Structure-directed reversion in the π-facial stereoselective alkylation of chiral bicyclic lactams

Article abstract of DOI:10.1021/jo801665k

(Chemical Equation Presented) The reversal of the π-facial diastereoselectivity observed in the benzyl bromide alkylation of the enolate of phenylglycinol-derived oxazolopiperidone is examined by means of theoretical calculations and experimental assays.

Full text of DOI:10.1021/jo801665k

Structure-Directed Reversion in the π-Facial Stereoselective
Alkylation of Chiral Bicyclic Lactams
Ignacio Soteras,† Oscar Lozano,‡ Carmen Escolano,‡ Modesto Orozco,§ Mercedes Amat,‡
Joan Bosch,*,‡ and F. Javier Luque*,†
Department of Physical Chemistry and Laboratory of Organic Chemistry, Faculty of Pharmacy, and
Institute of Biomedicine (IBUB), UniVersity of Barcelona, 08028 Barcelona, Spain, Molecular Modeling
and Bioinformatics Unit, Institute of Biomedical Research, Barcelona Scientific Park, 08028 Barcelona,
Spain, Department of Life Sciences, Barcelona Supercomputing Centre, 08034 Barcelona, Spain, and
Department of Biochemistry, Faculty of Biology, UniVersity of Barcelona, Barcelona 08028, Spain
fjluque@ub.edu; joanbosch@ub.edu
ReceiVed July 29, 2008
The reversal of the π-facial diastereoselectivity observed in the benzyl bromide alkylation of the enolate
of phenylglycinol-derived oxazolopiperidone is examined by means of theoretical calculations and
experimental assays. When the angular carbon adopts an R configuration, the endo addition is intrinsically
favored due to the lower torsional strain induced in the TS, but for the reaction with benzyl bromide,
which affords the exo product. This reversal is attributed to the formation of a C-H · · · π hydrogen bond
between the C-H unit of the C8a angular position and the benzene ring of the alkylating reagent. A
similar secondary interaction explains the stereochemical reversion observed in the benzyl alkylation of
the enolate of oxazolopyrrolidinone. These findings point out that the delicate balance between factors
that dictate the diastereoselectivity in the alkylation of chiral byciclic lactams can be unexpectedly altered
by weak secondary interactions. The results presented here provide valuable guidelines to tune the selective
preparation of enantiopure bioorganic and pharmaceutical compounds.
SCHEME 1
Introduction
The alkylation of enolates generated from chiral bicyclic
lactams is a useful method of carbon-carbon bond formation
involving the generation of new stereogenic centers, which is
highly relevant in the asymmetric synthesis of complex mol-
ecules.1 Meyers and co-workers, in their pioneering and
extensive work in this field, demonstrated that the double
alkylation of chiral bicyclic [3.3.0] lactams A (Scheme 1) is a
powerful tool to generate quaternary stereocenters with excellent
control over the absolute stereochemistry.2 However, the endo:
exo diastereofacial selectivity is largely dependent on both the
nature of the fused ring and the presence of substituents.3 A
similar sensitivity in the stereochemical alkylation outcome has
also been reported for bicyclic [4.3.0] lactams B.4 The finding
† Department of Physical Chemistry, Faculty of Pharmacy, and IBUB,
University of Barcelona.
‡ Laboratory of Organic Chemistry, Faculty of Pharmacy, and IBUB,
University of Barcelona.
§ Institute of Biomedical Research, Barcelona Supercomputing Centre, and
Department of Biochemistry, University of Barcelona.
(2) For reviews, see: (a) Romo, D.; Meyers, A. I. Tetrahedron 1991, 47,
9503. (b) Meyers, A. I. In Stereocontrolled Organic Synthesis; Blackwell
Scientific Publications: Oxford, 1994; pp 145-175. (c) Meyers, A. I.; Brengel,
G. P. Chem. Commun. 1997, 1. (d) Groaning, M. D.; Meyers, A. I. Tetrahedron
2000, 56, 9843.
(1) Ho¨gberg, H.-E. In StereoselectiVe Synthesis, Methods of Organic
Chemistry (Houben-Weyl); Helmchen, G., Hoffman, R. W., Mulzer, J., Schau-
mann, E., Eds.; Thieme: Stuttgart, 1996; E21, Vol. 2, pp 791-915.
7756 J. Org. Chem. 2008, 73, 7756–7763
10.1021/jo801665k CCC: $40.75 2008 American Chemical Society
Published on Web 09/03/2008

ReVersion in π-Facial StereoselectiVe Alkylation
TABLE 1. Alkylation of Oxazolopiperidone Lactams 1 and 2
that highly stereoselective alkylations occur in monocyclic
imidazolidinone C, lactam D (X ) NR2), and lactone D (X )
O) enolates is even more intriguing,5 because it suggests that
factors other than the steric hindrance in the concave-convex
faces of the enolate can play an important role.
The nature of the factors that explain the π-facial stereose-
lectivity in the above alkylations is still a challenging question.
For the enolate of imidazolidinone C, the involvement of
stereoelectronic effects based on the pyramidalization on the
C5 carbon has been proposed.5b The diastereofacial selectivity
in D (X ) NR2) has, nevertheless, been ascribed to the
competition between the anti-stereoelectronic directing effect
of the nitrogen lone pair and steric factors6 or alternatively to
the chelation of the enolate by the metal (lithium) ion.7 In
contrast, Houk and co-workers have rationalized the stereo-
chemical selectivity of D (X ) NR2) from the differences in
the torsional strain in the transition states leading to endo and
exo products.8 This factor has also been used to explain the
stereoselective alkylation of 4-substituted γ-butyrolactones D
(X ) O).9 Taken together, these results highlight the extreme
sensitivity of the stereochemical outcome to apparently minor
chemical modifications, which affect the interplay between
different factors.
We have recently examined the origin of the π-facial
stereoselectivity in the alkylation of the enolates of oxazolopi-
peridones 1 and 2,10 which stems from a delicate balance
between torsional strain and steric hindrance. Thus, the dias-
tereoselectivity in 1 is dictated by the greater intermolecular
steric hindrance between the incoming alkylating reagent and
the piperidone ring of the enolate in the endo transition state.
However, when the stereocenter at the angular position adopts
an R configuration (2), the phenyl ring at position 3 forces the
pseudo-planarity of the bicyclic lactam, and the diastereoselec-
tivity is dictated by the internal torsional strain induced in the
transition state, leading preferentially to the endo product.
R2
1a endo:exo 1b endo:exo 2a endo:exo 2b endo:exo
Me
Et
15:85
0:100
32:68
10:90
23:77
5:95
76:24
68:32
88:12
48:52
8:92
71:29
CH2CO2tBu
CH2CHdCH2
CH2C6H5
23:77
17:83
31:69
83:17
48:52
10:90
Whereas the preceding rationale justifies the exo diastereose-
lectivity observed for the alkylation of 1 using a variety of
reagents (alkyl halides, allyl bromide, or benzyl bromide), this
is not true for 2, since the endo preference observed in the
enolate alkylation with alkyl halides is completely reversed with
benzyl bromide (see Table 1).11
The reversal in the facial selectivity, or at least notable
changes in the diastereomeric ratio, observed in lactam enolate
alkylations when using benzyl bromide instead of alkyl halides
has been noted in previous studies, including [3.3.0] lactams,
and attributed to steric or mechanistic (SN1 versus SN2)
factors.12,13 However, no theoretical studies devoted to rational-
izing the anomalous behavior of benzyl bromide have been
reported so far. The aim of this work is to identify the origin of
the π-facial stereoselective reversion observed in the benzyl
alkylation of bicyclic lactams. To this end, the results of
theoretical and experimental studies performed for the benzyl
alkylation of the oxazolopiperidone 2 and its 8a-substituted
analog 3 and the oxazolopyrrolidinone 4 will be discussed.
(3) (a) Roth, G. P.; Leonard, S. F.; Tong, L. J. Org. Chem. 1996, 61, 5710.
(b) Meyers, A. I.; Seefeld, M. A.; Lefker, B. A. J. Org. Chem. 1996, 61, 5712.
(c) Meyers, A. I.; Seefeld, M. A.; Lefker, B. A.; Blake, J. F.; Williard, P. G.
J. Am. Chem. Soc. 1998, 120, 7429. (d) Bailey, J. H.; Byfield, A. T. J.; Davis,
P. J.; Foster, A. C.; Leech, M.; Moloney, M. G.; Mu¨ller, M.; Prout, C. K. J. Chem.
Soc., Perkin Trans. 1 2000, 1977, and references therein. (e) Oda, K.; Meyers,
A. I. Tetrahedron Lett. 2000, 41, 8193. (f) Anwar, M.; Bailey, J. H.; Dickinson,
L. C.; Edwards, H. J.; Goswami, R.; Moloney, M. G. Org. Biomol. Chem. 2003,
1, 2364.
(4) (a) Reeder, M. R.; Meyers, A. I. Tetrahedron Lett. 1999, 40, 3115. (b)
Hughes, R. C.; Dvorak, C. A.; Meyers, A. I. J. Org. Chem. 2001, 66, 5545. (c)
Amat, M.; Escolano, C.; Lozano, O.; Bosch, J. Org. Lett. 2003, 5, 3139. (d)
Amat, M.; Lozano, O.; Escolano, C.; Molins, E.; Bosch, J. J. Org. Chem. 2007,
72, 4431.
(5) (a) Tomioka, K.; Cho, Y.-S.; Sato, F.; Koga, K. J. Org. Chem. 1988, 53,
4094. (b) Baldwin, J. E.; Miranda, T.; Moloney, M. Tetrahedron 1989, 45, 7459.
(c) Seebach, D.; Maetzke, T.; Petter, W.; Klo¨tzer, B.; Plattner, D. A. J. Am.
Chem. Soc. 1991, 113, 1781. (d) Ezquerra, J.; Pedregal, C.; Rubio, A.;
Yruretagoyena, B.; Escribano, A.; Sa´nchez-Ferrando, F. Tetrahedron 1993, 49,
8665. (e) Bren˜a-Valle, L. J.; Carreo´n, R.; Cruz-Almanza, R. Tetrahedron:
Asymmetry 1996, 7, 1019. (f) Charrier, J.-D.; Duffy, J. E. S.; Hitchcock, P. B.;
Young, D. W. Tetrahedron Lett. 1998, 39, 2199. (g) Maldaner, A. O.; Pilli,
R. A. Tetrahedron 1999, 55, 13321.
Results and Discussion
Stereoselective Alkylation of the Enolate of Oxazolo-
piperidone 2a. To investigate the origin of the π-facial
selectivity reversion observed in the benzylation of the enolate
of 2a (see Table 1), the transition states (TS) leading to endo
and exo additions were located. Two TSs were identified for
the exo attack, but only one was identified for the endo addition
(Figure 1). The distance from the benzyl carbon atom in benzyl
chloride (denoted Cb hereafter) to the nucleophilic carbon atom
(CR) of the enolate was similar in all the TSs (∼2.68 Å).
(6) (a) Meyers, A. I.; Seefeld, M. A.; Lefker, B. A.; Blake, J. F. J. Am. Chem.
Soc. 1997, 119, 4565. (b) Meyers, A. I.; Seefeld, M. A.; Lefker, B. A.; Blake,
J. F.; Williard, P. G. J. Am. Chem. Soc. 1998, 120, 7429.
(11) Amat, M.; Escolano, C.; Lozano, O.; Go´mez-Esque´, A.; Griera, R.;
Molins, E.; Bosch, J. J. Org. Chem. 2006, 71, 3804.
(7) (a) Durkin, K. A.; Liotta, D. J. Am. Chem. Soc. 1990, 112, 8162. (b)
Ikuta, Y.; Tomoda, S. Tetrahedron Lett. 2003, 44, 5931. (c) Ikuta, Y.; Tomoda,
S. Org. Lett. 2004, 6, 189.
(8) Ando, K.; Green, N. S.; Li, Y.; Houk, K. N. J. Am. Chem. Soc. 1999,
121, 5334.
(12) (a) Schwarz, J. B.; Meyers, A. I. J. Org. Chem. 1998, 63, 1619. (b)
Mills, C. E.; Heightman, T. D.; Hermitage, S. A.; Moloney, M. G.; Woods,
G. A. Tetrahedron Lett. 1998, 39, 1025. (c) Zhang, R.; Brownewell, F.;
Madalengoitia, J. S. Tetrahedron Lett. 1999, 40, 2707. (d) Brewster, A. G.;
Broady, S.; Davies, C. E.; Heightman, T. D.; Hermitage, S. A.; Hughes, M.;
Moloney, M. G.; Woods, G. Org. Biomol. Chem. 2004, 2, 1031. See also refs
3d and 5e.
(13) (a) Armstrong, R. W.; DeMattei, J. A. Tetrahedron Lett. 1991, 32, 5749.
(b) Beard, M. J.; Bailey, J. H.; Cherry, D. T.; Moloney, M. G.; Shim, S. B.;
Statham, K. A. Tetrahedron 1996, 52, 3719.
(9) Ando, K. J. Am. Chem. Soc. 2005, 127, 3964.
(10) (a) Soteras, I.; Lozano, O.; Go´mez-Esque´, A.; Escolano, C.; Orozco,
M.; Amat, M.; Bosch, J.; Luque, F. J. J. Am. Chem. Soc. 2006, 128, 6581. (b)
For a recent review on chiral oxazolopiperidone lactams, see: Escolano, C.; Amat,
M.; Bosch, J. Chem. Eur. J. 2007, 72, 4431.
J. Org. Chem. Vol. 73, No. 19, 2008 7757

Soteras et al.
FIGURE 1. Transition state structures for the endo (top) and exo (bottom) addition of benzyl chloride to the enolate of oxazolopiperidone 2a,
together with the Newman projections viewed from the directions along the CR-Cꢀ and Cꢀ-Cγ bonds (the forming bond between CR and Cb is
shown by a dotted line in Newman projections). Distances and torsional angles are in angstroms and degrees, respectively.
TABLE 2. Gibbs Free Energy Differences (at 298 K; kcal/mol)
between endo and exo Transition States for Addition of Benzyl
Chloride to the Enolate of Oxazolopiperidone 2a
SCHEME 2
solvent
endo-a
exo-a
exo-b
gasa
2.5 (3.1)
3.3 (3.8)
2.5 (3.0)
2.6
2.7
2.2
4.0 (4.8)
4.7 (5.4)
2.9 (4.7)
4.0
3.8
3.7
0.0 (0.0)
0.0 (0.0)
0.0 (0.0)
0.0
0.0
0.0
carbon tetrachloride
chloroform
octanol
ring (distance of 2.92 Å; Figure 1). Such a C-H · · · π
hydrogen bond, which is mainly stabilized by dispersion
forces,14 is topologically reflected in the formation of a cage
critical point in the electron density (Fcage ) 0.0030 au), as
already noted in the complexation of benzene with CH4.15
To corroborate the role of the C-H · · · π hydrogen bond,
the reaction with benzyl bromide was performed with the
enolate of the substituted oxazolopiperidone 3, which bears
an additional methyl group at the angular position. According
to our hypothesis, such a chemical modification would
prevent the formation of the C-H · · · π interaction and
sterically destabilize the exo-b TS shown in Figure 1. As
expected, the cis relationship found between the hydrogen
atoms at positions 3 and 6 from the NMR data (see
Supporting Information) confirmed that the major product
corresponds to the endo alkylation (3a:3b 70:30) (Scheme
2). To further explore our hypothesis, additional reactions
were performed for p-nitro and p-methoxy derivatives of
benzyl bromide. When the alkylation of oxazolopiperidone
2a was performed with p-methoxybenzyl bromide, only the
water
2.6
3.1
0.0
a Relative stabilities in the gas phase determined at the (top) MP2/
6-31+G(d), (middle) MP2/aug-cc-pVDZ, and (bottom) composite MP2/
aug-cc-pVDZ + (CCSD-MP2)/6-31G(d) levels. This latter value was
used to obtain the relative stabilities in solution. The relative energies
corrected by zero-point energies are given in parentheses.
The endo TS is destabilized relative to the exo (exo-b TS)
attack by 2.5 kcal/mol in the gas phase (Table 2). The origin
of such an stabilization is mainly due to dispersion, as the
two TSs have similar energies at the Hartree-Fock level
(energy differences of -0.1 and 0.8 kcal/mol for endo and
exo-a relative to exo-b). In agreement with our previous
studies,10a such a difference in stability is little affected upon
solvation. At first sight, the destabilization of the endo TS is
surprising, since the Newman projection along the enolate
carbon atoms CR-Cꢀ (Figure 1) indicates that the torsional
strain is lower in the TS leading to the endo product, as
already noted for the attack of alkyl halides.10a Nevertheless,
the preferential stability of the exo-b TS can be realized from
the secondary interaction between the C-H unit at the angular
8a-position and the benzene ring of the attacking group,
where the C-H is pointing toward the center of the benzene
(14) (a) Hobza, P.; Spirko, V.; Selzle, H. L.; Schlag, E. W. J. Phys. Chem.
A 1998, 102, 2501. (b) Hobza, P.; Spirko, V.; Havlas, Z.; Buchhold, K.; Reimann,
B.; Barth, H. D.; Brutschy, B. Chem. Phys. Lett. 1999, 299, 180. (c) Cubero, E.;
Orozco, M.; Hobza, P.; Luque, F. J. J. Phys. Chem. A 1999, 103, 6394.
(15) Cubero, E.; Orozco, M.; Luque, F. J. J. Phys. Chem. A 1999, 103, 315.
7758 J. Org. Chem. Vol. 73, No. 19, 2008

ReVersion in π-Facial StereoselectiVe Alkylation
FIGURE 2. Transition state structures for the endo (top) and exo (bottom) addition of benzyl chloride to the enolate of oxazolopiperidone 2a
coordinated to a Li+ cation, together with the Newman projections viewed from the directions along the CR-Cꢀ and Cꢀ-Cγ bonds (the forming
bond between CR and Cb is shown by a dotted line in Newman projections). Distances and torsional angles are in angstroms and degrees, respectively.
exo isomer was detected. In contrast, the alkylation of 2a
with p-nitrobenzyl bromide afforded a 2:3 mixture of endo/
exo isomers, thus suggesting that the presence of an electron-
withdrawing group on the benzyl moiety results in a less
effective C-H · · · π interaction. Overall, present results
suggest that the intrinsic diastereoselectivity preference in
the alkylation of the enolate of 2a, which would favor the
endo addition due to the larger torsional strain in the exo
TS, is reverted when the alkylating reagent is benzyl due to
the formation of the secondary C-H · · · π interaction.
lar conclusions were found in our previous studies about the
stereoselective alkylation of the enolate of 2a.10a However,
because cation-π interactions are strongly stabilizing,16we
decided to further explore whether Li+ coordination to the
enolate or to the benzyl group might affect the stereochemical
outcome.
The structures of the TSs corresponding to the addition of
benzyl chloride to the enolate of 2a coordinated to a Li+
cation were located (Figure 2). Due to computational
limitations, dimethyl ether was used to mimic two solvating
Several studies have suggested that coordination with Li+
cation might affect the stereochemical outcome in Meyers-
type enolate alkylations.7 Nevertheless, the potential influence
exerted by Li+ coordination is not supported by the experi-
mental results reported by Romo and Meyers2a nor by
theoretical studies recently reported by Ando9 for the
stereoselective alkylation of γ-butyrolactones enolates. Simi-
(16) (a) Ma, J.; Dougherty, D. A. Chem. ReV. 1997, 97, 1303. (b) Cubero,
E.; Luque, F. J.; Orozco, M. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 5976. (c)
Tsuzuki, S.; Yoshida, M.; Uchimaru, T.; Mikami, M. J. Phys. Chem. A 2001,
105, 769. (d) Reddy, A. S.; Zipse, H.; Sastry, G. N. J. Phys. Chem. B 2007,
111, 11546.
(17) (a) McKeever, L. D.; Waack, R.; Doran, M. A.; Baker, E. B. J. Am.
Chem. Soc. 1969, 91, 1057. (b) Bauer, W.; Winchester, W. R.; Schleyer, P. V. R.
Organometallics 1987, 6, 2371.
J. Org. Chem. Vol. 73, No. 19, 2008 7759

Soteras et al.
TABLE 3. Gibbs Free Energy Differences (at 298 K; kcal/mol)
between endo and exo Transition States for Addition of Benzyl
Chloride to the Enolate of Li+-Coordinated Oxazolopiperidone 2a
TABLE 4. Alkylation of Oxazolopiperidone Lactam 2a under
Different Experimental Conditions
entry
base
solvent
cosolvent
endo:exo
methoda
endo-a
endo-b
exo-a
exo-b
exo-c
1
2
3
4
5
LiHMDS
LiHMDS
LiHMDS
KHMDS
KHMDS
THF
toluene
THF
toluene
THF
1:9
3:7
1:9
3:7
1:9
MP2/6-31+G(d)
2.8 (2.9) 1.2 (0.7) 3.8 (4.6) 0.0 (0.0) 3.4 (3.0)
MP2/6-311++(d,p) 2.9 (3.0) 1.2 (0.7) 4.2 (5.1) 0.0 (0.0) 3.5 (3.2)
HMPA
a The relative electronic energies corrected by zero-point energies in
the gas phase are given in parentheses.
TABLE 5. Alkylation of Oxazolopyrrolidinone Lactam 4
THF molecules, as done previously by Ando.9 Moreover,
even though several aggregation states have been described
for alkyl lithiums in THF, there is a tendency to form
monomers as the size of the coordinating alkyl group is
enlarged.17 For instance, whereas methyl lithium is found to
be tetrameric in THF, tert-butyl lithium is a monomer.
Accordingly, for our purposes here, a model system formed
by direct coordination of solvated Li+ cation to the enolate
of the oxazolopiperidone seems plausible.
R1
endo:exo
5:1;13a2.9:113b
1:2;13a1:2.113b
Me
CH2C6H5
The same TSs (endo-a, exo-a, and exo-b) shown in Figure
1 were identified in the structures optimized for the Li+-
coordinated TSs (Figure 2). The major differences are the
shortening of the CR-Cb bonds by 0.11-0.16 Å (CR-Cb
distances of ∼2.53 Å), and the lengthening of the Cb-Cl
bond by 0.16-0.19 Å. Moreover, the exo-b TS retains the
C-H · · · π hydrogen bond (distance to the benzene ring of
2.98 Å), but pointing toward one of the benzene carbon
atoms, leading to the formation of a bond critical point (Fbond
) 0.0074 au).
merely resulted in a change in the endo:exo ratio (3:7), the
major isomer again being the exo product.
In summary, both theoretical and experimental results
indicate that the diastereomeric reversal observed for the
reaction with benzyl chloride reflects intrinsic differences in
the TSs leading to endo and exo products related to the
formation of weak C-H · · · π interactions. Both experimental
and theoretical studies of the benzene-methane cluster in the
gas phase indicate that the complexation energy amounts to
-1.1 kcal/mol.18 Even though this is a modest contribution,
the relative stability found between the endo and exo TSs
corresponding to the methyl alkylation of 2a was estimated
to be less than 1 kcal/mol.10a Accordingly, the formation of
a C-H · · · π bond in the benzyl alkylation of the enolates of
oxazolopiperidones has a direct role in modulating the
π-facial stereoselective outcome.
Stereoselective Alkylation of the Pyrrolidinone Enolate
4. The dramatic effect of a benzyl alkylating group on the
stereoselectivity of enolate alkylations has also been noticed
experimentally for the enolate of oxazolopyrrolidinone 4
(Table 5).13 It might then be speculated that the structure-
directed π-facial reversion mechanism reported above for the
benzyl addition to the enolate of oxazolopiperidone 2a could
also operate for the pyrrolidinone enolate. To verify this
hypothesis, computations were also performed to locate the
TSs for the alkylation of the pyrrolidinone ring.
Three and two TSs were identified for the endo and exo
additions, respectively (Figure 3). The CR-Cb distance varies
from 2.58 to 2.61 Å in both endo and exo TSs. According to
the Newman projections along the enolate carbon atoms
CR-Cꢀ, there is larger torsional strain in the TS leading to
the exo product, which would therefore favor the endo
addition. However, comparison of the relative stabilities
between TSs (Table 6) indicates that the most favorable
benzylation occurs at the exo face, the endo addition being
destabilized by 1.6 kcal/mol. Inspection of the optimized
structures reveals the formation of C-H · · · π interactions in
both endo-c (Fcage ) 0.0041 au) and exo-a (Fbond ) 0.0084
au) TSs, which involve hydrogens of the C-2 and C-7
methylene groups, respectively. The former TS, however, is
destabilized by 2.0 kcal/mol relative to the exo-a TS, as
In addition, two new TSs where Li+, besides interacting
with the carbonyl oxygen of the enolate, forms a cation · · · π
interaction with the benzene ring of the incoming reactant
were located (endo-b and exo-c in Figure 2). Compared to
the preceding TSs, the length of the CR-Cb bond is shorter
(endo-b 2.42 Å; exo-c 2.40 Å), while the length of the Cb-Cl
bond is mostly unaffected (endo-b 2.35 Å; exo-c 2.36 Å).
The distance of the Li+ cation to the center of the benzene
ring (endo-b 3.31 Å; exo-c 3.13 Å) is much larger than the
value optimized for the benzene · · · Li+ complex (1.94 Å at
the MP2/6-311++G(d,p) level).16d This finding, in conjunc-
tion with the fact that Li+ deviates notably from the normal
to the ring, suggests that the cation · · · π interaction is weak,
as expected from the dominant contribution played by the
electrostatic attraction with the negative charge of the enolate.
The relative stabilities of the endo-a, exo-a, and exo-b TSs
are little affected by Li+ coordination (Table 3) and reflect
the values reported in Table 2. Moreover, the endo-b and
exo-c TSs are destabilized with regard to the exo-b TS by
1.2 and 3.4 kcal/mol, respectively, which can be attributed
to repulsive contacts between the benzylic CH2 unit and
hydrogen atoms of the enolate (see Figure 2). Overall, the
results point out that the exo-b TS is the preferred route for
the addition of benzyl chloride to the enolate of oxazolopi-
peridone 2a.
To further check the preceding conclusion, the stereo-
chemical outcome in the reaction of the enolate of lactam
2a with benzyl bromide was experimentally examined under
different conditions in order to explore the effect of solvent
(THF vs toluene), cation (Li+ vs K+), and cosolvent (HMPA)
on the diastereomeric endo:exo ratio. The results (Table 4)
demonstrate that the use of KHMDS as the base instead of
LiHMDS or the addition of HMPA did not change the ratio
of the final products, while performing the reaction in toluene
(18) Shibasaki, K.; Fujii, A.; Mikami, N.; Tsuzuki, S. J. Phys. Chem. A 2006,
110, 4397.
7760 J. Org. Chem. Vol. 73, No. 19, 2008

ReVersion in π-Facial StereoselectiVe Alkylation
FIGURE 3. Transition structures for the endo and exo addition of benzyl chloride to the enolate of pyrrolidinone 4, together with the Newman
projections viewed from the directions along the CR-Cꢀ and Cꢀ-Cγ bonds (the forming bond between CR and Cb is shown by a dotted line in
theNewman projections). Distances and torsional angles are in angstroms and degrees, respectively.
TABLE 6. Gibbs Free Energy Differences (at 298 K; kcal/mol)
between endo and exo Transition States for Addition of Benzyl
Chloride to the Enolate of Pyrrolidinone 4
Conclusion
Previous studies have shown that C-H · · · π interactions
modulate the diastereselectivity of host-guest complexes.19
The results presented here demonstrate that the reversion of
the π-facial selectivity observed experimentally in the benzyl
bromide alkylation of the enolates of both ozaxolopiperidone
2a and oxazolopyrrolidinone 4 can be attributed to the
formation of a C-H · · · π bond between the enolate and the
benzene ring of the incoming reagent. Whereas in 2a such
an interaction involves the C-H unit of the C8a angular
position, the stereochemical reversion observed in 4 is due
to the interaction with one of the C-7 methylene hydrogens.
Therefore, the subtle balance among different factors that
mediate the stereochemical outcome in the alkylation of chiral
bicyclic [4.3.0] and [3.3.0] lactams can be altered by
unexpected weak secondary interactions. These findings may
methoda
endo-a endo-b endo-c exo-a
exo-b
MP2/6-31+G(d)
3.1 (3.6) 1.8 (2.8) 2.6 (2.6) 0.0 (0.0) 1.4 (1.9)
4.1 (4.5) 2.3 (3.2) 2.8 (2.7) 0.0 (0.0) 1.8 (2.3)
2.9 (3.4) 1.6 (2.7) 2.0 (1.9) 0.0 (0.0) 1.5 (2.0)
MP2/aug-cc-pVDZ
MP2/aug-cc-pVDZ +
(CCSD-MP2)/6-31G(d)
a The relative electronic energies corrected by zero-point energies in
the gas phase are given in parentheses.
expected from the more crowded structure of the endo-c TS,
which is in fact slightly less stable than the endo-b TS (see
Table 6).
(19) Frontera, A.; Garau, C.; Quin˜onero, D.; Ballester, P.; Costa, A.; Deya`,
P. M. Org. Lett. 2003, 5, 1135.
J. Org. Chem. Vol. 73, No. 19, 2008 7761

Soteras et al.
Gas phase calculations were carried out using Gaussian03.28
MST calculations were performed using a locally modified
version of Gaussian-03.
provide useful guidelines to tune the selective preparation
of enantiopure derivatives, which is a crucial aspect in the
design of bioorganic and pharmaceutical compounds.
Experimental Procedures. The generation of the enolate of
2a was carried out at low temperature (-78 °C) for 1 h using
lithium bis(trimethylsilyl)amide (1 M in THF, 1.5 equiv) as the
base, followed by addition of the alkylating reagent (2.5 equiv).
The reaction mixture was then stirred at -78 °C for 2 h and
quenched by addition of a saturated aqueous solution of sodium
chloride. Reactions were performed using THF as the solvent.
Additional experiments were carried out upon lactam 2a to
evaluate the effect that the solvent (THF, toluene), the cation of
the base (Li+, K+), and cosolvent (HMPA, 2 equiv) could have
on the diastereomeric endo:exo ratio.
Experimental Section
Computational Details. The computational protocol used here
follows the main trends adopted in our previous study about
the origin of the π-facial stereoselectivity in the methyl alkylation
of the enolates of oxazolopiperidones 1 and 2.10a On the basis
of previous studies,10a,20 which showed that single point MP2
energies at the B3LYP geometries provide a reasonable estimate
of the lithium enolate activation barrier, full geometry optimiza-
tions were performed with the B3LYP21 density functional
method using the 6-31+G(d) basis set, and single-point MP2/
6-31+G(d) and MP2/aug-cc-pVDZ calculations were subse-
quently performed to determine the relative energies of the endo
and exo transition states. Higher-order electron correlation effects
were estimated from CCSD/6-31G(d) calculations, and the best
estimate of the energy difference between transition states was
determined by combining the relative energies computed at the
MP2/aug-cc-pVDZ level and the energy correction obtained from
CCSD and MP2 calculations performed with the 6-31G(d) basis
set [denoted in the following as MP2/aug-cc-pVDZ + (CCSD-
MP2)/6-31G(d)]. The nature of the stationary points was verified
by inspection of the vibrational frequencies, which were used
to calculate zero point, thermal and entropic corrections within
the framework of the harmonic oscillator-rigid rotor at 1 atm
and 298 K. These corrections were added to the electronic
energies to estimate the free energy differences in the gas phase.
[3R,6R(and 6S),8aR]-6-Benzyl-8a-methyl-5-oxo-3-phenyl-
2,3,6,7,8,8a-hexahydro-5H-oxazolo[3,2-a]pyridine (3a and 3b)
and (3R,8aR)-6,6-Dibenzyl-8a-methyl-5-oxo-3-phenyl-2,3,6,7,8,8a-
hexahydro-5H-oxazolo[3,2-a]pyridine (3c). A solution of the
lactam 329 (200 mg, 0.87 mmol) in THF (2 mL) was added to
a cooled (-78 °C) solution of LiHMDS (1 M in THF, 1.0 mL,
1.0 mmol) in THF (10 mL). After the solution stirred at -78
°C for 1 h, benzyl bromide (0.15 mL, 1.31 mmol) was added,
and stirring was continued for 2 h. The reaction was quenched
by the addition of saturated aqueous NaCl, and the resulting
mixture was extracted with EtOAc and CH2Cl2. The combined
organic extracts were dried and concentrated, and the resulting
residue was chromatographed (1:1 hexane-EtOAc) to afford a
mixture of epimers 3a and 3b [165 mg, 59%; (70:30 calculated
by GC/MS)], dialkylated product 3c (50 mg, 14%), and starting
lactam 3 (34 mg, 17%). 3a: IR (film) 1654 cm-1; 1H NMR (300
MHz) δ 1.34 (s, 3H, CH3), 1.56 (m, 1H, H-7), 1.72 (m, 1H,
H-7), 1.90-1.97 (m, 2H, H-8), 2.39 (m, 1H, H-6), 2.46 (dap, J
) 9.6 Hz, 1H, CH2Ph), 2.99 (dap, J ) 9.6 Hz, 1H, CH2Ph),
3.83 (dd, J ) 9.0, 2.0 Hz, 1H, H-2), 4.33 (dd, J ) 9.0, 7.0 Hz,
1H, H-2), 4.81 (dd, J ) 7.0, 2.0 Hz, 1H, H-3), 6.96-7.22 (m,
10H, ArH); 13C NMR (75.4 MHz) δ 21.7 (C-7), 23.8 (CH3),
32.5 (C-8), 37.1 (CH2Ph), 40.7 (C-6), 58.8 (C-3), 71.5 (C-2),
93.1 (C-8a), 126.0 (CH), 126.3 (2CH), 127.3 (CH), 128.2 (2CH),
128.4 (2CH), 129.1 (2CH), 139.7 (C i), 141.7 (C i), 169.4 (NCO);
[R]22D +43 (c 0.9, MeOH); MS-EI m/z 321 (M+, 43), 306 (100),
187 (12), 162 (39), 120 (92), 91 (89); HMRS calcd for
C21H23NO2 (M+ + 1) 322.1801, found 322.1804. 3b: IR (film)
The relative stabilities in solution were estimated by com-
bining the free energy difference in the gas phase and the
differences in solvation free energy determined from MST
calculations.22 QM SCRF continuum calculations were per-
formed by using the B3LYP/6-31G(d) optimized version of the
MST(IEF) model,23 which relies on the integral equation
formalism (IEF)24 of the polarizable continuum model.25 The
MST model determines the solvation free energy by three terms,
which account for electrostatic, cavitation, and van der Waals
contributions associated with the transfer of the solute from the
gas phase to solution. Since the MST model has not been
parametrized to treat solvation in tetrahydrofuran (i.e., the solvent
used in experimental assays; see below),26 calculations were
performed for the solvation in water, octanol, chloroform and
carbon tetrachloride in order to explore the effect of varying
the polarity of the solvent.23,27
1
1647 cm-1; H NMR (300 MHz) δ 1.12 (s, 3H, CH3), 1.57 (m,
1H, H-7), 1.77-1.91 (m, 2H, H-7, H-8), 2.09 (dt, J ) 13.0, 4.0
Hz, 1H, H-8), 2.43 (m, 1H, H-6), 2.79 (dd, J ) 13.5, 8.0 Hz,
1H, CH2Ph), 3.07 (dd, J ) 13.5, 4.0 Hz, 1H, CH2Ph), 3.86 (dd,
J ) 9.5, 2.0 Hz, 1H, H-2), 4.37 (dd, J ) 9.5, 7.5 Hz, 1H, H-2),
4.85 (dd, J ) 7.5, 2.0 Hz, 1H, H-3), 7.09-7.26 (m, 10H,
ArH);13C NMR (75.4 MHz) δ 23.1 (CH3), 23.3 (C-7), 34.3 (C-
8), 37.3 (CH2Ph), 42.8 (C-6), 59.2 (C-3), 71.4 (C-2), 93.1 (C-
8a), 126.2 (3CH), 127.3 (CH), 128.2 (2CH), 128.5 (2CH), 129.4
(20) Pratt, L. M.; Van Nguyen, N.; Ramachandran, B. J. Org. Chem. 2005,
70, 4279.
(21) (a) Becke, A. B. J. Chem. Phys. 1993, 98, 5648. (b) Becke, A. B. Phys.
ReV. A 1998, 38, 3098. (c) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988,
37, 785.
(22) Since the MST model was parametrized from experimental free energies
of solvation at 298 K, the gas phase free energy difference was also determined
at this temperature for the sake of internal consistency.
(23) (a) Bachs, M.; Luque, F. J.; Orozco, M. J. Comput. Chem. 1994, 15,
446. (b) Curutchet, C.; Orozco, M.; Luque, F. J. J. Comput. Chem. 2001, 22,
1180. (c) Soteras, I.; Curutchet, C.; Bidon-Chanal, A.; Orozco, M.; Luque, F. J.
J. Mol. Struct. (THEOCHEM) 2005, 727, 29.
(24) (a) Cance`s, E.; Mennucci, B. J. Math. Chem. 1998, 23, 309. (b) Cance`s,
E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107, 3032. (c) Mennucci, B.;
Cance`s, E.; Tomasi, J. J. Phys. Chem. B 1997, 101, 10506.
(25) (a) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117. (b)
Miertus, S.; Tomasi, J. Chem. Phys. 1982, 65, 239.
(26) The dielectric permittivity is 7.5 at 295 K. Data taken from Handbook
of Chemistry and Physics, 80th ed.; Lide, D. R., Ed.; CRC: Boca Raton, 1999.
(27) (a) Orozco, M.; Bachs, M.; Luque, F. J. J. Comput. Chem. 1995, 16,
563. (c) Luque, F. J.; Zhang, Y.; Aleman, C.; Bachs, M.; Gao, J.; Orozco, M. J.
Phys. Chem. 1996, 100, 4269. (d) Luque, F. J.; Alema´n, C.; Bachs, M.; Orozco,
M. J. Comput. Chem. 1996, 17, 806.
(2CH), 139.3 (C i), 141.8 (C i), 168.6 (NCO); [R]22 -129 (c
D
0.58, MeOH); MS-EI m/z 321 (M+, 34), 306 (100), 187 (12),
162 (30), 120 (82), 91 (84); HMRS calcd for C21H23NO2 (M+
1
+ 1) 322.1801, found 322.1797. 3c: IR (film) 1642 cm-1; H
NMR (300 MHz) δ 0.63 (s, 3H, CH3), 1.77-1.92 (m, 4H, H-7,
H-8), 2.23 (d, J ) 12.5 Hz, 1H, CH2Ph), 2.67 (d, J ) 13.5 Hz,
1H, CH2Ph), 2.94 (d, J ) 13.5 Hz, 1H, CH2Ph), 3.42 (d, J )
12.5 Hz, 1H, CH2Ph), 3.92 (dd, J ) 9.2, 2.2 Hz, 1H, H-2), 4.35
(dd, J ) 9.2, 7.2 Hz, 1H, H-2), 4.83 (dd, J ) 7.2, 2.2 Hz, 1H,
H-3), 6.99-7.02 (m, 3H, ArH), 7.16-7.41 (m, 12H, ArH); 13C
NMR (75.4 MHz) δ 22.0 (CH3), 23.7 (C-7), 31.9 (C-8), 42.8
(CH2Ph), 44.9 (CH2Ph), 47.8 (C-6), 59.1 (C-3), 71.4 (C-2), 93.1
(C-8a), 126.3 (CH), 126.5 (CH), 126.9 (2CH), 127.3 (CH), 128.0
(4CH), 128.3 (2CH), 131.0 (2CH), 131.0 (2CH), 137.0 (C i),
(28) Frisch, M. J. et al. Gaussian 03, ReVision B.04; Gaussian, Inc.: Pittsburgh
PA, 2003.
(29) Munchhorf, M. J.; Meyers, A. I. J. Org. Chem. 1995, 60, 7084.
7762 J. Org. Chem. Vol. 73, No. 19, 2008

ReVersion in π-Facial StereoselectiVe Alkylation
138.0 (C i), 141.9 (C i), 170.7 (NCO); [R]22 -42 (c 1.0,
gratefully acknowledged. Thanks are also due to the MEC for
fellowships to I.S. and O.L.
D
MeOH); MS-EI m/z 411 (M+, 1), 396 (39), 320 (100), 200 (6),
120 (22), 91 (72); HMRS calcd for C28H29NO2 (M+ + 1)
412.2171, found 412.2265.
Supporting Information Available: Complete ref 22, B3LYP/
6-31+G(d) geometries of the transition states for the reaction
of enolates derived from 2a and 4 with benzyl chloride, and
Acknowledgment. Financial support from the Spanish Min-
isterio de Educacio´n y Ciencia (MEC; projects CTQ2006-02390/
BQU and CTQ2005-08797-C02-01/BQU), the DURSI, Gener-
alitat de Catalunya (Grant 2005SGR-0603), and computational
facilities from the Centre de Supercomputacio´ de Catalunya are
1
copies of the H and 13C NMR spectra of compounds 3a, 3b,
and 3c. This material is available free of charge via the Internet
at http://pubs.acs.org.
JO801665K
J. Org. Chem. Vol. 73, No. 19, 2008 7763