two imines were
In order to improve our understanding of the mechanism chosen as models for imines 3 and 2, respectively, while si-
of this reaction, we also performed low-temperature 13C lylketene was used as a model for (trimethylsilyl)ketene (1).
Scheme 3.
2600
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2599–2606

Silylated β-Lactams from Silylketenes
FULL PAPER
Reaction between Silylketene and trans-(Methoxycarbonyl)-
N-methylformaldimine in the Presence of BF3
The energies of all the critical structures of both the re-
verse electron demand and the catalyzed Staudinger mecha-
nisms are given in Table 1 and Figure 2 shows the corre-
sponding energy profiles in solution. The B3LYP/6-
311+G(d,p)/PCM//B3LYP/6-311+G(d,p)/Onsager energies,
including the zero-point vibrational energy (ZPVE) correc-
tion from the Onsager frequencies, are given in the text.
Reverse Electron Demand Mechanism
Figure 1S (available in the Supporting Information; see
also the footnote on the first page of this article), shows the
optimized geometries of the critical structures located in the
reverse electron demand mechanism.
The BF3 catalyst coordinated to the imine gives rise to a
complex, C1cN, that is 17.8 kcalmol–1 more stable than the
reactants. This complex has been experimentally detected
(vide supra). C1cN and the ketene can react by three dif-
ferent routes: when the two reactants approach with the
silyl and methoxycarbonyl substituents in a trans arrange-
ment, both concerted and stepwise mechanisms are possible
whereas the corresponding cis interaction only gives rise to
a stepwise path. C1cN reacts by the trans concerted route
through transition state TS3cNtrans, which is 8.9 kcalmol–1
Figure 1. B3LYP/6-311+G(d,p)-optimized geometries of the transi-
tion states for the electrocyclic conrotatory closure step for the re-
actions between methylketene and formaldimine by a) the anti and
b) the syn approaches, and between silylketene and formaldimine
by c) the anti and d) the syn approaches. Lengths are given in
Ångstrøms.
The catalyst may bond to the nitrogen atom of the imine above the reactants giving an energy barrier of
and induce a reverse electron demand mechanism or to the 26.7 kcalmol–1, to yield the corresponding β-lactam with
oxygen atom of the ketene and induce a catalyzed Staud- BF3 coordinated to the nitrogen atom, C2cNtrans
,
inger mechanism (see Scheme 4).
16.3 kcalmol–1 below the energy of the reactants. By the
Scheme 4.
Eur. J. Org. Chem. 2005, 2599–2606
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2601

B. Pelotier, M. Rajzmann, J.-M. Pons, P. Campomanes, R. López, and Tomás L. Sordo,
FULL PAPER
Table 1. Relative B3LYP/6-311+G(d,p) electronic energies in the plex, C2cNtrans, via the transition state TS2cNtrans, which is
8.4 kcalmol–1 less stable than the reactants. Hence, the first
gas phase (including ZPVE) (ΔE) and relative B3LYP/6-
311+G(d,p)/PCM//B3LYP/6-311+G(d,p)/Onsager energies (in-
step is rate-determining with an activation barrier of
cluding ZPVE corrections from Onsager frequency calculations)
25.3 kcalmol–1. When the reactants approach with the silyl
(ΔEsolution) of the structures located for the reaction of silylketene
and methoxycarbonyl substituents in the cis orientation, the
reaction between C1cN and the silylketene evolves along
a two-step mechanism analogous to the previous one via
transition state TS1cNcis, intermediate IcNcis and transition
state TS2cNcis (with relative energies with respect to the re-
actants of 11.1, –9.4, and 10.4 kcalmol–1, respectively) to
afford the cis-β-lactam–BF3 complex which is
18.1 kcalmol–1 more stable than the reactants. Therefore,
the first stage of this mechanism is rate-determining with
an energy barrier of 28.9 kcalmol–1.
with trans-(methoxycarbonyl)-N-methylformaldimine with (top)
and without (bottom) the BF3 catalyst.
Structures
ΔE/kcalmol–1
ΔEsolution
/
kcalmol–1
BF3 + ketene + imine
C1cN + ketene
TS1cNtrans
IcNtrans
TS2cNtrans
TS1cNcis
0.0
–14.0
9.1
–18.2
7.6
14.9
–12.2
10.4
14.1
–17.5
–37.8
–1.9
4.9
0.0
–17.8
7.5
–14.1
8.4
11.1
–9.4
10.4
8.9
–18.1
–16.3
–0.5
7.6
–3.5
8.7
9.3
–3.6
9.1
IcNcis
TS2cNcis
TS3cNtrans
C2cNcis
C2cNtrans
C1cO + imine
TS1cOanti
IcOanti
TS2cOanti
TS1cOsyn
IcOsyn
TS2cOsyn
C2cOtrans
C2cOcis
Catalyzed Staudinger Mechanism
Figure 3 shows the optimized geometries of the critical
structures located in the Staudinger mechanism.
–1.1
13.4
7.2
When the catalyst is coordinated to silylketene, complex
C1cO is formed which is 0.5 kcalmol–1 more stable than
the reactants. When the imine approaches C1cO in an anti
orientation with respect to the silyl substituent a two-step
mechanism results involving transition state TS1cOanti, in-
termediate IcOanti and transition state TS2cOanti with rela-
tive energies of 7.6, –3.5 and 8.7 kcalmol–1, respectively, to
4.2
13.5
–31.0
–29.6
–34.3
–32.7
Ketene + imine
TS2syn
TS2anti
trans-β-Lactam
cis-β-Lactam
0.0
0.0
28.5
27.3
–18.9
–17.5
28.0
27.1
–17.8
–16.0
yield
the
cis-β-lactam–BF3
complex,
C2cOcis,
32.7 kcalmol–1 lower than the reactants.
Thus, the second step is rate-determining with an energy
barrier of 12.2 kcalmol–1. When C1cO approaches the im-
ine with a syn orientation, the system evolves along an anal-
trans stepwise path, C1cN reacts via transition state ogous two-step route through transition state TS1cOsyn
TS1cNtrans, 7.5 kcalmol–1 above the reactants, to form an (9.3 kcalmol–1), intermediate IcOsyn (–3.6 kcalmol–1) and
intermediate, IcNtrans, that is 21.6 kcalmol–1 more stable. transition state TS2cOsyn (9.1 kcalmol–1) to give the trans-
Intermediate IcNtrans then yields the β-lactam–BF3 com- β-lactam–BF3 complex, C2cOtrans, that is 34.3 kcalmol–1
Figure 2. B3LYP/6-311+G(d,p)/PCM//B3LYP/6-311+G(d,p)/Onsager+ZPVE(B3LYP/Onsager) energy profiles for the reaction between
silylketene and trans-(methoxycarbonyl)-N-methylformaldimine in the presence of BF3.
2602
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2599–2606

Silylated β-Lactams from Silylketenes
FULL PAPER
more stable than the original reactants. Here again the sec-
ond stage of the reaction is rate-determining with an energy
barrier of 12.7 kcalmol–1.
Discussion and Comparison with Experiment
It is clear from these calculations that the catalyzed
Staudinger mechanism is the favoured one and hence the
following rationalization of the experimental results can be
proposed. Initially roughly half the ketene and imine are
complexed to BF3 leading to the imine–BF3 and ketene–
BF3 complexes. On one hand, given its relative thermo-
dynamic stability, the BF3–imine complex is involved in no
(or very little) cycloaddition resulting in a yield of β-lactam
of about 50%, in good agreement with the experimental
results obtained for the reaction between silylketene 1 and
glyoxylate imine 3 (see Scheme 3). On the other hand, de-
spite its low relative stability, most of the cycloaddition re-
action occurs from the ketene–BF3 complex because of the
low activation barriers in the routes starting from this
point. These relatively low activation barriers result from a
stronger complexation of BF3 to the oxygen atom of the
ketene in the transition state structures (1.691 or 1.684 Å)
than in the initial ketene–BF3 complex (2.591 Å). The dif-
ference in the rate-determining energy barrier between the
syn (12.7 kcalmol–1) and anti (12.2 kcalmol–1) approaches
of the C1cO and imine renders a theoretical cis:trans ratio
for the β-lactam–BF3 complex at –30 °C of 74:26, close to
the experimental value. Note that the asymmetric induction
experimentally observed, particularly in the case of the cis
isomers (see Scheme 3), can be attributed to the fact that
the difference between the two faces of the β-lactam ring
introduced by the chiral substituent on the nitrogen atom
of the imine employed in the experimental work has a larger
effect on the two cis isomers because in the corresponding
rate-determining transition states the two substituents ap-
pear on the same face whereas in the transition states for
the trans-β-lactams there is a substituent on each face.
Note that the β-lactam–BF3 complexes in which the cata-
lyst is coordinated to the nitrogen atom are considerably
less stable than those in which BF3 is coordinated to the
oxygen atom. This is clearly a consequence of the hindrance
of the conjugation between the lone pair on the nitrogen
atom and the C–O double bond in the β-lactam ring, which
is also reflected in a lengthening of the C–N bond length
of about 0.15 Å compared with the length of the C–N bond
in the isolated β-lactam ring (approximately 1.37 Å).
The catalytic effect of BF3 in these processes is very im-
portant. As shown at the bottom of Table 1 the rate-de-
termining energy barriers for the formation of the cis- and
trans-β-lactams in solution without the Lewis acid catalyst
are 27.1 and 28.0 kcalmol–1, respectively, more than double
the values for the most favourable catalyzed mechanisms.
Note also that in the presence of BF3 as catalyst, the
energy barriers for the most favourable mechanisms of the
reaction between silylketene and trans-N-methyl-methylfor-
maldimine (R2 = Me) are relatively high, in agreement with
Figure 3. B3LYP/6-311+G(d,p)/Onsager optimized geometries of
the critical structures for the catalyzed Staudinger mechanism for
the reaction between silylketene and trans-(methoxycarbonyl)-N-
methylformaldimine in the presence of BF3. Lengths are given in
Ångstrøms.
Eur. J. Org. Chem. 2005, 2599–2606
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2603

B. Pelotier, M. Rajzmann, J.-M. Pons, P. Campomanes, R. López, and Tomás L. Sordo,
FULL PAPER
230–400mesh) under low pressure. When appropriate, solvents and
reagents were dried by distillation from the usual drying agent prior
to use. Diethyl ether was distilled from Na/benzophenone and used
fresh. Dichloromethane was distilled from P2O5. IR spectra were
recorded with a Perkin–Elmer 1600 series FTIR spectrometer as
thin films supported on sodium chloride plates. Absorptions are
reported in cm–1 and defined as either strong (s), medium (m) or
weak (w). 1H NMR spectra were recorded in Fourier transform
mode with a Bruker AC 200 (200 MHz) or AM 400 (400 MHz)
spectrometer in [D]chloroform. Chemical shifts are reported in
ppm relative to residual CHCl3 (δ = 7.27 ppm). Multiplicities are
described using the following abbreviations: (s) singlet, (d) doublet,
(t) triplet, (q) quartet, (quint) quintet, (sext) sextet, (sept) septuplet,
(m) multiplet and (br) broad. 13C NMR spectra were recorded with
a Bruker AC 200 (50.3 MHz) or AC 400 (100.6 MHz) spectrometer
in [D]chloroform. Chemical shifts are reported in ppm relative to
the solvent (δ = 77.1 ppm). Multiplicities were determined by using
the distortionless enhancement by polarization transfer (DEPT)
spectral editing technique. The number of coupled protons is indi-
cated by an integer 0–3 in parentheses following the 13C chemical
shift value. Mass spectra were recorded on a VO 70–250-SE or
JEOL MStation JMS-700 spectrometer. Ion mass/charge (m/z) ra-
tios are reported as atomic mass units followed, in parentheses, by
the peak intensity relative to the base peak (100%) and where
shown, the proposed signal assignment. All compounds submitted
for mass spectral analysis were purified by either distillation or col-
umn chromatography and estimated to be at least 95% pure by
NMR and thin-layer chromatography.
the experimental observation that no reaction between the
silylketene 1 and n-hexanaldimine 2 occurs even with the
aid of a catalyst. Indeed, from Table 2 we can see that the
rate-determining energy barriers in solution for the anti and
syn mechanisms are 26.6 and 26.3 kcalmol–1, respectively.
Table 2. Relative B3LYP/6-311+G(d,p) electronic energies in the
gas phase (including ZPVE) (ΔE) and relative B3LYP/6-
311+G(d,p)/PCM//B3LYP/6-311+G(d,p)/Onsager energies (includ-
ing ZPVE corrections from Onsager frequency calculations)
(ΔEsolution) of the most significant structures located for the reac-
tion of silylketene with trans-N-methyl-methylformaldimine in the
presence of BF3.
Structures
ΔE/kcalmol–1
ΔEsolution/kcalmol–1
BF3 + ketene + imine 0.0
0.0
–0.5
4.9
–18.8
7.8
6.0
C1cO + imine
TS1cOanti
IcOanti
TS2cOanti
TS1cOsyn
IcOsyn
–1.9
1.8
–10.8
17.4
3.9
–5.7
16.4
–20.2
6.1
TS2cOsyn
Figure 2S, available in the Supporting Information,
shows the corresponding critical structures. Our theoretical
results clearly indicate that the increase in the rate-de-
termining energy barrier that arises when an electron-rich
imine is employed does not stem from a destabilisation of
the corresponding transition state with respect to the reac-
tants, but from an important stabilisation of the intermedi-
ate preceding this transition state that results from the pres-
ence of the methyl substituent on the carbon atom of the
imine.
(Trimethylsilyl)ketene (1):[8]
2-Ethoxy-1-(trimethylsilyl)acetylene (3.45 g, 24.2 mmol) was
quickly introduced into a flask fitted with a short-path distillation
apparatus pre-warmed to 130–135 °C in an oil bath. The pyrolysis
was followed by the bubbling of gaseous ethylene formed in situ.
Pure (trimethylsilyl)ketene (1) (b.p. 80–82 °C) was obtained as a
colourless oil after distillation (1.71 g, 15.0 mmol, 62%). IR (film):
Conclusions
A catalyzed Staudinger mechanism accounts for both the
overall yield and the diastereoselectivity of the [2+2] cyclo-
addition reaction of silylketenes with imines leading to β-
lactams. Despite the fact that reaction paths corresponding
to the reverse electron demand mechanism were found, they
cannot account for the occurrence of the reaction at low
temperature, mainly because of the high stability of the im-
ine–BF3 complex, or for its diastereoselectivity.
ν = 2110 (s), 1270 (m), 1250 (m), 1055 (w), 845 (s), 755 (w) cm–1.
˜
1H NMR (200 MHz, CDCl3): δ = 1.81 (s, 1 H), 0.19 (s, 9 H) ppm.
13C NMR (50.3 MHz, CDCl3): δ = 179.5 (0), 0.6 (3, 3C), –0.2
(1) ppm.
N-(1Ј-Phenylethyl)hexanaldimine (2):[15]
Experimental Section
All reactions were magnetically stirred and were monitored by thin-
layer chromatography (TLC) using Macherey–Nagel Düren Alug-
ram Si G/UV254 pre-coated aluminium foil sheets, layer thickness
0.25 mm. Compounds were visualised by UV (254 nm) detection
and then with 20 wt-% phosphomolybdic acid (PMA) in ethanol
with heating. Organic extracts were dried with MgSO4 unless other-
wise specified and solvents were evaporated at a water pump using
a Buchi rotary evaporator. Petroleum ether (boiling range 40–
60 °C) and diethyl ether (“ether”) were used as eluents in
chromatography and were distilled before use. Column chromatog-
raphy was performed using Merck silica gel 60 (0.04–0.063 mm,
A solution of freshly distilled hexanal (551 mg, 5.5 mmol) in
CH2Cl2 (15 mL) was introduced, under argon, into a flask contain-
ing 3-Å molecular sieves (1 g). 1-(Phenyl)ethylamine (645 μL,
5 mmol) was then added dropwise and the mixture was stirred at
room temperature for 30 min. The molecular sieves were then fil-
tered off and the solvent evaporated in vacuo to give crude imine
2 as a colourless oil which was used directly in the [2+2] cycload-
dition reaction. IR (film): ν = 1670 (s), 1490 (m), 1380 (m), 760
˜
2604
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2599–2606

Silylated β-Lactams from Silylketenes
FULL PAPER
(m), 700 (s) cm–1. 1H NMR (200 MHz, CDCl3): δ = 7.75 (t, J = of four diastereoisomers. Note: This diastereomeric ratio is based
5.1 Hz, 1 H, H-1), 7.40–7.20 (m, 5 H, H-2ЈЈ. H-3ЈЈ, H-4ЈЈ), 4.28 (q, on the integrals of the peaks corresponding to protons H-3 and H-
J = 6.6 Hz, 1 H, H-1Ј), 2.28 (td, J = 7.2, 5.1 Hz, 2 H, H-2), 1.50 4 of the four diastereoisomers.
(d, J = 6.6 Hz, 3 H, H-2Ј), 1.40–1.20 (m, 6 H, H-3, H-4, H-5), 0.89
IR (film): ν = 1750 (s), 1250 (s), 1185 (s), 1095 (m), 1030 (m), 850
˜
(t, J = 6.4 Hz, 3 H, H-6) ppm. 13C NMR (50.3 MHz, CDCl3): δ =
163.6 (1), 144.9 (0), 128.1 (1, 2C), 126.7 (1), 126.3 (1, 2C), 69.5 (1),
35.6 (2), 31.3 (2), 25.6 (2), 24.4 (3), 22.2 (2), 13.8 (3) ppm.
1
(s), 700 (m) cm–1. Major cis isomer: H NMR (400 MHz, CDCl3):
δ = 7.36–7.28 (m, 5 H, H-2ЈЈ. H-3ЈЈ, H-4ЈЈ), 4.99 (q, J = 7.1 Hz, 1
H, H-1Ј), 4.17–3.95 (m, AB syst, 2 H, H-1ЈЈЈ), 3.91 (d, J = 6.0 Hz,
1 H, H-4), 2.91 (d, J = 6.1 Hz, 1 H, H-3), 1.62 (d, J = 6.9 Hz, 3
H, H-2Ј), 1.60 (m, 2 H, H-2ЈЈЈ), 1.34 (sext, J = 7.4 Hz, 2 H, H-
3ЈЈЈ), 0.93 (t, J = 7.3 Hz, 3 H, H-4ЈЈЈ), 0.12 (s, 9 H, H-1ЈЈЈЈ) ppm.
13C NMR (100.6 MHz, CDCl3): δ = 171.2 (0), 169.0 (0), 139.6 (0),
128.6 (1, 2C), 127.7 (1), 127.6 (1, 2C), 65.2 (2), 52.6 (1), 52.1 (1),
45.8 (1), 30.4 (2), 19.1 (2), 18.8 (3), 13.6 (3), –2.0 (3, 3C) ppm.
Butyl-N-(1Ј-phenylethyl)glyoxaldimine (3):[13]
1
Minor cis isomer (distinct signals): H NMR (400 MHz, CDCl3):
δ = 4.65 or 4.57 (q, J = 7.1 Hz, 1 H, H-1Ј), 2.97 (d, J = 6.3 Hz, 1
H, H-3), 1.79 or 1.75 (d, J = 7.0 Hz, 3 H, H-2Ј) ppm. Major trans
isomer (distinct signals): 1H NMR (400 MHz, CDCl3): δ = 4.65 or
4.57 (q, J = 7.1 Hz, 1 H, H-1Ј), 3.64 (d, J = 2.8 Hz, 1 H, H-4), 2.67
(d, J = 2.7 Hz, 1 H, H-3), 1.79 or 1.75 (d, J = 7.0 Hz, 3 H, H-
2Ј) ppm. Minor trans isomer (distinct signals): 1H NMR
(400 MHz, CDCl3): δ = 3.58 (d, J = 2.5 Hz, 1 H, H-4), 2.69 (d, J
= 2.5 Hz, 1 H, H-3) ppm. LRMS (CI mode): m/z (%) = 348.2 (100)
[M + H]+, 234.2 (15). HRMS (CI mode): found [M + H]+ 348.1994;
C19H29NO3Si + H requires 348.1995.
A
solution of freshly prepared butyl glyoxylate (252 mg,
1.94 mmol) in CH2Cl2 (5 mL) was introduced, under argon, into a
flask containing 3 Å molecular sieves (400 mg). 1-(Phenyl)ethyl-
amine (263 μL, 2.04 mmol) was then added dropwise and the mix-
ture was stirred at room temperature for 30 min. The molecular
sieves were then filtered off and the solvent evaporated in vacuo to
give the crude imine. Purification by flash chromatography (petro-
leum ether/ether, 4:1) led to pure imine 3 (426 mg, 1.83 mmol) in
Computational Details: Quantum chemical computations were per-
formed using the Gaussian 98 series of programs[17] with the hybrid
density functional B3LYP.[18] The geometry of the critical struc-
tures located along the reaction coordinates were fully optimized
using the 6-311+G(d,p) basis set[19] and Schlegel’s algorithm.[20]
Harmonic vibrational frequencies were also calculated at this same
level of theory to characterize the critical points located and to
evaluate the zero-point vibrational energy (ZPVE).
94% yield. IR (film): ν = 1720 (s), 1650 (w), 1290 (m), 1195 (s),
˜
735 (m), 700 (s) cm–1. H NMR (200 MHz, CDCl3): δ = 7.74 (s, 1
1
H, H-2), 7.42–7.20 (m, 5 H, H-2ЈЈ. H-3ЈЈ, H-4ЈЈ), 4.61 (q, J =
6.6 Hz, 1 H, H-1Ј), 4.28 (t, J = 7.2 Hz, 2 H, H-1ЈЈЈ), 1.69 (quint, J
= 7.2 Hz, 2 H, H-2ЈЈЈ), 1.63 (d, J = 6.6 Hz, 3 H, H-2Ј), 1.41 (sext,
J = 7.2 Hz, 2 H, H-3ЈЈЈ), 0.95 (t, J = 7.2 Hz, 3 H, H-4ЈЈЈ) ppm. 13
C
NMR (50.3 MHz, CDCl3): δ = 163.2 (0), 152.2 (1), 142.7 (0), 128.6
(1, 2C), 127.4 (1), 126.8 (1, 2C), 69.6 (1), 65.5 (2), 30.5 (2), 23.8
(3), 19.0 (2), 13.6 (3) ppm.
In the preliminary evaluation of the rate-determining energy barri-
ers for the reactions of methylketene and silylketene with formaldi-
mine we carried out single-point polarisable continuum model
(PCM)[21] calculations on the gas-phase geometries. To take into
account the effects of the condensed phase in the mechanisms of
the reactions of silylketene with trans-(methoxycarbonyl)-N-meth-
ylformaldimine and trans-N-methyl-methylformaldimine we opti-
mized the geometries in solution using the Onsager continuum
model[22] and performed single point calculations on these geome-
tries with the PCM.[21] A relative permittivity of 4.34 was used in
the calculations to simulate ether as solvent. This methodology has
previously proved to be adequate in the study of related cycload-
dition reactions in solution.[23] The minima connected by each lo-
cated transition state (TS) have been checked by intrinsic reaction
coordinate (IRC) calculations[24] at the same level of theory.
3–BF3 Complexes:[16] The most significant changes in the 13C NMR
(50.3 MHz, CDCl3) spectrum of 3 upon complexation of BF3 are
as follows: Signals that have shifted: 163.2, 152.2, 142.7, 69.6 ppm;
new signals:[16] 161.5, 159.7, 157.5, 137.2, 136.2, 135.9, 67.9 ppm.
4-(Butoxycarbonyl)-N-(1Ј-phenylethyl)-3-(trimethylsilyl)-2-azetidin-
one (4):
Supporting Information: Figure 1S, B3LYP/6-311+G(d,p)/Onsager
optimized geometries for the reverse electron demand mechanism
for the reaction between silylketene and trans-(methoxycarbonyl)-
N-methylformaldimine in the presence of BF3. Figure 2S, most sig-
nificant optimized structures located with the B3LYP/6-
311+G(d,p)/Onsager level of theory for the catalyzed Staudinger
mechanism of the reaction between silylketene and trans-N-methyl-
methylformaldimine in the presence of BF3.
To a mixture of imine 3 (117 mg, 0.5 mmol) and (trimethylsilyl)
ketene (1) (68 mg, 0.6 mmol) in Et2O (2 mL) at –50 °C was added
BF3–Et2O (68 μL, 0.55 mmol). The reaction was stirred for 2 h be-
tween –50 and –30 °C before hydrolysis (2 mL H2O). After extrac-
tion with Et2O (3×3 mL), the organic layer was washed with brine
and dried with MgSO4. Evaporation of the solvent in vacuo led to
Acknowledgments
the crude product which was purified by chromatography (petro- We are grateful to Dr. L. Stella (CNRS, Marseille) and Prof. P. J.
leum ether/ether, 4:1) to yield 2-azetidinones 4 (99 mg, 0.28 mmol, Kocienski (Leeds University, UK) for fruitful discussions, and to
57%) as a colourless oil as a 60:15:15:10 (cis/cis/trans/trans) mixture Dr. M. Oblin (Marseille) for preliminary semiempirical calcula-
Eur. J. Org. Chem. 2005, 2599–2606
www.eurjoc.org
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 2605

B. Pelotier, M. Rajzmann, J.-M. Pons, P. Campomanes, R. López, and Tomás L. Sordo,
FULL PAPER
[9]
a) A. Pommier, J.-M. Pons, P. J. Kocienski, L. Wong, Synthesis
1994, 1294–1300; b) P. J. Kocienski, B. Pelotier, J.-M. Pons, H.
Prideaux, J. Chem. Soc., Perkin Trans. 1 1998, 1373–1382.
J.-M. Pons, M. Oblin, A. Pommier, M. Rajzmann, D. Liotard,
J. Am. Chem. Soc. 1997, 119, 3333–3338.
H. Wack, W. J. Drury, A. E. Taggi, D. Ferraris, T. Lectka, Org.
Lett. 1999, 1, 1985–1988.
B. Pelotier, Ph. D. Thesis, Université Paul Cézanne, Marseille,
1999.
tions. The Franco-British collaboration was supported by a RSC
Journal grant to J.-M. P. P. C., R. L. and T. L. S. thank the Princi-
pado de Asturias for financial support (PB02–045).
[10]
[11]
[12]
[1] H. Staudinger, Justus Liebigs Ann. Chem. 1907, 356, 51.
[2] a) G. I. Georg, The Organic Chemistry of β-Lactams, VCH,
New York, 1993; b) C. Palomo, J. M. Aizpurua, I. Ganboa, M.
Oiarbide, Eur. J. Org. Chem. 1999, 3223–3235.
[3] a) J. Pacansky, J. S. Chang, D. W. Brown, W. Schwarz, J. Org.
Chem. 1982, 47, 2233–2234; b) H. W. Moore, G. Hughes, K.
Srinivasachar, M. Fernández, N. V. Nguyen, D. Schoon, A.
Tranne, J. Org. Chem. 1985, 50, 4231–4238; c) J. E. Lynch,
S. M. Riseman, W. L. Laswell, D. M. Tschaen, R. P. Volante,
G. B. Smith, I. Shinkay, J. Org. Chem. 1989, 54, 3792–3796.
[4] a) J. A. Sordo, J. González, T. L. Sordo, J. Am. Chem. Soc.
1992, 114, 6249–6251; b) F. P. Cossío, J. M. Ugalde, X. López,
B. Lecea, C. Palomo, J. Am. Chem. Soc. 1993, 115, 995–1004;
c) F. P. Cossío, A. Arrieta, B. Lecea, J. M. Ugalde, J. Am.
Chem. Soc. 1994, 116, 2085–2093; d) X. Assfeld, M. F. Ruiz-
López, J. González, R. López, J. A. Sordo, T. L. Sordo, J. Com-
put. Chem. 1994, 15, 479–487; e) X. Assfeld, R. López, M. F.
Ruiz-López, J. González, T. L. Sordo, J. A. Sordo, J. Mol.
Struct. (Theochem) 1995, 331, 1–2; f) R. López, M. F. Ruiz-
López, D. Rinaldi, J. A. Sordo, T. L. Sordo, J. Phys. Chem.
1996, 100, 10600–10608; g) D. C. Fang, X. Y. Fu, Int. J. Quan-
tum Chem. 1996, 57, 1107–1114; h) A. Arrieta, B. Lecea, F. P.
Cossío, J. Org. Chem. 1998, 63, 5869–5876; i) T. N. Truong, J.
Phys. Chem. B 1998, 102, 7877–7881; j) P. V. Bharatam, R. S.
Kumar, M. P. Mahajan, Org. Lett. 2000, 2, 2725–2728; k) A.
Arrieta, F. P. Cossío, B. Lecea, J. Org. Chem. 2000, 65, 8458–
8464; l) A. Venturini, J. González, J. Org. Chem. 2002, 67,
9089–9092; m) C. Zhou, D. M. Birney, J. Am. Chem. Soc. 2002,
124, 5231–5241.
[5] For reviews on silylketene chemistry, see: a) T. T. Tidwell, Ke-
tenes, Wiley, New York, 1995, pp. 348–368; b) A. Pommier, P. J.
Kocienski, J.-M. Pons, J. Chem. Soc., Perkin Trans. 1 1998,
2105–2118.
[6] a) W. T. Brady, R. A. Owens, Tetrahedron Lett. 1976, 17, 1553–
1556; b) J. Kron, U. Schubert, J. Organomet. Chem. 1989, 373,
203–219.
[7] a) O. P. Novikova, L. I. Livantsova, G. S. Zaitseva, Zh. Obshch.
Khim. 1989, 59, 2630–2631; b) G. S. Zaitseva, O. P. Novikova,
L. I. Livanstova, V. S. Petrosyan, Y. I. Baukov, Zh. Obshch.
Khim. 1991, 61, 1389–1395.
[8] a) L. L. Shchukovskaya, R. I. Pal’Chick, A. N. Lazarev, Dokl.
Akad. Nauk SSSR 1965, 164, 357 (Chem. Abstr. 1965, 63,
18138g); b) J. L. Loebach, R. L. Danheiser in Encyclopedia of
Reagents for Organic Synthesis (Ed.: L. A. Paquette), Wiley,
New York, 1995, p. 5266.
[13]
[14]
H. Abraham, L. Stella, Tetrahedron 1992, 48, 9707–9718.
A. E. Taggi, A. M. Hafez, H. Wack, B. Young, D. Ferraris, T.
Lectka, J. Am. Chem. Soc. 2002, 124, 6626–6635, and refer-
ences therein.
C. Boga, D. Savoia, A. Umani-Ronchi, Tetrahedron: Asym-
metry 1990, 1, 291–294.
[15]
[16]
Since imine 3 bears several electronic doublets, several Lewis
base–Lewis acid type complexes may be formed.
[17] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Mont-
gomery, R. E. Stratmann, Jr., J. C. Burant, S. Dapprich, J. M.
Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C.
Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson,
P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rab-
uck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Or-
tiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komar-
omi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A.
Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Re-
plogle, J. A. Pople, Gaussian 98, Revision A.6, Gaussian, Inc.,
Pittsburgh PA, 1998.
[18] a) A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) C. Lee,
W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789; c) A. D.
Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[19] W. J. Hehre, L. Radom, J. A. Pople, P. v. R. Schleyer, Ab initio
Molecular Orbital Theory, John Wiley & Sons Inc., New York,
1986.
[20] H. B. Schlegel, J. Comput. Chem. 1982, 3, 214–218.
[21] a) J. Tomasi, M. Persico, Chem. Rev. 1994, 94, 2027–2094; b)
V. Barone, M. Cossi, J. Tomasi, J. Chem. Phys. 1997, 107, 3210–
3221.
[22] L. Onsager, J. Am. Chem. Soc. 1936, 58, 1486–1493.
[23] D. A. Singleton, Y. Wang, H. W. Yang, D. Romo, Angew.
Chem. Int. Ed. 2002, 41, 1572–1575.
[24] a) C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154–
2161; b) C. Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94,
5523–5527.
Received: December 23, 2004
2606
© 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2005, 2599–2606