Synthesis of enantiopure highly functionalized pyrrolizines and indolizines from natural α-amino acids: An experimental and theoretical investigation

Article abstract of DOI:10.1002/ejoc.200701217

Variously substituted enantiopure 2-benzylamino-1-hydroxymethyl-2,3- dihydro-1H-pyrrolizines and 6-benzylamino-5,6,7,8-tetrahydroindolizin-8-ols have been prepared. The reaction sequence starts from L-α-amino acids and involves an intramolecular cycloaddition of pyrrole-based nitrone intermediates. A theoretical investigation at the HCTH/6-311+G(d,p)//HCTH/6-31+G(d) level of theory was performed with the aim of rationalizing the effects of substituents on the regiochemistry of the cycloaddition reaction. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejoc.200701217

FULL PAPER
DOI: 10.1002/ejoc.200701217
Synthesis of Enantiopure Highly Functionalized Pyrrolizines and Indolizines
from Natural α-Amino Acids: An Experimental and Theoretical Investigation
Elena Borsini,[a] Gianluigi Broggini,*[a,b] Alessandro Contini,*[b,c] and Gaetano Zecchi[a,b]
Keywords: Cycloaddition / Nitrones / Regioselectivity / Density functional calculations / Pi interactions
Variously substituted enantiopure 2-benzylamino-1-hy-
droxymethyl-2,3-dihydro-1H-pyrrolizines and 6-benzyl-
amino-5,6,7,8-tetrahydroindolizin-8-ols have been prepared.
HCTH/6-311+G(d,p)//HCTH/6-31+G(d) level of theory was
performed with the aim of rationalizing the effects of substit-
uents on the regiochemistry of the cycloaddition reaction.
The reaction sequence starts from L-α-amino acids and in-
volves an intramolecular cycloaddition of pyrrole-based
nitrone intermediates. A theoretical investigation at the
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
Introduction
Pyrrolizine and indolizine skeletons, partially or totally
saturated, are present in a large array of alkaloids and re-
lated unnatural compounds. Such compounds exhibit po-
tent biological activities, particularly as glycosidase inhibi-
tors. They consequently have a cytostatic effect that has po-
tential applications in antitumour and antiviral chemother-
apy.[1] For the past few years, we have had an interest in
the synthesis of such molecules, with intramolecular nitrone
cycloaddition reactions[2] being our synthetic methodology
of choice. We have succeeded in the preparation of both
racemic and enantiopure pyrrolizidines and indolizidines
endowed with amino and hydroxy groups by using nitrones
derived from 1-allylpyrrole-2-carbaldehydes 1.[3] We be-
lieved that using nitrones derived from 1-(formylmethyl)-2-
vinylpyrroles 2 would allow flexibility in the positioning of
these important functionalities on the parent skeletons. On
the basis of the retrosynthetic analysis of such pyrroles, nat-
ural α-amino acids were devised as building blocks for the
construction of the pyrrole nucleus, allowing access to chi-
ral non-racemic species.
Although intramolecular nitrone cycloaddition reactions
have been known for a long time and their synthetic versa-
tility is well established,[2] the regio- and diastereoselective
outcomes of these reactions are a consequence of a subtle
interplay of electronic factors and geometric restraints, pre-
cluding a sound prediction in most cases. Consequently, we
present our experimental work along with the results of an
extensive and systematic theoretical investigation.
Results and Discussion
Synthesis
Following the procedure reported in the literature for
building the pyrrole nucleus on a primary amino group,[4]
we first transformed a series of commercial α-amino acids
into chiral non-racemic derivatives (S)-3a–e (Scheme 1). As
the heterocyclic ring of the latter required a vinyl residue in
the 2-position as a potential dipolarophile, a formyl group
was selectively introduced into this position by applying
Vilsmeier conditions to (S)-3a–e.[5] For the conversion of
the intermediates (S)-4a–e into the desired vinyl derivatives
(S)-5a–e, after considering the possible base-promoted race-
mization under classical Wittig reaction conditions, we
chose to carry out the olefination process by using zinc,
CH2I2 and TiCl4 in the presence of a catalytic amount of
PbCl2.[6] In order to introduce the nitrone moiety, the
methyl ester group of compounds (S)-5a–e was reduced
with DIBAL-H to give the aldehydes (S)-6a–e. However,
these aldehydes show a high tendency to degrade, even at a
[a] Dipartimento di Scienze Chimiche e Ambientali, Università
dell’Insubria,
Via Valleggio 11, 22100 Como, Italy
Fax: +39-0312386449
E-mail: gianluigi.broggini@uninsubria.it
[b] Centro Interuniversitario di Ricerca sulle Reazioni Pericicliche
e Sintesi di Sistemi Etero e Carbociclici,
Via Venezian 21, 20133 Milano, Italy
[c] Istituto di Chimica Organica “A. Marchesini”, Facoltà di Farm-
acia, Università di Milano,
Via Venezian 21, 20133 Milano, Italy
Fax: +39-0250314476
E-mail: alessandro.contini@unimi.it
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
2808
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2808–2816

Pyrrolizines and Indolizines from α-Amino Acids
Scheme 1. Cycloaddition process and isoxazolidine ring-opening step.
low temperature, so their use as precursors of nitrones was and indolizines.[8] The reductive cleavage of the N–O bond
crucial. Therefore we sought to use a one-pot protocol with was performed by heating at reflux compounds 8a–e and
no isolation of intermediates (S)-6a–e, namely, by treating 9d,e in dry THF with LiAlH4 for 24 h. This protocol led to
the methyl esters (S)-5a–e with DIBAL-H in toluene at 2,3-dihydropyrrolizines 10a–e and 5,6,7,8-tetrahydroindoli-
–78 °C for 1 h and then adding N-benzylhydroxylamine and zines 11d,e, respectively. The enantiomeric purity of the so-
heating the mixture at reflux for 24 h. Under these condi- obtained 1,3-amino alcohols was verified to be higher than
1
tions, the products resulting from the intramolecular cyclo- 98% in one case, namely 10b, by comparing its H NMR
addition reaction were obtained directly. The isolated yields spectrum with that of (Ϯ)-10b (synthesized starting from
were moderate, but it must be underlined that they refer to the racemic valine methyl ester) recorded in the presence of
three steps involving the labile aldehydes (S)-6a–e. The final (R)-O-acetylmandelic acid.
outcome, however, was somewhat different for the different
The most striking feature of the above results is the
substrates. The fused-ring cycloadducts 8a–e were all regiochemical outcome of the intramolecular cycloaddition
formed from the nitrones (S)-7a–e, whereas the bridged- in which the role of the R substituent seems at first glance
ring compounds 9d,e were observed only with (S)-7d,e.
incongruous. Indeed, it is quite hard to explain the ob-
The two regioisomeric structures were distinguished by served regioselectivity by only invoking the different steric
1H and 13C NMR spectroscopy, the different chemical hindrances of the substituents because, for example, 7c and
shifts of the endocyclic methylene group being particularly 7d (R = iBu and Bn, respectively) provide quite a different
diagnostic. On the other hand, by analogy to previous re- product distribution even if the encumbrance of the R
sults concerning the intramolecular cycloaddition of C-(1- group is comparable. A contrasting statement could be
allyl-2-imidazolyl)nitrones,[7] the stereochemical disposition made for 7a (R = Me) and 7c for which a remarkable differ-
of the isoxazolidine ring was assumed to be trans with re- ence in steric hindrance results in a similar regioselective
spect to the R substituent. The steric hindrance exerted by outcome. At the same time, it must be stressed that the
the R group plausibly justifies such a stereochemical prefer- regiochemistry of the cycloaddition determines whether the
ence.
pyrrolizine or indolizine derivatives are formed as the final
Next, the isoxazolidine ring of the cycloadducts was ma- products. Such considerations prompted us to undertake
nipulated in order to achieve functionalized pyrrolizines the following computational study.
Eur. J. Org. Chem. 2008, 2808–2816
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2809

E. Borsini, G. Broggini, A. Contini, G. Zecchi
FULL PAPER
Theoretical Calculations
sity functionals were compared with CCSD(T) calculations
describing dispersion-bound homomolecular dimers,[15] in-
Although several theoretical investigations on intermo- cluding benzene dimers already described by high-level cal-
lecular nitrone cycloadditions have been reported in the lit- culations.[16] Such a comparative investigation evidenced
erature,[9] fewer studies concerning intramolecular cycload- that the PW91 and the HCTH407 functionals were able to
ditions are available.[7,10] However, none of the reported predict an attractive interaction in all three possible ben-
studies provided us with an unequivocal interpretation of zene dimers.[17] Another recent article reported that the hy-
our experimental results, thus the need to perform an “ad brid density functional BH&H also provided a good de-
hoc” investigation. As the B3LYP/6-31G* methodology, scription of the π–π interactions, even if the authors as-
which combines good treatment of the electron correlation cribed the good performance to a fortuitous cancellation of
with a moderate need for computational power, was exten- errors.[18] For the reasons just mentioned, we decided to
sively adopted in the previously cited articles, it was our model the cycloadditions of 7e by using the HCTH, PW91
first choice for the geometry optimization of reactants 7a– and BH&H functionals. As the best concordance between
e, products 8a–e and 9a–e and the corresponding transition theoretical calculations and experimental evidence was ob-
structures TS-8a–e and TS-9a–e. In fact, the energetic re- tained with the HCTH functional, we undertook a full re-
sults obtained were not consistent with the experimental optimization of all the reactants, transition-state structures
evidence (see the Supporting Information for the B3LYP/6- and products at the HCTH/6-31+G(d) level of theory and
31G* energies). Indeed, in every modelled reaction the single-point energies were recalculated at the HCTH/6-
fused-ring structure 8 was calculated to be the kinetic 311+G(d,p) level in the gas phase or in solution by the
(∆∆E‡
= 2.5, 3.6, 2.8, 1.6 and 2.3 kcal/mol for com- CPCM model for toluene. In fact, the HCTH407 functional
TS9–TS8
pounds a, b, c, d and e, respectively) as well as the thermo- provided, at an affordable computational expense, a net im-
dynamic product with the only exception the methyl-substi- provement in the description of the kinetics and thermody-
tuted compound (∆∆E9–8 = –1.5, 2.0, 0.8, 0.01 and 0.7 kcal/ namics of the reactions considered herein. Indeed, the ener-
mol for compounds a, b, c, d and e, respectively). Therefore getic results summarized in Table 1 (see the Supporting In-
we turned our attention to other computational approaches. formation for the all energies) fit well with the experimen-
Recently it has been reported that diffuse functions are tally observed regiochemistry, which confirms that a good
necessary for a proper description of cycloaddition reac- description of the dispersive interactions is essential to the
tions involving nitrones, at least as single-point energy accuracy of our computational model.
evaluations.[9a] Moreover, in some cases, the inclusion of the
solvent effect was important for a reliable estimation of the
Table 1. Activation and reaction free-energy differences [kcal/mol]
energetic aspect of these processes.[9b,9e] Indeed, Kuznetsov
for the formation of 9 and 8.[a,b]
and Kukushkin showed that the reactivity of cyclic and acy-
Compound
(R)
Gas phase
∆∆G
Solvent
∆∆G
clic nitrones in 1,3-dipolar cycloadditions was well de-
scribed by single-point CPCM calculations.[9b] For this
reason we recomputed the energy of all the stationary
points previously located by single-point energy evaluations
at the B3LYP/6-311+G(d,p) level of theory, including the
solvent contribution by the CPCM model for toluene. Un-
fortunately, even at this more accurate theoretical level, the
computed kinetic and thermodynamic data were overall in
discordance with the observed regiochemical outcomes (see
the Supporting Information). At the same time, a thorough
analysis of the optimized geometries showed the possibility
that π–π or CH–π interactions between the R substituent
and the nitrone benzyl moiety could play some role in de-
termining the regioselectivity in the transition state. The im-
∆∆G‡
∆∆G‡
a (Me)
b (iPr)
c (iBu)
d (Bn)
e (Ph)
0.25
1.12
0.90
–2.21
–0.38
0.96
2.98
4.03
1.76
1.49
–0.81
0.38
–0.98
–3.54
–2.13
0.20
3.25
1.80
1.92
0.61
[a] Gas-phase and solution Gibbs free energies are calculated as the
sum of the single-point HCTH/6-311+G(d,p) energies obtained in
vacuo and in solution and the free-energy corrections resulting
from the thermochemical analysis. [b] ∆∆G‡ and ∆∆G are the total
Gibbs free-energy differences between TS-9 and TS-8 and the prod-
ucts 9 and 8, respectively.
Note that the gas-phase energies are slightly more con-
portance of π–π and CH–π interactions in governing the cordant with experiments than those computed in solution,
selectivity of cycloaddition reactions is documented in the probably because in this latter case an overstabilization of
literature.[11,12] Furthermore, it is known that the descrip- TS-8a,e arose from the lack of explicit π–π solute–solvent
tion of dispersive interactions by DFT methods is quite interactions in the implicit solvation model. However, be-
challenging and that in dispersion-bound systems the cause all the cycloaddition reactions were performed in an
B3LYP functional often predicts a repulsive instead of an apolar solvent such as toluene, gas-phase results can be
attractive interaction.[13] Although several theoretical considered accurate enough for further analyses. The first
methods that accurately describe dispersive interactions are evidence for the two reaction mechanisms concerns the de-
reported in the literature,[14] most of them could not be ap- gree of synchronicity. In a recent article we defined the de-
plied to our investigations owing to either the large de- gree of synchronicity for cycloaddition reactions involving
mands on computational power or the need for non-stan- heteroatoms as the difference between the ratios of the
dard codes. However, in a recent theoretical study 25 den- forming bond lengths in the TS and the corresponding
2810
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2808–2816

Pyrrolizines and Indolizines from α-Amino Acids
bond lengths in the product, that is, ∆dTS/P = |(C–O)TS/(C– C8-linked hydrogen and such an eclipsed conformation is
O)P – (C–C)TS/(C–C)P|. Note that a complete synchronous due to the attempt by the phenyl ring to maximize the at-
reaction gives ∆dTS/P = 0.[18]
tractive π–π interaction with the nitrone benzyl group. Con-
Analysis of the vibrational motion associated with the versely, no unfavourable interactions are observed within
unique imaginary frequency computed for TS-8 and TS-9 TS-9e, thus explaining its greater stability. Therefore, be-
as well as IRC analyses show that both reaction paths are cause 8e is the thermodynamic product and 9e is slightly
concerted but, as shown in Figure 1 (see Table 1 of the Sup- kinetically favoured, a mixture of regioisomers can be ex-
porting Information for specific values), the cycloadditions pected. In conclusion, the experimentally observed out-
leading to the fused-ring compounds 8 are quite synchro- comes can be rationalized for both nitrones 7a and 7e.
nous (0.17 Ͻ ∆dTS/P Ͻ 0.21, 8d and 8a, respectively),
whereas the reactions providing the bridged-ring products
9 are decidedly asynchronous (0.71 Ͻ ∆dTS/P Ͻ 0.77, 9d
and 9b,e, respectively). Noticeably for Diels–Alder cycload-
ditions to unsymmetrically substituted dienophiles, it has
been reported that the more asynchronous is the TS, the
lower is the activation barrier.[19] However, owing to the
lack of rigorous theoretical justification, such a statement
has become merely a sort of empirical rule that holds for
several DA reactions.[18,20] To the best of our knowledge, a
correlation between synchronicity and activation barrier in
dipolar cycloadditions has never been reported and, seeking
an analogy between 1,3-dipolar and DA cycloadditions, the
most asynchronous pathways should be the favoured ones.
However, neither experimental results nor the energies re-
ported in Table 1 can confirm or exclude any correlation
between synchronicity and activation barriers. Indeed, the
steric hindrance of the R substituent or its ability to estab-
lish attractive dispersive interactions can modulate the sta-
bility of the regioisomeric TSs, so raising or lowering the
activation barriers of the competing cycloadditions. The
steric effect emerges by comparing the methyl-substituted
TS-8,9a with the phenyl-substituted TS-8,9e. Despite the
low encumbrance of the methyl group, Figure 2 shows that
an unfavourable steric interaction between the methyl and
Figure 2. Transition states for the formation of 8a,d,e and 9a,d,e.
vinyl groups destabilizes TS-9a and causes the slight ener-
getic preference computed for TS-8a. Together with the
thermodynamic preference for 8a, only one product is thus
expected. On the other hand, the phenyl-substituted TS-8e
presents a steric clash between the ortho-hydrogen and the
Additional Probing Work
In order to determine the intrinsic stability difference be-
tween fused- and bridged-ring TSs, we should eliminate the
steric effect of the R substituent. For this reason, we also
modelled the two possible cycloaddition paths of the un-
substituted nitrone 15. The predicted energy difference be-
tween the corresponding cycloadducts 16 and 17 was
0.6 kcal/mol in favour of the fused-ring compound 16,
whereas TS-17 was favoured over TS-16 with a ∆∆G‡
17–16
of –1.9 kcal/mol. On the basis of these theoretical calcula-
tions, a regioisomeric mixture with the kinetically favoured
product 17 predominating can be expected. Moreover, by
eliminating the effect of the R substituent, a marked prefer-
ence for the most asynchronous TS-17 emerges, which sug-
gests that the relationship between synchronicity and acti-
vation barriers observed for DA reactions also holds in
principle for 1,3-dipolar cycloadditions. This prediction was
confirmed experimentally by the synthetic sequence illus-
trated in Scheme 2.
Figure 1. Average TS bond lengths [Å] and degrees of synchronicity
∆dTS/P for TS-8 and TS-9.
Eur. J. Org. Chem. 2008, 2808–2816
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2811

E. Borsini, G. Broggini, A. Contini, G. Zecchi
FULL PAPER
Scheme 3. Cycloaddition reaction of the N-methyl-substituted
nitrone.
Conclusions
In conclusion, we have described a systematic investiga-
tion, both experimental and theoretical, of the intramolecu-
lar cycloaddition of a series of variously substituted, pyr-
role-based nitrones. Depending on the regiochemistry of the
cycloaddition, the final products of the synthetic sequence
were
2-benzylamino-2,3-dihydro-1H-pyrrolizine-1-meth-
anols or 6-benzylamino-5,6,7,8-tetrahydroindolizin-8-ols.
As the regiochemistry of the cycloaddition determines the
synthetic targets, its origin was investigated theoretically at
different levels of sophistication. Our calculations indicate
that the competition between the regioisomeric paths is es-
sentially governed by spatial interactions between the sub-
stituents: steric repulsions in the case of alkyl groups and
dispersive attractions in the case of aryl groups. Probing
experiments were carried out to gain evidence for this ratio-
nalization. Furthermore, by constructing the pyrrole nu-
cleus from commercially available -α-amino acids, we ac-
quired the final products in enantiopure form.
Scheme 2. The cycloaddition process and computed energies for
the formation of regioisomeric compounds 16 and 17 (solution en-
ergies are reported in parentheses).
Experimental Section
Turning to the behaviour of the benzyl-substituted
nitrone 7d, steric hindrance by itself is not able to explain
the experimental evidence because the fused-ring product
8d is expected owing to the bulkiness of the benzyl group.
However, as shown in Figure 2, TS-8d exhibits a T-shaped
Melting points were determined by the capillary method with a
Büchi B-540 apparatus and are uncorrected. Optical rotations were
measured with a Jasco P-1010 polarimeter. NMR spectra were re-
corded with an AVANCE 400 Bruker spectrometer at 400 MHz for
CH–π interaction between the two aromatic rings that 1H NMR and 100 MHz for 13C NMR spectroscopy. Chemical
shifts are given as δ values in ppm relative to residual solvent peaks
(CHCl3 or C6H6) as the internal reference. 13C NMR spectra are
1H-decoupled and the multiplicities were determined by the APT
pulse sequence. IR spectra were measured with a Jasco FT/IR 5300
spectrometer. Mass spectra were recorded with a VG-7070 EQ-HF
instrument. Elemental analyses were executed using a Perkin-Elmer
CHN Analyzer Series II 2400 instrument. Thin-layer chromato-
graphic separations were performed on Merck silica-gel 60-F254
precoated plates. Preparative separations were performed by flash
chromatography using Merck silica gel (0.035–0.070 mm).
forces the benzylic methylene to clash sterically with the
pyrrole 5-H.[21] On the other hand, the PD-shaped π–π in-
teraction between the two aromatic rings observed in TS-9d
locks the C-5-linked benzyl group into a sterically allowed
position, thus justifying the strong preference for the
bridged TS-9d. Seeking further proof of the role of weak
hydrogen or CH–π interactions in determining the regio-
chemistry of cycloaddition, we hypothesized that the ab-
sence of the N-benzyl group should unlock the C-5 benzyl
and enhance its steric hindrance, thus raising the amount
of the fused-ring product. In support of this view, we
treated the intermediate 6d with N-methylhydroxylamine
and obtained the fused-ring compound 19 as the only prod-
uct (Scheme 3).
Theoretical Calculations: Reactants 7a,e and 15, products 8a,e, 9a,e,
16 and 17 and all the corresponding TSs were fully optimized in
the gas phase at the HCTH/6-31+G(d) level of theory. As the con-
formations of the R substituents and the nitrone benzyl group were
found to be critical, a rigid conformational search was performed
2812
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2808–2816

Pyrrolizines and Indolizines from α-Amino Acids
on TSs and products at the same level of theory. Vibrational fre-
quencies were computed at the same level of theory in order to
define optimized geometries as minima (no imaginary frequencies)
or transition states (a unique imaginary frequency corresponding
to the vibrational stretching of the forming/breaking bonds) and to
calculate the ZPVEs and thermochemical corrections to electronic
energies (1 atm, 298.15 K). Single-point energies were computed at
the HCTH/6-311+G(d,p) level of theory in the gas phase or in
toluene with the CPCM solvent model.[22] All calculations were
performed with the Gaussian 03 package.[23]
(c = 0.91, CHCl3). 1H NMR (400 MHz, CDCl3): δ = 3.33 (dd, J =
8.7, 13.8 Hz, 1 H), 3.57 (dd, J = 6.5, 13.8 Hz, 1 H), 3.76 (s, 3 H),
5.05 (dd, J = 6.5, 8.7 Hz, 1 H), 5.09 (dd, J = 1.3, 11.1 Hz, 1 H),
5.52 (dd, J = 1.3, 17.2 Hz, 1 H), 6.27 (d, J = 3.5 Hz, 1 H), 6.42 (d,
J = 3.5 Hz, 1 H), 6.48 (dd, J = 11.1, 17.2 Hz, 1 H), 6.93–6.95
(m, 1 H), 7.08–7.14 (m, 2 H), 7.25–7.35 (m, 3 H) ppm. 13C NMR
(100 MHz, CDCl3): δ = 39.6 (t), 53.0 (q), 60.4 (d), 107.1 (d), 109.5
(d), 113.3 (t), 120.5 (d), 125.3 (d), 127.5 (d), 129.0 (d), 129.3 (d),
132.8 (s), 136.7 (s), 171.0 (s) ppm. MS: m/z = 255 [M]+. C16H17NO2
(255.32): calcd. C 75.27, H 6.71, N 5.49; found C 75.03, H 6.98, N
5.29.
General Procedure for the Preparation of 2-Vinylpyrroles 5a–e:
CH2I2 (1.0 mL, 12.4 mmol) and TiCl4 (8.0 mL of 1  solution in
CH2Cl2) were added to a suspension of zinc (activated in 1  HCl)
(2.4 g, 37.3 mmol) and PbCl2 (104.0 mg, 0.37 mmol) in dry THF
(25 mL) under N2. After stirring for 30 min, a solution of 4a–e
(8.2 mmol) in dry THF (7 mL) was added dropwise and the result-
ant suspension was stirred overnight at room temperature. The
mixture was diluted with AcOEt (25 mL) and washed with a satu-
rated aqueous solution of NH4Cl (55 mL). The layers were sepa-
rated and the aqueous phase was extracted with AcOEt
(3ϫ20 mL). The organic phases were collected, washed with a 1 
solution of Na2S2O3 and dried with Na2SO4. The solvent was evap-
orated and the crude mixture was purified by chromatography on
a silica gel column with light petroleum/AcOEt (4:1) as eluent to
give 5a–e.
Methyl (S)-2-(2-Vinylpyrrol-1-yl)-2-phenylacetate (5e): Yield 1.25 g
(63%); pale-yellow oil. IR (Nujol): ν = 1745 cm–1. [α]23 = –4.9 (c
˜
D
= 0.67, CHCl3). 1H NMR (400 MHz, CDCl3): δ = 3.84 (s, 3 H),
5.12 (d, J = 11.2 Hz, 1 H), 5.55 (d, J = 17.2 Hz, 1 H), 6.08 (s, 1
H), 6.17 (t, J = 3.3 Hz, 1 H), 6.42 (d, J = 3.3 Hz, 1 H), 6.48 (dd,
J = 11.2, 17.2 Hz, 1 H), 6.65 (d, J = 3.2 Hz, 1 H), 7.28–7.30 (m, 2
H), 7.40–7.44 (m, 3 H) ppm. 13C NMR (100 MHz, CDCl3): δ =
53.2 (q), 62.7 (d), 107.7 (d), 109.1 (d), 113.2 (t), 122.0 (d), 125.3
(d), 128.4 (d), 129.3 (d), 129.4 (d), 132.8 (s), 135.3 (s), 170.4
(s) ppm. MS: m/z = 241 [M]+. C15H15NO2 (241.29): calcd. C 74.67,
H 6.27, N 5.80; found C 74.75, H 6.36, N 5.89.
General Procedure for the One-Pot Reduction of 5a–e and 13 and
Cycloaddition of the Nitrones Arising from the Aldehyde Intermedi-
ates: Preparation of mixture A: a solution of 5a–e (or 13) (4 mmol)
in dry toluene (75 mL) was cooled to –78 °C under nitrogen and a
1.5  DIBAL-H solution (3.5 mL, 5.2 mmol) was added dropwise.
The mixture was stirred at –78 °C for 1 h. Preparation of mixture
B: a suspension of N-benzylhydroxylamine hydrochloride (0.64 g,
4 mmol), NaHCO3 (0.68 g, 8 mmol) and MgSO4 (4.44 g, 37 mmol)
in toluene (60 mL) was stirred for 15 min at room temperature.
Mixture B was slowly added to mixture A and the resultant suspen-
sion was heated at reflux for 24 h. After cooling, the mixture was
washed with 1  HCl (150 mL) and the aqueous layer was extracted
with AcOEt (3ϫ150 mL). The collected organic phases were
washed with H2O, 5% aqueous solution of NaHCO3 and dried
with Na2SO4. The solvent was evaporated and the crude mixture
was purified by chromatography on a silica gel column with light
petroleum/AcOEt (4:1) as eluent to give 8a–e and 9d,e [or 16 and
17, or 19 (in the case of 19, N-methylhydroxylamine hydrochloride
was used in the preparation of mixture B)].
Methyl (S)-2-(2-Vinylpyrrol-1-yl)propanoate (5a): Yield 778 mg
(53%); colourless oil. IR (Nujol): ν = 1749 cm–1. [α]23 = –52.2 (c =
˜
D
0.46, CHCl3). 1H NMR (400 MHz, CDCl3): δ = 1.75 (d, J =
7.2 Hz, 3 H), 3.75 (s, 3 H), 4.97 (q, J = 7.2 Hz, 1 H), 5.12 (dd, J =
1.0, 11.1 Hz, 1 H), 5.54 (dd, J = 1.0, 17.2 Hz, 1 H), 6.22 (t, J =
3.2 Hz, 1 H), 6.56 (dd, J = 11.1, 17.2 Hz, 1 H), 6.82 (dd, J = 1.6,
3.2 Hz, 1 H), 7.02 (dd, J = 1.6, 3.2 Hz, 1 H) ppm. 13C NMR
(100 MHz, CDCl3): δ = 18.5 (q), 30.7 (q), 54.1 (d), 107.3 (d), 109.2
(d), 113.0 (t), 119.8 (d), 125.9 (d), 132.5 (s), 172.0 (s) ppm. MS: m/z
= 179 [M]+. C10H13NO2 (179.22): calcd. C 67.02, H 7.31, N 7.82;
found C 66.95, H 7.55, N 7.99.
Methyl (S)-2-(2-Vinylpyrrol-1-yl)-3-methylbutanoate (5b): Yield
509 mg (30%); colourless oil. IR (Nujol): ν = 1743 cm–1. [α]23
=
˜
D
1
–127.4 (c = 0.75, CHCl3). H NMR (400 MHz, CDCl3): δ = 0.73
(d, J = 6.6 Hz, 3 H), 1.03 (d, J = 6.6 Hz, 3 H), 2.44 (dqq, J = 10.5,
6.6, 6.6 Hz, 1 H), 3.74 (s, 3 H), 4.35 (d, J = 10.5 Hz, 1 H), 5.11
(dd, J = 1.5, 11.2 Hz, 1 H), 5.53 (dd, J = 1.5, 17.2 Hz, 1 H), 6.19
(dd, J = 3.2, 3.5 Hz, 1 H), 6.36 (dd, J = 1.4, 3.2 Hz, 1 H), 6.62 (dd,
J = 11.2, 17.2 Hz, 1 H), 6.91 (dd, J = 1.4, 3.5 Hz, 1 H) ppm. 13C
NMR (100 MHz, CDCl3): δ = 19.1 (q), 19.9 (q), 30.7 (q), 52.6 (d),
65.0 (d), 106.7 (d), 109.5 (d), 113.1 (t), 120.3 (d), 125.9 (d), 133.1
(s), 171.2 (s) ppm. MS: m/z = 207 [M]+. C12H17NO2 (207.27): calcd.
C 69.54, H 8.27, N 6.76; found C 69.35, H 8.41, N 6.97.
(3aS,8S,8aR)-1-Benzyl-8-methyl-3,3a,8,8a-tetrahydro-1H-isoxazolo-
[4,3-a]pyrrolizine (8a): Yield 244 mg (24%); colourless oil. [α]2D3
=
+78.1 (c = 0.46, CHCl3). 1H NMR (400 MHz, CDCl3): δ = 1.34
(d, J = 6.6 Hz, 3 H), 3.85, 4.17 (AB system, J = 12.8 Hz, 2 H),
3.89 (dd, J = 4.5, 7.7 Hz, 1 H), 3.91 (dd, J = 3.7, 8.0 Hz, 1 H), 4.15
(dq, J = 3.7, 6.6 Hz, 1 H), 4.17 (dt, J = 4.5, 7.7 Hz, 1 H), 4.33 (t,
J = 7.7 Hz, 1 H), 5.84 (d, J = 3.3 Hz, 1 H), 6.25 (dd, J = 2.9,
3.3 Hz, 1 H), 6.50 (d, J = 2.9 Hz, 1 H), 7.30–7.45 (m, 5 H) ppm.
13C NMR (100 MHz, CDCl3): δ = 20.6 (q), 46.0 (d), 60.7 (t), 62.8
(d), 72.4 (t), 81.6 (d), 99.9 (d), 112.8 (d), 113.3 (d), 127.5 (d), 128.3
(d), 129.5 (d), 136.8 (s), 137.0 (s) ppm. MS: m/z = 254 [M]+.
C16H18N2O (254.33): calcd. C 75.56, H 7.13, N 11.01; found C
75.39, H 6.97, N 10.79.
Methyl (S)-2-(2-Vinylpyrrol-1-yl)-4-methylpentanoate (5c): Yield
744 mg (41%); pale-yellow oil. IR (Nujol): ν = 1748 cm–1. [α]23
=
˜
D
–65.3 (c = 0.29, CHCl3). 1H NMR (400 MHz, CDCl3): δ = 0.96 (d,
J = 6.8 Hz, 6 H), 1.29–1.51 (m, 1 H), 1.79–2.10 (m, 2 H), 3.75 (s,
3 H), 4.90 (dd, J = 6.1, 9.6 Hz, 1 H), 5.14 (dd, J = 1.4, 11.1 Hz, 1
H), 5.56 (dd, J = 1.4, 17.2 Hz, 1 H), 6.23 (t, J = 3.3 Hz, 1 H), 6.42
(d, J = 3.3 Hz, 1 H), 6.63 (dd, J = 11.1, 17.2 Hz, 1 H), 6.85 (d, J
= 3.2 Hz, 1 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 22.1 (q),
23.2 (q), 25.0 (d), 41.7 (t), 52.9 (q), 56.8 (d), 107.1 (d), 109.4 (d),
113.1 (t), 120.3 (d), 125.5 (d), 132.8 (s), 171.9 (s) ppm. MS: m/z =
221 [M]+. C13H19NO2 (221.30): calcd. C 70.56, H 8.65, N 6.33;
found C 70.37, H 8.54, N 6.51.
(3aS,8S,8aR)-1-Benzyl-8-isopropyl-3,3a,8,8a-tetrahydro-1H-
isoxazolo[4,3-a]pyrrolizine (8b): Yield 192 mg (17%); colourless oil.
1
[α]2D3 = +59.6 (c = 0.13, CHCl3). H NMR (400 MHz, C6D6): δ =
0.55 (d, J = 6.8 Hz, 3 H), 0.61 (d, J = 6.8 Hz, 3 H), 1.78 (dqq, J =
3.0, 6.8, 6.8 Hz, 1 H), 3.66 (dd, J = 2.6, 8.2 Hz, 1 H), 3.69, 3.96
(AB system, J = 12.5 Hz, 2 H), 3.74 (ddd, J = 3.7, 7.5, 8.2 Hz, 1
H), 3.85 (dd, J = 3.7, 7.8 Hz, 1 H), 3.96 (dd, J = 7.5, 7.8 Hz, 1 H),
4.02 (dd, J = 2.6, 3.0 Hz, 1 H), 6.02 (dd, J = 2.9, 3.2 Hz, 1 H), 6.49
Methyl (S)-2-(2-Vinylpyrrol-1-yl)-3-phenylpropanoate (5d): Yield
565 mg (27%); yellow oil. IR (Nujol): ν = 1752 cm–1. [α]23 = –28.8
˜
D
Eur. J. Org. Chem. 2008, 2808–2816
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2813

E. Borsini, G. Broggini, A. Contini, G. Zecchi
FULL PAPER
(dd, J = 1.2, 2.9 Hz, 1 H), 6.58 (dd, J = 1.2, 3.2 Hz, 1 H), 7.23–
(1S,4R,5S)-3-Benzyl-5-phenyl-4,5-dihydro-1H,3H-1,4-methano-
7.45 (m, 5 H) ppm. 13C NMR (100 MHz, C6D6): δ = 17.2 (q), 18.2 pyrrolo[1,2-e][1,2,5]oxadiazepine (9e): Yield 126 mg (10 %); white
(q), 32.3 (d), 46.7 (d), 60.9 (t), 68.6 (d), 72.6 (t), 75.8 (d), 98.9 (d),
113.1 (d), 113.4 (d), 128.0 (d), 128.8 (d), 129.7 (d), 137.0 (s), 137.9
(s) ppm. MS: m/z = 282 [M]+. C18H22N2O (282.39): calcd. C 76.56,
H 7.85, N 9.92; found C 76.55, H 8.07, N 9.79.
crystals; m.p. 118–120 °C (diisopropyl ether). [α]2D3 = +7.5 (c = 0.33,
CHCl3). 1H NMR (400 MHz, CDCl3): δ = 2.19–2.23 (m, 1 H), 2.34
(d, J = 11.4 Hz, 1 H), 3.66 (d, J = 4.7 Hz, 1 H), 3.87, 4.33 (AB
system, J = 12.6 Hz, 2 H), 5.29 (d, J = 4.7 Hz, 1 H), 5.34 (s, 1 H),
6.11 (dd, J = 1.1, 3.4 Hz, 1 H), 6.16 (dd, J = 3.0, 3.4 Hz, 1 H), 6.44
(dd, J = 1.1, 3.0 Hz, 1 H), 6.73–6.75 (m, 3 H), 7.25–7.44 (m, 5 H),
7.47–7.49 (m, 2 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 29.3
(t), 63.5 (t), 66.3 (d), 66.7 (d), 71.3 (d), 104.7 (d), 109.1 (d), 120.8
(d), 126.8 (d), 128.1 (d), 128.2 (d), 129.0 (d), 129.1 (d), 129.6 (d),
132.7 (s), 137.6 (s), 140.5 (s) ppm. MS: m/z = 330 [M]+. C21H20N2O
(316.41): calcd. C 79.72, H 6.37, N 8.85; found C 79.61, H 6.61, N
9.02.
(3aS,8S,8aR)-1-Benzyl-8-isobutyl-3,3a,8,8a-tetrahydro-1H-
isoxazolo[4,3-a]pyrrolizine (8c): Yield 308 mg (26%); white crystals;
m.p. 82–84 °C (diisopropyl ether). [α]2D3 = +119.8 (c = 0.49, CHCl3).
1H NMR (400 MHz, CDCl3): δ = 0.81 (d, J = 6.3 Hz, 3 H), 0.91
(d, J = 6.2 Hz, 3 H), 1.26–1.45 (m, 2 H), 1.58–1.71 (m, 1 H), 3.84,
4.13 (AB system, J = 12.3 Hz, 2 H), 3.92 (dd, J = 3.0, 8.3 Hz, 1
H), 3.94 (d, J = 7.8 Hz, 1 H), 4.11–4.15 (m, 1 H), 4.18 (dd, J =
3.0, 7.6, 7.8 Hz, 1 H), 4.32 (d, J = 7.6 Hz, 1 H), 5.83 (d, J = 3.1 Hz,
1 H), 6.24 (d, J = 3.1 Hz, 1 H), 6.50 (br. s, 1 H), 7.28–7.43 (m, 5
H) ppm. 13C NMR (100 MHz, CDCl3): δ = 22.7 (q), 23.3 (q), 24.9
(d), 45.2 (t), 46.0 (d), 60.8 (t), 62.1 (d), 72.5 (t), 80.0 (d), 99.4 (d),
113.0 (d), 113.2 (d), 128.1 (d), 128.9 (d), 129.8 (d), 136.9 (s), 137.1
(s) ppm. MS: m/z = 296 [M]+. C19H24N2O (296.42): calcd. C 76.99,
H 8.16, N 9.45; found C 77.11, H 8.33, N 9.27.
(3aR*,8aS*)-1-Benzyl-3,3a,8,8a-tetrahydro-1H-isoxazolo[4,3-a]-
pyrrolizine (16): Yield 125 mg (13 %); colourless oil. 1H NMR
(400 MHz, CDCl3): δ = 3.87–3.94 (m, 3 H), 4.10 (dd, J = 7.6,
11.1 Hz, 1 H), 4.14 (dd, J = 4.0, 12.9 Hz, 1 H), 4.19 (dd, J = 3.8,
7.8 Hz, 1 H), 4.30–4.38 (m, 2 H), 5.86 (d, J = 3.3 Hz, 1 H), 6.25
(dd, J = 2.9, 3.3 Hz, 1 H), 6.52 (d, J = 2.9 Hz, 1 H), 7.29–7.45 (m,
5 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 46.9 (d), 51.7 (t),
60.7 (t), 72.3 (t), 73.9 (d), 99.8 (d), 113.4 (d), 114.2 (d), 128.1 (d),
129 (d), 129.5 (d), 136.9 (s), 137.7 (s) ppm. MS: m/z = 240 [M]+.
C15H16N2O (240.31): calcd. C 74.97, H 6.71, N 11.66; found C
75.11, H 6.54, N 11.52.
(3aS,8S,8aR)-1,8-Dibenzyl-3,3a,8,8a-tetrahydro-1H-isoxazolo[4,3-
a]pyrrolizine (8d): Yield 343 mg (26 %); white crystals; m.p. 84–
86 °C (diisopropyl ether). [α]2D3 = +108.9 (c = 0.47, CHCl3). 1H
NMR (400 MHz, CDCl3): δ = 2.88 (dd, J = 6.6, 14.0 Hz, 1 H),
3.01 (dd, J = 5.8, 14.0 Hz, 1 H), 3.67 (d, J = 12.8 Hz, 1 H), 3.86
(dd, J = 2.6, 8.2 Hz, 1 H), 3.96–3.97 (m, 2 H), 4.03 (d, J = 12.8 Hz,
1 H), 4.28 (dd, J = 3.6, 8.2 Hz, 1 H), 4.43 (dd, J = 5.8, 6.6 Hz, 1
H), 5.82 (d, J = 3.3 Hz, 1 H), 6.24 (d, J = 3.3 Hz, 1 H), 6.34 (dd,
J = 2.6, 3.3 Hz, 1 H), 6.93–6.95 (m, 2 H), 7.25–7.29 (m, 3 H), 7.35–
7.41 (m, 5 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 41.1 (t), 46.4
(d), 60.7 (t), 64.3 (d), 72.6 (t), 78.8 (d), 99.7 (d), 113.6 (d), 115.4
(d), 127.4 (d), 128.2 (d), 129.1 (d), 129.8 (d), 130.0 (d), 130.5 (d),
137.1 (s), 137.2 (s), 137.7 (s) ppm. MS: m/z = 330 [M]+. C22H22N2O
(330.43): calcd. C 79.97, H 6.71, N 8.48; found C 80.08, H 6.69, N
8.59.
(1R*,4S*)-3-Benzyl-4,5-dihydro-1H,3H-1,4-methanopyrrolo[1,2-e]-
[1,2,5]oxadiazepine (17): Yield 250 mg (26%); white crystals; m.p.
1
116–118 °C (diisopropyl ether). H NMR (400 MHz, CDCl3): δ =
2.18 (d, J = 11.2 Hz, 1 H), 2.48 (ddd, J = 5.3, 5.6, 11.2 Hz, 1 H),
3.70–3.77 (ddd, J = 2.2, 2.8, 5.3 Hz, 1 H), 3.81 (d, J = 12.7 Hz, 1
H), 3.97 (dd, J = 2.8, 11.8 Hz, 1 H), 4.05 (dd, J = 2.2, 11.8 Hz, 1
H), 4.23 (d, J = 12.7 Hz, 1 H), 5.22 (d, J = 5.3 Hz, 1 H), 5.97 (dd,
J = 1.5, 3.5 Hz, 1 H), 6.08 (dd, J = 3.2, 3.5 Hz, 1 H), 6.52 (dd, J
= 1.8, 3.2 Hz, 1 H), 7.23–7.45 (m, 5 H) ppm. 13C NMR (100 MHz,
CDCl3): δ = 33.2 (t), 52.8 (t), 58.4 (d), 63.5 (t), 71.3 (d), 104.6 (d),
108.8 (d), 120.2 (d), 128.0 (d), 128.9 (d), 129.6 (d), 132.3 (s), 137.6
(s) ppm. MS: m/z = 240 [M]+. C15H16N2O (240.31): calcd. C 74.97,
H 6.71, N 11.66; found C 74.79, H 6.93, N 11.61.
(3aS,8S,8aR)-1-Benzyl-8-phenyl-3,3a,8,8a-tetrahydro-1H-isoxazolo-
[4,3-a]pyrrolizine (8e): Yield 202 mg (16%); white crystals; m.p. 93–
95 °C (diisopropyl ether). [α]2D3 = +39.4 (c = 0.08, CHCl3). 1H
NMR (400 MHz, CDCl3): δ = 3.88 (d, J = 12.9 Hz, 1 H), 3.99 (dd,
J = 3.1, 7.9 Hz, 1 H), 4.07 (dd, J = 3.3, 7.5 Hz, 1 H), 4.15 (d, J =
12.9 Hz, 1 H), 4.27 (ddd, J = 3.1, 7.4, 7.5 Hz, 1 H), 4.34 (dd, J =
7.4, 7.9 Hz, 1 H), 5.19 (d, J = 3.3 Hz, 1 H), 5.93 (d, J = 3.3 Hz, 1
H), 6.30 (dd, J = 2.4, 3.3 Hz, 1 H), 6.39 (d, J = 2.4 Hz, 1 H), 6.87–
6.90 (m, 2 H), 7.26–7.38 (m, 5 H), 7.38–7.40 (m, 3 H) ppm. 13C
NMR (100 MHz, CDCl3): δ = 46.5 (d), 60.8 (t), 68.0 (d), 72.4 (t),
83.6 (d), 99.6 (d), 113.9 (d), 114.3 (d), 126.3 (d), 128.0 (d), 128.3
(d), 128.9 (d), 129.2 (d), 129.5 (d), 136.8 (s), 138.1 (s), 144.3
(s) ppm. MS: m/z = 316 [M]+. C21H20N2O (316.41): calcd. C 79.72,
H 6.37, N 8.85; found C 79.90, H 6.16, N 8.99.
(3aS,8S,8aR)-8-Benzyl-1-methyl-3,3a,8,8a-tetrahydro-1H-isoxazolo-
[4,3-a]pyrrolizine (19): Yield 274 mg (27%); colourless oil. [α]2D3
=
+96.7 (c = 0.40, CHCl3). 1H NMR (400 MHz, CDCl3): δ = 2.52
(s, 3 H), 3.04 (dd, J = 7.1, 13.9 Hz, 1 H), 3.16 (dd, J = 6.4, 13.9 Hz,
1 H), 3.71–3.86 (m, 2 H), 3.93–3.99 (m, 1 H), 4.24 (t, J = 7.9 Hz,
1 H), 4.51 (br. s, 1 H), 5.79 (d, J = 3.3 Hz, 1 H), 6.22 (dd, J = 2.6,
3.3 Hz, 1 H), 6.37 (d, J = 2.6 Hz, 1 H), 7.12–7.18 (m, 2 H), 7.23–
7.37 (m, 3 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 41.8 (t), 43.7
(q), 46.4 (d), 64.4 (d), 72.0 (t), 81.4 (d), 99.5 (d), 113.4 (d), 124.4
(d), 127.5 (d), 129.1 (d), 129.8 (d), 137.2 (s), 137.7 (s) ppm. MS:
m/z = 254 [M]+. C16H18N2O (254.33): calcd. C 75.56, H 7.13, N
11.01; found C 75.26, H 7.31, N 10.75.
(1S,4R,5S)-3,5-Dibenzyl-4,5-dihydro-1H,3H-1,4-methanopyrrolo-
[1,2-e][1,2,5]oxadiazepine (9d): Yield 224 mg (17%); colourless oil.
[α]2D3 = +61.7 (c = 0.68, CHCl3). 1H NMR (400 MHz, CDCl3): δ =
2.28 (m, 2 H), 2.85 (dd, J = 9.8, 14.3 Hz, 1 H), 3.23 (dd, J = 5.3,
14.3 Hz, 1 H), 3.60 (br. s, 1 H), 3.75, 4.18 (AB system, J = 12.6 Hz,
2 H), 4.48 (dd, J = 5.3, 9.8 Hz, 1 H), 5.20 (t, J = 2.6 Hz, 1 H), 6.00
General Procedure for the Isoxazolidine Ring-Opening with LiAlH4:
A 1  solution of LiAlH4 (0.90 mL) in THF was added to a solu-
tion of 8a–e or 9d,e (0.14 mmol) in dry THF (2 mL) under nitro-
gen. The resultant solution was heated at reflux for 24 h and then,
(dd, J = 1.5, 3.5 Hz, 1 H), 6.12 (d, J = 3.5 Hz, 1 H), 6.59 (d, J = after cooling, MeOH (18 mL) was added dropwise. After concen-
1.5 Hz, 1 H), 6.98–7.01 (m, 2 H), 7.19–7.28 (m, 8 H) ppm. 13C tration, a solution of AcONa (690 mg, 8.37 mmol) in H2O (3.5 mL)
NMR (100 MHz, CDCl3): δ = 30.1 (t), 40.8 (t), 61.2 (d), 63.0 (d),
63.3 (t), 71.4 (d), 104.7 (d), 108.9 (d), 119.4 (d), 127.1 (d), 127.9
(d),128.8 (d), 129.0 (d), 129.2 (d), 129.5 (d), 131.8 (s), 135.6 (s),
137.0 (s) ppm. MS: m/z = 330 [M]+. C22H22N2O (330.43): calcd. C
79.97, H 6.71, N 8.48; found C 79.98, H 6.56, N 8.24.
was added and the mixture was extracted with CHCl3 (4ϫ4 mL).
The organic layer was dried with Na2SO4 and the solvent was re-
moved under reduced pressure. The residue was purified by
chromatography on a silica gel column with light petroleum/AcOEt
(2:1) as eluent to give 10a–e or 11d,e, respectively.
2814
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2808–2816

Pyrrolizines and Indolizines from α-Amino Acids
(1S,2R,3S)-(2-Benzylamino-3-methyl-2,3-dihydro-1H-pyrrolizin-1-
3.53 (dt, J = 6.0, 5.8 Hz, 1 H), 3.85–3.89 (m, 3 H), 3.96 (d, J =
6.5 Hz, 2 H), 5.06 (d, J = 6.5 Hz, 1 H), 5.97 (d, J = 2.6 Hz, 1 H),
6.27 (d, J = 2.6 Hz, 1 H), 6.61 (s, 1 H), 7.16 (dd, J = 5.6, 7.1 Hz,
4 H), 7.26 (d, J = 7.1 Hz, 2 H), 7.38 (d, J = 5.6 Hz, 4 H) ppm. 13C
NMR (100 MHz, CDCl3): δ = 42.1 (d), 53.1 (t), 63.3 (t), 67.8 (d),
72.7 (d), 100.7 (d), 112.5 (d), 114.5 (d), 127.4 (d), 127.7 (d), 128.4
(d), 128.7 (d), 128.9 (d), 129.3 (d), 135.9 (s), 139.5 (s), 139.9
(s) ppm. MS: m/z = 318 [M]+. C21H22N2O (318.42): calcd. C 79.21,
H 6.96, N 8.80; found C 79.33, H 6.74, N 8.59.
yl)methanol (10a): Yield 15 mg (42%); colourless oil. IR (Nujol): ν
˜
= 3320, 2910 cm–1. [α]2D3 = –5.3 (c = 0.33, CHCl3). 1H NMR
(400 MHz, CDCl3): δ = 1.50 (d, J = 6.3 Hz, 3 H), 2.12 (br. s, 2 H,
missing after deuteriation), 3.44 (dt, J = 7.2, 5.8 Hz, 1 H), 3.58 (dd,
J = 6.8, 7.2 Hz, 1 H), 3.87 (d, J = 5.8 Hz, 2 H), 3.92, 4.01 (AB
system, J = 13.0 Hz, 2 H), 4.05 (qd, J = 6.3, 6.8 Hz, 1 H), 5.88 (d,
J = 3.4 Hz, 1 H), 6.23 (dd, J = 2.5, 3.4 Hz, 1 H), 6.60 (d, J = 2.5 Hz,
1 H), 7.30–7.39 (m, 5 H) ppm. 13C NMR (100 MHz, CDCl3): δ
= 18.7 (q), 43.5 (d), 53.4 (t), 59.0 (d), 63.4 (t), 70.9 (d), 100.6 (d),
111.9 (d), 113.0 (d), 127.9 (d), 128.6 (d), 129.0 (d), 135.2 (s), 139.9
(s) ppm. MS: m/z = 256 [M]+. C16H20N2O (256.35): calcd. C 74.97,
H 7.86, N 10.93; found C 75.22, H 7.61, N 11.10.
(5S,6R,8S)-5-Benzyl-6-benzylamino-5,6,7,8-tetrahydroindolizin-8-ol
(11d): Yield 15 mg (33%); colourless oil. IR (Nujol): ν = 3324,
˜
2906 cm–1. [α]2D3 = +5.1 (c = 0.36, CHCl3). 1H NMR (400 MHz,
CDCl3): δ = 2.12 (ddd, J = 2.9, 6.8, 14.9 Hz, 1 H), 2.30 (ddd, J =
3.2, 6.4, 14.9 Hz, 1 H), 2.34 (br. s, 2 H, missing after deuteriation),
2.86 (dd, J = 8.5, 13.9 Hz, 1 H), 3.03 (dd, J = 5.6, 13.9 Hz, 1 H),
3.67–3.72 (m, 1 H), 3.70, 3.88 (AB system, J = 13.0 Hz, 2 H), 4.31
(dd, J = 5.6, 8.5 Hz, 1 H), 4.86 (dd, J = 2.9, 3.2 Hz, 1 H), 6.20 (dd,
J = 2.9, 3.5 Hz, 1 H), 6.23 (d, J = 3.5 Hz, 1 H), 6.48 (d, J = 2.9 Hz,
1 H), 7.03–7.06 (m, 2 H), 7.13–7.15 (m, 2 H), 7.23–7.38 (m, 6
H) ppm. 13C NMR (100 MHz, CDCl3): δ = 43.4 (t), 51.9 (t), 53.9
(d), 61.8 (d), 62.7 (d), 63.1 (t), 107.2 (d), 109.1 (d), 119.9 (d), 127.3
(d), 127.7 (d), 128.5 (d), 129.0 (d), 129.2 (d), 129.5 (d), 131.2 (s),
137.7 (s), 139.4 (s) ppm. MS: m/z = 332 [M]+. C22H24N2O (332.45):
calcd. C 79.48, H 7.28, N 8.43; found C 79.67, H 7.09, N 8.19.
(1S,2R,3S)-(2-Benzylamino-3-isopropyl-2,3-dihydro-1H-pyrrolizin-
1-yl)methanol (10b): Yield 25 mg (63%); colourless oil. IR (Nujol):
ν = 3310, 2902 cm–1. [α]23 = +21.6 (c = 0.10, CHCl3). 1H NMR
˜
D
(400 MHz, CDCl3): δ = 0.93 (d, J = 6.9 Hz, 3 H), 0.98 (d, J =
6.9 Hz, 3 H), 2.08 (dt, J = 4.8, 6.9 Hz, 1 H), 2.30 (br. s, 2 H, missing
after deuteriation), 3.51 (d, J = 5.1 Hz, 1 H), 3.79 (d, J = 8.4 Hz,
1 H), 3.83, 3.95 (AB system, J = 12.9 Hz, 2 H), 3.88 (d, J = 8.4 Hz,
1 H), 3.97 (d, J = 5.1 Hz, 1 H), 4.01 (d, J = 4.8 Hz, 1 H), 5.80 (d,
J = 3.4 Hz, 1 H), 6.23 (dd, J = 2.7, 3.4 Hz, 1 H), 6.61 (d, J = 2.7 Hz,
1 H), 7.28–7.40 (m, 5 H) ppm. 13C NMR (100 MHz, CDCl3): δ
= 18.4 (q), 19.5 (q), 32.2 (d), 42.3 (d), 51.8 (t), 61.9 (t), 65.5 (d),
68.8 (d), 98.6 (d), 112.4 (s), 112.5 (d), 115.0 (d), 125.4 (d), 128.7
(d), 129.4 (d), 135.3 (s) ppm. MS: m/z = 284 [M]+. C18H24N2O
(284.40): calcd. C 76.02, H 8.51, N 9.85; found C 75.97, H 8.80, N
9.68.
(5S,6R,8S)-6-Benzylamino-5-phenyl-5,6,7,8-tetrahydroindolizin-8-ol
(11e): Yield 13 mg (30%); colourless oil. IR (Nujol): ν = 3304,
˜
2915 cm–1. [α]2D3 = –4.8 (c = 0.05, CHCl3). 1H NMR (400 MHz,
CDCl3): δ = 2.01–2.07 (m, 2 H), 2.25 (br. s, 2 H, missing after
deuteriation), 3.34 (dd, J = 2.9, 6.8 Hz, 1 H), 3.90, 4.07 (AB system,
J = 13.1 Hz, 2 H), 4.97 (t, J = 4.0 Hz, 1 H), 5.27 (d, J = 2.9 Hz, 1
H), 6.26 (dd, J = 3.2, 3.5 Hz, 1 H), 6.33 (dd, J = 1.9, 3.5 Hz, 1 H),
6.65 (dd, J = 1.9, 3.2 Hz, 1 H), 6.64–6.66 (m, 2 H), 7.25–7.40 (m,
8 H) ppm. 13C NMR (100 MHz, CDCl3): δ = 30.1 (t), 58.9 (d),
63.1 (d), 64.3 (d), 68.5 (t), 107.1 (d), 109.7 (d), 121.0 (d), 126.5 (d),
127.8 (d), 128.1 (d), 128.6 (d), 129.1 (d), 129.2 (d), 135.6 (s), 139.4
(s), 142.1 (s) ppm. MS: m/z = 318 [M]+. C21H22N2O (318.42): calcd.
C 79.21, H 6.96, N 8.80; found C 79.02, H 7.24, N 8.97.
(1S,2R,3S)-(2-Benzylamino-3-isobutyl-2,3-dihydro-1H-pyrrolizin-1-
yl)methanol (10c): Yield 24 mg (59%); pale-yellow oil. IR (Nujol):
ν = 3328, 2915 cm–1. [α]23 = +20.1 (c = 0.14, CHCl3). 1H NMR
˜
D
(400 MHz, CDCl3): δ = 1.02 (d, J = 6.6 Hz, 3 H), 1.03 (d, J =
6.6 Hz, 3 H), 1.53 (ddd, J = 6.5, 7.2, 13.8 Hz, 1 H), 1.68 (ddd, J =
5.9, 7.5, 13.8 Hz, 1 H), 1.92 (ddt, J = 6.5, 7.5, 6.6 Hz, 1 H), 2.16
(br. s, 2 H, missing after deuteriation), 3.50 (d, J = 7.1 Hz, 1 H),
3.65 (dd, J = 5.5, 7.1 Hz, 1 H), 3.86–3.96 (m, 4 H), 4.11 (ddd, J =
5.5, 5.9, 7.2 Hz, 1 H), 5.85 (d, J = 3.2 Hz, 1 H), 6.22 (dd, J = 2.5,
3.2 Hz, 1 H), 6.63 (d, J = 2.5 Hz, 1 H), 7.29–7.37 (m, 5 H) ppm.
13C NMR (100 MHz, CDCl3): δ = 22.8 (q), 23.6 (q), 30.1 (t), 30.7
(d), 41.7 (d), 53.1 (t), 61.7 (d), 62.8 (t), 69.7 (d), 100.2 (d), 112.1
(d), 114.1 (d), 127.9 (d), 128.6 (d), 129.1 (d), 135.4 (s), 139.8
(s) ppm. MS: m/z = 298 [M]+. C19H26N2O (298.43): calcd. C 76.47,
H 8.78, N 9.39; found C 76.23, H 8.89, N 9.51.
Supporting Information (see also the footnote on the first page of
this article): Cartesian coordinates and energies of the stationary
points.
Acknowledgments
(1S,2R,3S)-(3-Benzyl-2-benzylamino-2,3-dihydro-1H-pyrrolizin-1-
The authors thank the Ministero dell’Istruzione, dell’Università e
della Ricerca (MIUR) (First and PRIN) for financial support and
the Centro Interuniversitario Lombardo per l’Elaborazione Auto-
matica (CILEA) for computational facilities.
yl)methanol (10d): Yield 24 mg (52%); colourless oil. IR (Nujol): ν
˜
= 3305, 2912 cm–1. [α]2D3 = +20.2 (c = 0.30, CHCl3). 1H NMR
(400 MHz, CDCl3): δ = 2.21 (br. s, 2 H, missing after deuteriation),
3.02 (dd, J = 7.2, 13.8 Hz, 1 H), 3.17 (dd, J = 6.4, 13.8 Hz, 1 H),
3.40 (dd, J = 7.0, 4.9 Hz, 1 H), 3.67–3.75 (m, 3 H), 3.84 (dd, J =
7.0, 12.2 Hz, 1 H), 3.89 (dd, J = 4.9, 12.2 Hz, 1 H), 4.33 (dd, J =
6.4, 7.2 Hz, 1 H), 5.85 (d, J = 3.3 Hz, 1 H), 6.20 (dd, J = 2.6,
3.3 Hz, 1 H), 6.46 (d, J = 2.6 Hz, 1 H), 7.11–7.13 (m, 2 H), 7.19–
7.21 (m, 2 H), 7.26–7.39 (m, 6 H) ppm. 13C NMR (100 MHz,
CDCl3): δ = 41.1 (t), 41.6 (d), 53.0 (t), 63.3 (t), 64.3 (d), 68.7 (d),
100.3 (d), 112.2 (d), 113.8 (d), 127.4 (d), 127.8 (d), 128.7 (d), 128.9
(d), 129.3 (d), 129.6 (d), 135.6 (s), 137.5 (s), 139.5 (s) ppm. MS: m/z
= 332 [M]+. C22H24N2O (332.44): calcd. C 79.48, H 7.28, N 8.43;
found C 79.22, H 7.41, N 8.55.
[1] a) A. R. Mattocks, Chemistry and Toxicology of Pyrrolizidine
Alkaloids; Academic Press, New York, 1986; b) A. D. Elbein,
R. J. Molyneux in Alkaloids: Chemical and Biological Perspec-
tives (Ed.: S. W. Pelletier), Wiley, New York, 1987, vol. 5, chap-
ter 1; c) A. D. Elbein, J. E. Tropea, R. J. Molineux, U.S. Pat.
Appl. 289907, 1990 [Chem. Abstr. 1990, 113, P91444p; d) T.
Granier, F. Gaiser, L. Hintermann, A. Vasella, Helv. Chim.
Acta 1997, 80, 1443–1456; e) L. H. Zalkow, J. A. Glinski, L. T.
Gelbaum, T. J. Fleischmann, L. S. McGowan, M. M. Gordon,
J. Med. Chem. 1985, 28, 687–694; f) R. A. Gruters, J. J. Neefjes,
M. Tersmette, R. A. Y. de Goede, A. Tulp, H. G. Huisman, F.
Miedema, H. L. Ploegh, Nature 1987, 330, 74–77; g) A.
Karpas, G. W. J. Fleet, R. A. Dwek, S. Petursson, S. K.
Namgoong, N. G. Ramsden, G. S. Jacob, T. W. Rademacher,
Proc. Natl. Acad. Sci. USA 1988, 85, 9229–9233; h) J. W. Den-
(1S,2R,3S)-(2-Benzylamino-3-phenyl-2,3-dihydro-1H-pyrrolizin-1-
yl)methanol (10e): Yield 16 mg (35%); colourless oil. IR (Nujol): ν
˜
= 3308, 2900 cm–1. [α]2D3 = –5.9 (c = 0.14, CHCl3). 1H NMR
(400 MHz, CDCl3): δ = 2.44 (br. s, 2 H, missing after deuteriation),
Eur. J. Org. Chem. 2008, 2808–2816
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2815

E. Borsini, G. Broggini, A. Contini, G. Zecchi
[11] M. Nishio, Tetrahedron 2005, 61, 6923–6950.
[12] C. N. Alves, O. A. S. Romero, A. B. F. Da Silva, Int. J. Quan-
tum Chem. 2003, 95, 133–136.
[13] M. J. Allen, D. J. Tozer, J. Chem. Phys. 2002, 117, 11113–11120.
[14] a) Y. Zhao, N. E. Schultz, D. G. Truhlar, J. Chem. Theory Com-
put. 2006, 2, 364–382; b) Y. Zhao, D. G. Truhlar, J. Phys. Chem.
A 2005, 109, 4209–4212; c) S. Grimme, J. Comput. Chem. 2004,
25, 1463–1473; d) Y. Zhao, D. G. Truhlar, J. Phys. Chem. A
2004, 108, 6908–6918; e) S. Grimme, J. Chem. Phys. 2003, 118,
9095–9102; f) M. Kamiya, T. Tsuneda, K. Hirao, J. Chem.
Phys. 2002, 117, 6010–6015.
FULL PAPER
nis, Cancer Res. 1986, 46, 5131–5136; i) M. J. Humphries, K.
Matsumoto, S. L. White, K. Olden, Cancer Res. 1986, 46,
5215–5222; j) P. C. Das, J. D. Roberts, S. L. White, K. Olden,
Oncology Res. 1995, 7, 425–433; k) J. Gubin, J. Lucchetti, J.
Mahaux, D. Nisato, G. Rosseels, M. Clinet, P. Polster, P. Chate-
lain, J. Med. Chem. 1992, 35, 981–988; l) J. R. Liddell, Nat.
Prod. Rep. 2002, 19, 773–781; m) J. P. Michael, Nat. Prod. Rep.
2005, 22, 603–626; n) L. F. James, K. E. Panter, W. Gaffield,
R. J. Molyneux, J. Agric. Food Chem. 2004, 52, 3211–3230.
[2] a) P. N. Confalone, E. M. Huie, Org. React. 1988, 36, 1–172;
b) A. Padwa, W. H. Pearson, Synthetic Applications of 1,3–Di-
polar Cycloaddition Chemistry Toward Heterocycles and Natural
Products, Wiley, New York, 2003, chapter 1, pp. 1–81; c) U.
Chiacchio, A. Rescifina, G. Romeo in Targets in Heterocyclic
Systems (Eds.: O. A. Attanasi, D. Spinelli), SCI, Rome, 1997,
vol. I, pp. 225–276; d) M. Frederickson, Tetrahedron 1997, 53,
403–425.
[15] E. R. Johnson, R. A. Wolkow, G. A. DiLabio, Chem. Phys.
Lett. 2004, 394, 334–338.
[16] M. O. Sinnokrot, E. F. Valeev, C. D. Sherrill, J. Am. Chem. Soc.
2002, 124, 10887–10893.
[17] A. D. Boese, N. C. Handy, J. Chem. Phys. 2001, 114, 5497–
5503.
[3] a) A. Arnone, G. Broggini, D. Passarella, A. Terraneo, G. Zec- [18] A. Contini, S. Leone, S. Menichetti, C. Viglianisi, P. Trimarco,
chi, J. Org. Chem. 1998, 63, 9279–9284; b) G. Broggini, C.
La Rosa, T. Pilati, A. Terraneo, G. Zecchi, Tetrahedron 2001,
57, 8323–8332; c) E. M. Beccalli, G. Broggini, A. Farina, L.
Malpezzi, A. Terraneo, G. Zecchi, Eur. J. Org. Chem. 2002,
2080–2086.
J. Org. Chem. 2006, 71, 5507–5514.
[19] W. L. Jorgensen, D. Lim, J. F. Blake, J. Am. Chem. Soc. 1993,
115, 2936–2942.
[20] a) J. I. Garcia, V. Martinez-Merino, J. A. Mayoral, L. Salvat-
ella, J. Am. Chem. Soc. 1998, 120, 2415–2420; b) H. Wang, Y.
Wang, K.-L. Han, X.-J. Peng, J. Org. Chem. 2005, 70, 4910–
4917.
[4] C. Kashima, T. Maruyama, K. Harada, S. Hibi, Y. Omote, J.
Chem. Res. 1988, 62–63 (S), 601–645 (M).
[5] A. S. Demir, N. T. Subasi, E. Sahin, Tetrahedron: Asymmetry
2006, 17, 2625–2631.
[6] a) L. Lombardo, Tetrahedron Lett. 1982, 23, 4293–4296; b) K.
Takai, T. Kakiuchi, Y. Kataoka, K. Utimoto, J. Org. Chem.
1994, 59, 2668–2670.
[7] E. Beccalli, G. Broggini, A. Contini, I. De Marchi, G. Zecchi,
C. Zoni, Tetrahedron: Asymmetry 2004, 15, 3181–3187.
[8] a) K. B. G. Torssell, Nitrile Oxides, Nitrones, and Nitronates in
Organic Synthesis, VCH, New York, 1988; b) G. Broggini, I.
De Marchi, M. Martinelli, G. Zecchi in Targets in Heterocyclic
Systems (Eds: O. A. Attanasi, D. Spinelli), SCI, Rome, 2005,
vol. 9, pp. 211–229.
[9] a) A. Milet, Y. Gimbert, A. E. Greene, J. Comput. Chem. 2006,
27, 157–162; b) M. L. Kuznetsov, V. Y. Kukushkin, J. Org.
Chem. 2006, 71, 582–592; c) A. Ponti, G. Molteni, Chem. Eur.
J. 2006, 12, 1156–1161; d) L. Manzoni, D. Arosio, L. Belvisi,
A. Bracci, M. Colombo, D. Invernizzi, C. Scolastico, J. Org.
Chem. 2005, 70, 4124–4132; e) G. Wagner, Chem. Eur. J. 2003,
9, 1503–1510; f) P. Merino, J. Revuelta, T. Tejero, U. Chiacchio,
A. Rescifina, G. Romeo, Tetrahedron 2003, 59, 3581–3592; g)
P. Merino, J. A. Mates, J. Revuelta, T. Tejero, U. Chiacchio, G.
Romeo, D. Iannazzo, R. Romeo, Tetrahedron: Asymmetry
2002, 13, 173–190; h) M. A. Silva, J. M. Goodman, Tetrahedron
2002, 58, 3667–3671; i) K. Kavitha, P. Venuvanalingam, J.
Chem. Soc. Perkin Trans. 2 2002, 12, 2130–2139; j) R. Herrera,
A. Nagarajan, M. A. Morales, F. Méndez, H. A. Jiménez-
Vázquez, L. G. Zepeda, J. Tamariz, J. Org. Chem. 2001, 66,
1252–1263; k) M. Carda, R. Portolés, J. Murga, S. Uriel, J. A.
Marco, L. R. Domingo, R. Zaragozá, H. Röper, J. Org. Chem.
2000, 65, 7000–7009.
[21] The nomenclature used herein for describing aromatic ring in-
teractions follows that reported in ref.[15] for the two stable ben-
zene dimers, namely the T-shaped and the parallel-displaced
(PD).
[22] F. Eckert, A. Klamt, AIChE J. 2002, 48, 369–385.
[23] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. A. Montgomery Jr, T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tom-
asi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota,
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda,
O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratch-
ian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli,
J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Sal-
vador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Ba-
boul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Lia-
shenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T.
Keith, M. A. Al–Laham, C. Y. Peng, A. Nanayakkara, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W.
Wong, C. Gonzalez, J. A. Pople, Gaussian 03, Revision B.04,
Gaussian, Inc., Pittsburgh, PA, 2003
Received: December 21, 2007
Published Online: April 9, 2008
[10] a) T. K. M. Shing, A. W. F. Wong, T. Ikeno, T. Yamada, J. Org.
Chem. 2006, 71, 3253–3263; b) S. Torrente, B. Noya, V. Branch-
adell, R. Alonso, J. Org. Chem. 2003, 68, 4772–4783.
2816
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 2808–2816