Synthesis of ethyl 5-hydroxyisoxazolidine-4-carboxylates via michael addition/intramolecular hemiketalisation

Article abstract of DOI:10.1002/ejoc.200800719

The 1,4-addition of N,O-bis(trimethylsilyl)hydroxylamine to alkylideneacetoacetates gave, in high yield, new 5-hydroxyisoxazolidine-4- carboxylates. The results of the accurate computational investigation on the mechanism at the DFT level are in complete

Full text of DOI:10.1002/ejoc.200800719

FULL PAPER
DOI: 10.1002/ejoc.200800719
Synthesis of Ethyl 5-Hydroxyisoxazolidine-4-carboxylates via Michael
Addition/Intramolecular Hemiketalisation
Fides Benfatti,[a] Andrea Bottoni,[a] Giuliana Cardillo,*[a] Luca Gentilucci,[a]
Magda Monari,[a] Elisa Mosconi,[a] Marco Stenta,*[a] and Alessandra Tolomelli[a]
Keywords: Michael addition / Alkylideneacetoacetates / N,O-Heterocycles / Hemiketalisation / Computer chemistry / Reac-
tion mechanism
The 1,4-addition of N,O-bis(trimethylsilyl)hydroxylamine to
alkylideneacetoacetates gave, in high yield, new 5-hydroxy-
isoxazolidine-4-carboxylates. The results of the accurate
computational investigation on the mechanism at the DFT
level are in complete agreement with the experimental evi-
dence and the crystallographic data.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
report the results of an accurate computational investiga-
tion at the Density Functional Theory (DFT) level of the
mechanism of this reaction. The computational finding sat-
isfactorily explains the stereochemical outcome.
Introduction
The conjugate addition of nitrogen-containing nucleo-
philes to electron-deficient olefins represents one of the
most employed and versatile methods for C–N bond con-
struction in organic chemistry.[1] The reaction may be per-
formed under catalytic conditions, and the use of chiral
Lewis acids[2] or, more recently, of organocatalysts[3] has
been described in the literature.
Results and Discussion
Alkylideneacetoacetates have been prepared by
a
Knøvenagel reaction between ethyl acetoacetate and vari-
ous aldehydes in the presence of a catalytic amount of pro-
line (Scheme 1).[10] Proline was chosen among other bases
since it was demonstrated to afford the best results in terms
of yields and (Z)/(E) product selectivity.
Alkylidenemalonates have been intensively employed as
Michael acceptors.[4] In particular, a chiral Lewis acid cata-
lysed Michael addition of hydroxylamine derivatives to alk-
ylidenemalonates has been reported by our group.[5] Herein,
we describe the highly stereocontrolled synthesis of ethyl
5-hydroxyisoxazolidine-4-carboxylate through a Lewis acid
induced Michael addition of hydroxylamine derivatives to
alkylideneacetoacetates, followed by intramolecular hemi-
ketal formation. The use of acetoacetates in this field is
rather unusual and has the advantage of introducing a reac-
tive keto functionality that may be further elaborated.
Isoxazolidines are interesting heterocyclic compounds
that may be regarded as unusual constrained β-amino ac-
ids[6] or as furanose mimetics,[7] and have also been ex-
ploited as analogues of natural products.[8]
Scheme 1. Synthesis of alkylideneacetoacetates 1a–d.
The stereochemical assignment for 1a–d was made on the
basis of DPFGSE-NOE (Double Pulse Field Gradient Spin
Echo NOE) experiments. We found that the major isomer
(Z) exhibited a strong NOE between the vinyl H atom and
the ketone CH3 group, whereas the minor isomer (E)
showed a strong NOE between the vinyl H atom and the
ester CH2 group.
An organocatalytic synthesis of 5-hydroxyisoxazolidine
from N-protected hydroxylamines and α,β-unsaturated al-
dehydes has been previously reported by Córdova and co-
workers.[9] Our contribution concerns the description of a
novel and straightforward Lewis acid catalysed protocol
and the study of more complex substrates. In addition, we
The conjugate addition of N,O-bis(trimethylsilyl)hydrox-
ylamine (TMSONHTMS) to alkylideneacetoacetates was
first examined in the absence of catalysts (Scheme 2). The
reaction conditions were optimized in order to induce the
regioselective formation of 1,4-addition products rather
than oximes, which would derive from a 1,2-addition pro-
[a] Department of Chemistry “G. Ciamician,” University of
Bologna,
Via Selmi 2, 40126 Bologna, Italy
Fax: +39-051-2099456
E-mail: giuliana.cardillo@unibo.it
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2008, 6119–6127
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6119

G. Cardillo, M. Stenta, et al.
FULL PAPER
cess.[11] The reactions were completed in 16 h, and the syn mediates of the 1,4-addition. The conjugate addition of
and anti 1,4-adducts 2a–d were obtained in quantitative TMSONHTMS to the methyl ester of isopropylideneace-
yields in a 60:40 (Z)/(E) ratio, as indicated by the 1H NMR toacetate gave results in complete agreement with those ob-
spectrum of the crude reaction. The loss of the trimethyl- tained for the corresponding ethyl esters.[13] This allowed
silyl group on the nitrogen atom occurred during reaction for the simplification of the model system as reported in
workup. In order to characterize the intermediates, a very Scheme 5. It is reasonable to believe that, after the C–N
fast chromatography on silica gel allowed 2a–d to be ob- bond formation, the resulting species has a zwitterionic na-
tained as inseparable mixtures of syn/anti isomers. The ture, with a positive charge on the nitrogen atom and a
1
stereochemical assignment was made by comparing the H negative charge delocalised on the enolate moiety.[14] The
NMR coupling constants of 2a–d with data available in the optimum structures of this intermediate, either as a free spe-
literature[12] (Jsyn = 3.6 Hz, Janti = 7.2 Hz).
cies (IntF) or as a complex with the Lewis acid (IntC), have
been computed (Scheme 5).
Scheme 2. Michael addition of TMSONHTMS to alkylideneace-
toacetates 1a–d.
Notably, the reaction carried out on the pure (Z) or (E)
isomer of 1a–d always gave 2a–d in the same syn/anti ratio,
as did the mixture of isomeric starting material. When this
reaction was performed in the presence of a catalyst [scan-
dium(III) triflate or copper(II) triflate], the 1,4-addition
quickly occurred (0.5–1 h), and the resulting adducts were
isolated as silyl enol ether derivatives 3a–d (Scheme 3).
Scheme 5. Keto and enol resonance structures for intermediates
IntC and IntF.
A schematic representation of the two possible resonance
structures (keto and enol) for these two intermediate species
is given in Scheme 5. A comparison between the C–C and
the C–O enolate bond lengths in these molecules shows
that, in the case of IntF, the keto form is dominant, whereas
for IntC the enol form predominates (Figure 1). The accu-
rate analysis performed on IntC and IntF is reported in the
Supporting Information.
Scheme 3. Lewis acid catalysed Michael addition of
TMSONHTMS to alkylideneacetoacetates 1a–d.
The structure of 3a–d has been proposed on the basis of
1
their H and 13C NMR spectra. Further evidence of the
nature of 3a–d was obtained by treating the 1,4-adduct 2a
with LHMDSA and TMSCl (Scheme 4).
Scheme 4. Conversion of adducts 2a into silyl enol ether 3a.
To summarize the results thus far, starting from an in-
separable 60:40 mixture of syn/anti isomers of 2a, the reac-
tion gave 3a, whose (Z) configuration was assigned by
means of NOE experiments and could be extended to 3b–
d on the basis of the complete regularity of their NMR
spectra.
Figure 1. Optimum structures of the two intermediates IntF and
IntC. Bond lengths [Å] and Wiberg bond orders are reported in
round parentheses and square brackets, respectively (see the Sup-
porting Information for details).
An explanation for the different outcomes of the uncata-
lysed (adducts 2a–d in Scheme 2) and catalysed (silyl enol
This finding suggests that IntF is likely to end up with a
ethers 3a–d in Scheme 3) conjugate additions may come proton transfer to the α-carbon atom to give products 2a–
from DFT theoretical computations on the plausible inter- d. On the other hand, it is reasonable to believe that IntC,
6120
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 6119–6127

Synthesis of Ethyl 5-Hydroxyisoxazolidine-4-carboxylates
because of its structural features, may transfer a TMS group source that facilitates ready desilylation of the reagent (e.g.,
from the nitrogen atom to the enol oxygen atom to afford treatment with aqueous acid solution or acidic resins).[15]
products 3a–d.
The ethyl 5-hydroxyisoxazolidine-4-carboxylates 4a–d were
Adducts 2a–d and silyl enol ethers 3a–d spontaneously fully characterised by NMR spectroscopy (DEPT,
converted to ethyl 5-hydroxyisoxazolidine-4-carboxylates HETCOR, COSY, and NOE).
4a–d at room temperature. In both cases, the intramolecular
We carried out a theoretical investigation on the reaction
hemiketalisation required, on average, 24 h to be completed. reported in Scheme 6 with the model system depicted in
Remarkably, both classes of 2a–d and 3a–d gave, with quan- Figure 2. In the following discussion, the atom numbers
titative conversion, mixtures ranging from an 80:20 to a given in parentheses refer to the numbering shown in Fig-
75:25 ratio of the two C-5 stereoisomers of (3,4)-trans-hy- ure 2. The chosen model system (2e) emulates the ring-clo-
droxyisoxazolidines 4a–d (Scheme 6). The intramolecular sure process to afford 4a–d starting from the desilylated
hemiketalisation can be accelerated by the addition of silica keto species 2a–d, which represents the active form for the
gel to the crude mixture or by the use of another acidity reaction. The enol forms 3a–d were not explicitly consid-
ered since the species should immediately transform to 2a–
d after hydrolysis and expulsion of the TMS groups.
Figure 2. Schematic representation of the model system emulating
the ring closure.
For both the anti (S*,R*) and syn (S*,S*) isomers of the
model system, only three stable conformers exist. These are
Scheme 6. Intramolecular hemiketalisation to 4a–d.
characterized by a different H-bond pattern. In M1a and
Figure 3. Schematic representation of the various reaction pathways for the nucleophilic attack of O(12) on the Re and Si face of the
(S,R) and (S,S) isomers, respectively. The reported energy values [kcalmol–1] are the sum of the molecular total energy (electronic energy
+ nuclear energy) and the COSMO solvation energy. All values in round parentheses are relative to M1a. ∆E# values are the computed
activation energies.
Eur. J. Org. Chem. 2008, 6119–6127
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6121

G. Cardillo, M. Stenta, et al.
FULL PAPER
M1b (see Figure 3) H(13) and H(14) form H bonds with were determined. In both cases (the corresponding transi-
O(3) and O(5), respectively. A slight rotation around the tion states are TS1a and TS1b) the nucleophilic attack and
N(11)–C(9) bond (see column D1 in Table 1) leads to M2a the proton transfer appear to be concerted (see Figure 4).
and M2b, where both H(13) and H(14) point toward the The nucleophilic attack on the Si face of the carbonyl group
same oxygen atom, O(5). The lost interaction with O(3) al- has also been explored. We found two reaction pathways
lows a rotation around the C(2)–C(4) bond and, conse- leading to M5a and M5b from M3a (transition state TS2a)
quently, a different orientation of the C(2)–O(3) carbonyl and M3b (transition state TS2b), respectively (see Figure 5).
group (see column D2 in Table 1). The energy values of Thus, M4b and M5b represent the two possible trans prod-
these four isomers range from –1.14 to 2.96 kcalmol–1 (see ucts originating from the (S*,S*) isomer, whereas M4a and
Figure 3). In both M2a and M2b, the relative orientation of M5a are the cis products originating from the (S*,R*) iso-
the various atoms becomes suitable for the intramolecular mer.
nucleophilic attack of O(12) on C(2) on the Re face of the
carbonyl group. However, a further analysis of the confor-
mational space showed the existence of two additional con-
formers (M3a and M3b), where a rotation around the C(9)–
C(4) bond (see column D3 in Table 1) allows for the attack
of the nucleophilic oxygen atom O(12) on the Si face of the
carbonyl group.
The computed activation barriers and the relative energy
values of the various transition states indicated that the for-
mation of M4b and M5b was favoured over M4a and M5a;
it was evident from the relative energy values reported in
Figure 3 that in the former case the corresponding transi-
tion states (TS1b and TS2b) were significantly lower in en-
ergy than those of the latter (TS1a and TS2a). This can be
ascribed to a more effective H-bond network characterized
by a better co-linearity of the atoms involved in proton
transfer. Also, since the energy difference between TS1b
and TS2b is only 1.0 kcalmol–1, the reaction should afford
both cyclic trans products differing in the configuration at
C-5, in agreement with the experimental observation. How-
ever, since the anti and syn isomers are simultaneously pres-
ent in the reaction mixture, some amounts of the cyclic cis
products should also be experimentally observed. Alterna-
tively, because of the high energy of TS1a and TS2a, a cer-
tain amount of the unreacted anti species should be de-
tected in the final product mixture.
Table 1. Computed values for the various dihedral angles describ-
ing the different conformations of reactants.
Entry
Conformer
Dihedral angle
D2[b]
D1[a]
D3[c]
1
2
3
4
5
6
M1a
M2a
M3a
M1b
M2b
M3b
80.0
39.8
–83.2
76.1
41.4
–74.3
–92.8
10.8
152.4
–88.6
2.3
–69.0
–72.1
76.5
–70.0
–68.4
56.9
118.7
[a] Defined as C(12)–N(11)–C(9)–C(4). [b] Defined as O(3)–C(2)–
C(4)–C(6). [c] Defined as N(11)–C(9)–C(4)–C(2).
Since only the two (3,4)-trans isomer products were ob-
The intramolecular nucleophilic attack of O(12) on the tained experimentally, this suggests that a syn/anti isomer-
Re carbonyl face of C(2) and the simultaneous proton ization took place during the cyclization, giving only the
transfer of H(13) from O(12) to O(3) leads from M2a and kinetically favoured (3,4)-trans derivatives. NOE experi-
M2b to the cyclic adducts M4a and M4b, respectively. Two ments on 4a confirmed the relative stereochemistry of the
unique reaction pathways (one for M2a and one for M2b) ring substituents. A further validation of the structure of
Figure 4. Schematic representation of reactants (M2a and M2b), transition states (TS1a and TS1b), and products (M4a and M4b) for
the nucleophilic attack on the Re face of the carbonyl group.
6122
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 6119–6127

Synthesis of Ethyl 5-Hydroxyisoxazolidine-4-carboxylates
Figure 5. Schematic representation of reactants (M3a and M3b), transition states (TS2a and TS2b), and products (M5a and M5b) for
the nucleophilic attack on the Si face of the carbonyl group.
ethyl
5-hydroxy-5-methyl-3-isopropylisoxazolidine-4-car-
DFT computations were carried out on the model system
boxylate (4a) was obtained by X-ray diffraction analysis of depicted in Figure S5 (see the Supporting Information) to
the derivative 5a, easily obtained by the treatment of 4a investigate the structure and the stereochemistry of the N-
with 3,5-dinitrobenzoyl chloride and triethylamine, as (3,5-dinitrobenzoyl) derivative 5a obtained from the isox-
shown in Scheme 7.
azolidine 4a. All eight potential isomers were considered.[16]
We found a strong preference for the isomers having a trans
configuration of the amide bond. In particular, the
(3S*,4S*,5R*) isomer was the most stable structure. Its ab-
solute configuration is in agreement with the results of the
X-ray diffraction analysis of derivative 5a, thus confirming
the reliability of the computational approach adopted in
this work (Figure 6).
Scheme 7. Derivatisation of 4a leading to 5a.
Note that the relative configurations at C3 and C4 in the
ORTEP drawing are in agreement with those of the domi-
nant products forecasted by the computational results (Fig-
ure 6). Starting from a 78:22 diastereomeric mixture of 4a,
only one (3,4)-trans isomer (J3-H,4-H = 7.2 Hz) of 5a was
obtained, suggesting that an isomerisation occurred at the
hemiketal position during the reaction. Under the same
conditions, 4b–c afforded exclusively (3,4)-trans-5b–c. The
transformation of 4d to 5d gave unsatisfactory results and
a partial degradation of the starting material. Complete
regularity in the 1H NMR chemical shifts allowed us to
extend the stereochemical attribution of 5a to 5b–c.
Conclusions
In this paper we described a new protocol for the prepa-
ration of functionalized 3,4-disubstituted 5-hydroxyisox-
azolidines through the 1,4-addition of TMSONHTMS to
alkylideneacetoacetates. We carried out a computational
DFT investigation of the reaction mechanism leading to 5-
hydroxyisoxazolidine-4-carboxylates. Our computations
satisfactorily explain the experimentally observed stereo-
chemical outcome of the reaction and are in agreement with
the crystallographic analysis.
Experimental Section
General: All chemicals were purchased from commercial suppliers
and used without further purification. Anhydrous solvents were
purchased in Sure-seal bottles over molecular sieves and used with-
out further drying. Flash chromatography was performed with sil-
ica gel (230–400 mesh). NMR spectra were recorded with 200, 300,
or 600 MHz spectrometers. Chemical shifts were reported as δ val-
ues (ppm) relative to the solvent peak of CDCl3 set at δ = 7.27
(1H) or δ = 77.0 (13C). The preparation of acetoacetate 1 and the
complete characterization of 1a and 1c have already been re-
ported.[17] 1b was prepared from the corresponding racemic alde-
Figure 6. ORTEP drawing of 5a.
Eur. J. Org. Chem. 2008, 6119–6127
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6123

G. Cardillo, M. Stenta, et al.
FULL PAPER
hyde. Due to their easy conversion to 4a–d, 2a–d and 3a–d were
not submitted to elemental analysis.
CHCH2CH3, CHCH3), 1.26 (t, J = 6.9 Hz, 3 H, OCH2CH3), 1.46
(m, 1 H, CH3CHCH2), 2.23 (s, 3 H, OCCH3), 3.36 (m, 1 H, NCH),
3.42 (d, J = 9.6 Hz, 1 H, CHCO), 4.18 (m, 2 H, OCH2CH3), 6.0
(br. s, 1 H, NH) ppm. 13C NMR (CDCl3, 300 MHz): δ = –1.1
(3CH3), 11.2 (CH3), 14.0 (CH3), 16.3 (CH3), 26.4 (CH2), 28.6
(CH3), 35.6 (CH), 60.0 (CH2), 60.9 (CH), 67.5 (CH), 170.0 (C),
202.8 (C) ppm.
anti-2b: 1H NMR (CDCl3, 300 MHz): δ = 0.06 [s, 9 H, OSi-
(CH3)3], 0.66–1.06 (m, 8 H, CHCH2CH3, CHCH2CH3, CHCH3),
1.18 (m, 1 H, CH3CHCH2), 1.25 (t, J = 6.9 Hz, 3 H, OCH2CH3),
2.27 (s, 3 H, OCCH3), 3.69 (d, J = 8.1 Hz, CHCO), 3.83 (m, 1 H,
NCH), 4.18 (m, 2 H, OCH2CH3), 6.2 (br. s, 1 H, NH) ppm. 13C
NMR (CDCl3, 300 MHz): δ = –0.6 (3 CH3), 11.8 (CH3), 13.9
(CH3), 15.6 (CH3), 26.7 (CH2), 29.7 (CH3), 36.0 (CH), 59.3 (CH2),
61.2 (CH), 68.4 (CH), 168.9 (C), 200.3 (C) ppm.
(Z)-1b: Yield 66%, colourless clear oil. 1H NMR (CDCl3,
300 MHz): δ = 0.75 (t, J = 6.9 Hz, 3 H, CH3CH2CH), 0.97 (d, J =
6.6 Hz, 3 H, CHCH3), 1.22 (t, J = 7.2 Hz, 3 H, CH3CH2O), 1.34
(m, 2 H, CH3CH2CH), 2.21 (s, 3 H, COCH3), 2.35 (m, 1 H,
CH3CH2CH), 4.20 (q, J = 7.2 Hz, 2 H, CH3CH2O), 6.50 (d, J =
10.8 Hz, 1 H, CH=) ppm. 13C NMR (CDCl3, 300 MHz): δ = 11.5
(CH3), 13.9 (CH3), 19.3 (CH), 26.5 (CH2), 29.1 (CH3), 36.3 (CH3),
60.9 (CH2), 136.0 (C), 152.8 (CH), 166.4 (C), 195.0 (C) ppm.
1
(E)-1b: H NMR (CDCl3, 300 MHz): δ = 0.74 (t, J = 6.9 Hz, 3 H,
CH3CH2CH), 0.93 (d, J = 6.9 Hz, 3 H, CHCH3), 1.20 (t, J =
7.2 Hz, 3 H, CH3CH2O), 1.34 (m, 2 H, CH3CH2CH), 2.25 (s, 3 H,
COCH3), 2.35 (m, 1 H, CH3CH2CH), 4.16 (q, J = 7.2 Hz, 2 H,
CH3CH2O), 6.56 (d, J = 10.8 Hz, 1 H, CH=) ppm. 13C NMR
(CDCl3, 300 MHz): δ = 11.6 (CH3), 13.9 (CH3), 19.2 (CH), 26.3
(CH2), 29.0 (CH3), 35.4 (CH3), 61.0 (CH2), 134.5 (C), 153.0 (CH),
164.1 (C), 201.2 (C) ppm. C11H18O3 (198.26): calcd. C 66.64, H
9.15; found C 66.72, H 9.14.
(Z)-1d: Yield 72%, yellow oil. 1H NMR (CDCl3, 200 MHz): δ =
1.28 (t, J = 7.2 Hz, 3 H, CH3CH2O), 2.43 (s, 3 H, COCH3), 4.34
(q, J = 7.2 Hz, 2 H, CH3CH2O), 7.28–7.48 (m, 5 H, Ph), 7.58 (s, 1
H, CH=) ppm. 13C NMR (CDCl3, 300 MHz): δ = 13.8 (CH3), 26.5
(CH3), 61.7 (CH2), 128.8 (CH), 129.5 (CH), 130.7 (CH), 132.9 (C),
134.6 (C), 141.2 (CH), 167.7 (C), 194.6 (C) ppm.
(E)-1d: 1H NMR (CDCl3, 200 MHz): δ = 1.34 (t, J = 7.2 Hz, 3
H, CH3CH2O), 2.36 (s, 3 H, COCH3), 4.30 (q, J = 7.2 Hz, 2 H,
CH3CH2O), 7.34–7.39 (m, 5 H, Ph), 7.68 (s, 1 H, CH=) ppm. 13C
NMR (CDCl3, 300 MHz): δ = 14.1 (CH3), 31.1 (CH3), 61.5 (CH2),
128.8 (CH), 129.6 (CH), 130.3 (CH), 132.8 (C), 134.0 (C), 140.4,
(CH), 164.3 (C), 203.3 (C) ppm. C13H14O3 (218.25): calcd. C 71.54,
H 6.47; found C 71.59, H 6.48.
1
syn-2c: Yield 75%, pale yellow oil. H NMR (CDCl3, 300 MHz):
δ = 0.07 [s, 9 H, OSi(CH3)3], 0.90–1.22 (m, 5 H, cyclohexyl), 1.27
(t, J = 7.2 Hz, 3 H, OCH2CH3), 1.44–1.80 (m, 6 H, cyclohexyl),
2.25 (s, 3 H, OCCH3), 3.36 (m, 1 H, NCH), 3.44 (d, J = 3.3 Hz, 1
H, CHCO), 4.12 (m, 2 H, OCH2CH3), 6.20 (br. s, 1 H, NH) ppm.
13C NMR (CDCl3, 300 MHz): δ = –1.1 (3CH3), 14.0 (CH3), 26.2
(2CH2), 28.6 (CH), 29.2 (CH2), 30.4 (CH2), 30.9 (CH2), 38.8 (CH3),
59.2 (CH), 60.9 (CH2), 69.0 (CH), 170.2 (C), 203.1 (C) ppm.
anti-2c: 1H NMR (CDCl3, 300 MHz): δ = 0.13 [s, 9 H, OSi-
(CH3)3], 0.90–1.22 (m, 5 H, cyclohexyl), 1.25 (t, J = 7.2 Hz, 3 H,
OCH2CH3), 1.44–1.80 (m, 6 H, cyclohexyl), 2.28 (s, 3 H, OCCH3),
3.36 (m, 1 H, NCH), 3.72 (d, J = 3.3 Hz, 1 H, CHCO), 4.20 (m, 2
H, OCH2CH3), 6.20 (br. s, 1 H, NH) ppm. 13C NMR (CDCl3,
300 MHz): δ = –1.2 (3CH3), 14.1 (CH3), 26.0 (CH2), 26.5 (CH2),
29.3 (CH), 30.3 (CH2), 30.6 (CH2), 31.0 (CH2), 39.2 (CH3), 59.7
(CH), 61.2 (CH2), 67.5 (CH), 169.4 (C), 204.0 (C) ppm.
1
syn-2d: Yield 66%, pale yellow oil. H NMR (CDCl3, 300 MHz):
δ = 0.03 [s, 9 H, OSi(CH3)3], 0.94 (t, J = 7.2 Hz, 3 H, OCH2CH3),
2.36 (s, 3 H, OCCH3), 3.90 (q, J = 7.2 Hz, 2 H, OCH2CH3), 4.15
(d, J = 7.8 Hz, 1 H, CHCO), 4.68 (d, J = 7.8 Hz, 1 H, NCH), 5.59
(br. s, 1 H, NH), 7.20–7.50 (m, 5 H, Ph) ppm. 13C NMR (CDCl3,
300 MHz): δ = –0.8 (3 CH3), 13.6 (CH3), 29.6 (CH3), 61.3 (CH2),
63.0 (CH), 65.8 (CH), 128.0 (CH), 128.4 (CH), 128.8 (CH), 137.6
(C), 167.2 (C), 201.7 (C) ppm.
anti-2d: 1H NMR (CDCl3, 300 MHz): δ = 0.01 [s, 9 H, OSi-
(CH3)3], 1.26 (t, J = 7.2 Hz, 3 H, OCH2CH3), 2.06 (s, 3 H,
OCCH3), 4.05 (d, J = 8.7 Hz, 1 H, CHCO), 4.20 (q, J = 7.2 Hz, 2
H, OCH2CH3), 4.65 (d, J = 8.7 Hz, 1 H, NCH), 5.91 (br. s, 1 H,
NH), 7.20–7.50 (m, 5 H, Ph) ppm. 13C NMR (CDCl3, 300 MHz):
δ = –1.2 (3CH3), 13.9 (CH3), 29.8 (CH3), 61.5 (CH2), 62.4 (CH),
65.8 (CH), 127.8 (CH), 128.2 (CH), 129.5 (CH), 137.6 (C), 168.6
(C), 201.3 (C) ppm.
General Procedure for the 1,4 Addition of N,O-Bis(trimethylsilyl)-
hydroxylamine to Alkylideneacetoacetate 1: To a stirred solution of
1a–d (0.5 mmol) in dry CH2Cl2 (2.5 mL) at 0 °C under nitrogen,
N,O-bis(trimethylsilyl)hydroxylamine (2 equiv., 1 mmol) was added
in one portion. The reaction was monitored by TLC and quenched
after 16 h with H2O. The residue was then diluted with CH2Cl2
(10 mL) and washed with H2O (2ϫ10 mL). The organic layer was
dried with Na2SO4, and the solvent was removed under reduced
pressure. Compounds 2a–d were purified by flash chromatography
with silica gel (cyclohexane/EtOAc, 9:1).
1
syn-2a: Yield 77%, pale yellow oil. H NMR (CDCl3, 300 MHz):
δ = 0.08 [s, 9 H, OSi(CH3)3], 0.95 (d, J = 6.9 Hz, 3 H, CHCH3),
1.03 (d, J = 6.9 Hz, 3 H, CHCH3), 1.28 (t, J = 7.2 Hz, 3 H,
OCH2CH3), 1.68 (m, 1 H, CH3CHCH3), 2.25 (s, 3 H, OCCH3),
3.26 (m, 1 H, NCH), 3.45 (d, J = 3.6 Hz, 1 H, CHCO), 4.17 (m, 2
H, OCH2CH3), 6.22 (d, J = 11.4 Hz, 1 H, NH) ppm. 13C NMR
(CDCl3, 300 MHz): δ = –1.1 (3 CH3), 14.0 (CH3), 20.5 (2 CH3),
29.4 (CH), 30.1 (CH3), 59.7 (CH), 60.9 (CH2), 67.5 (CH), 170.1
(C), 202.9 (C) ppm.
anti-2a: 1H NMR (CDCl3, 300 MHz): δ = 0.11 [s, 9 H, OSi-
(CH3)3], 0.93 (m, 6 H, CHCH3), 1.26 (t, J = 7.2 Hz, 3 H,
OCH2CH3), 1.94 (m, 1 H, CH3CHCH3), 2.29 (s, 3 H, OCCH3),
3.37 (m, 1 H, NCH), 3.71 (d, J = 7.2 Hz, 1 H, CHCO), 4.17 (m, 2
H, OCH2CH3), 5.55 (d, J = 9.6 Hz, 1 H, NH) ppm. 13C NMR
(CDCl3, 300 MHz): δ = –0.6 (3 CH3), 14.0 (CH3), 18.5 (CH3), 21.6
(CH3), 28.6 (CH), 30.1 (CH3), 60.3 (CH), 61.3 (CH2), 69.8 (CH),
168.7 (C), 200.6 (C) ppm.
General Procedure for the Catalyzed 1,4-Addition of N,O-Bis(tri-
methylsilyl)hydroxylamine to Alkylideneacetoacetate 1: To a stirred
solution of 1a–d (0.5 mmol) in dry CH2Cl2 (2.5 mL) under nitro-
gen, a catalytic amount of scandium(III) triflate or copper(II) trifl-
ate (0.025 mmol) was added. The reaction mixture was cooled to
0 °C, and N,O-bis(-trimethylsilyl)hydroxylamine (2 equiv., 1 mmol)
was added in one portion. The mixture was stirred for 1 h at the
same temperature and then filtered through a Celite pad. The sol-
vent was removed under reduced pressure to afford 3a–d.
3a: Yield Ͼ95%, colourless clear oil. 1H NMR (CDCl3, 300 MHz):
δ = 0.17 [s, 9 H, OSi(CH3)3], 0.26 [s, 9 H, OSi(CH3)3], 0.81 (d, J =
6.9 Hz, 3 H, CHCH3), 1.07 (d, J = 6.6 Hz, 3 H, CHCH3), 1.28 (t,
J = 6.9 Hz, 3 H, OCH2CH3), 1.54–1.64 (m, 1 H, CH3CHCH3),
1.92 (s, 3 H, OCCH3), 3.12 (t, J = 10.8 Hz, 1 H, NCH), 4.09–4.20
1
syn-2b: Yield 70%, pale yellow oil. H NMR (CDCl3, 300 MHz):
δ = 0.09 [s, 9 H, OSi(CH3)3], 0.66–1.06 (m, 8 H, CHCH2CH3,
6124
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 6119–6127

Synthesis of Ethyl 5-Hydroxyisoxazolidine-4-carboxylates
(m, 2 H, OCH2CH3), 5.88 (d, J = 11.4 Hz, 1 H, NH) ppm. 13C (CH), 32.6 (CH3), 60.1 (CH2), 61.3 (CH), 67.9 (CH), 106.5 (C),
NMR (CDCl3, 300 MHz): δ = –0.9 (3 CH3), 0.4 (3 CH3), 14.3 170.7 (C) ppm. Minor anomer: 1H NMR (C6D6, 600 MHz): δ =
(CH3), 20.5 (CH3), 21.3 (CH3), 28.9 (CH), 59.7 (CH2), 69.7 (CH), 0.86 (d, J = 7.2 Hz 3 H, CHCH3), 0.88 (d, J = 6.6 Hz, 3 H,
112.8 (C), 157.3 (C), 168.4 (C) ppm.
CHCH3), 0.91 (t, J = 7.2 Hz, 3 H, OCH2CH3), 1.45 (s, 3 H,
HOCCH3), 1.50 (m, 1 H, CH3CHCH3), 3.03 (d, J = 7.8 Hz 1 H,
CHCO), 3.88 (dd, J = 7.8, 7.2 Hz, 1 H, HNCH), 3.92 (m, 2 H,
OCH2CH3) ppm. 13C NMR (CDCl3, 300 MHz): δ = 14.2 (CH3),
19.6 (CH3), 20.4 (CH), 22.2 (CH3), 30.1 (CH3), 61.1 (CH2), 62.0
(CH), 71.9 (CH), 108.4 (C), 171.5 (C) ppm. C10H19NO4 (217.26):
calcd. C 55.28, H 8.81, N 6.45; found C 55.31, H 8.79, N 6.43.
3b: Yield Ͼ95%, colourless clear oil. 1H NMR (CDCl3, 300 MHz):
δ = 0.10 [s, 9 H, OSi(CH3)3], 0.20 [s, 9 H, OSi(CH3)3], 0.79 (d, J =
6.9 Hz, 3 H, CHCH2CH3), 0.82–0.90 (m, 2 H, CHCH2CH3), 1.05
(d, J = 6.6 Hz, 3 H, CHCH3), 1.23–1.30 (m, 3 H, OCH2CH3), 1.35–
1.49 (m, 1 H, CH3CHCH2), 1.92 (s, 3 H, OCCH3), 3.22 (bt, J =
9.3 Hz, 1 H, NCH), 4.07–4.27 (m, 2 H, OCH2CH3), 5.88 (br. s, 1
1
H, NH) ppm. 13C NMR (CDCl3, 300 MHz): δ = –0.8 (3 CH3), 0.5 4b: Yield 90%, pale yellow oil. Major anomer: H NMR (CDCl3,
(3 CH3), 10.5 (CH3), 14.4 (CH3), 16.9 (CH3), 20.5 (CH), 26.6
(CH2), 34.8 (CH3), 59.7 (CH2), 68.2 (CH), 112.8 (C), 156.9 (C),
168.5 (C) ppm.
300 MHz): δ = 0.80–1.06 (m, 8 H, CHCH2CH3, CHCH2CH3,
CHCH3) 1.30 (t, J = 7.2 Hz, 3 H, OCH2CH3), 1.58 (m, 1 H,
CH3CH CH2CH3), 1.67 (s, 3 H, HOCCH3), 2.96 (d, J = 6.0 Hz, 1
H, CHCO), 3.67 (t, J = 6.0 Hz, 1 H, HNCH), 4.23 (m, 2 H,
OCH2CH3) ppm. 13C NMR (CDCl3, 300 MHz): δ = 11.2 (CH3),
14.1 (CH3), 15.2 (CH3), 23.8 (CH), 26.2 (CH2), 39.0 (CH3), 59.8
(CH), 61.3 (CH2), 66.6 (CH), 106.1 (C), 170.7 (C) ppm. Minor
anomer: 1H NMR (CDCl3, 300 MHz): δ = 0.80–1.06 (m, 8 H,
CHCH2CH3, CHCH2CH3, CHCH3) 1.29 (t, J = 7.2 Hz, 3 H,
3c: Yield Ͼ95%, colourless clear oil. 1H NMR (CDCl3, 300 MHz):
δ = 0.14 [s, 9 H, OSi(CH3)3], 0.24 [s, 9 H, OSi(CH3)3], 0.86–1.40
(m, 6 H, cyclohexyl), 1.32 (t, J = 7.2 Hz, 3 H, OCH2CH3), 1.65–
1.80 (m, 4 H, cyclohexyl), 1.96 (s, 3 H, OCCH3), 2.27 (bd, J =
13.2 Hz, 1 H, cyclohexyl), 3.29 (d, J = 10.5 Hz, 1 H, NCH), 4.12–
4.23 (m, 2 H, OCH2CH3), 5.92 (br. s, 1 H, NH) ppm. 13C NMR
(CDCl3, 300 MHz): δ = –0.8 (3 CH3), 0.5 (3 CH3), 14.4 (CH3), 20.6
(CH), 25.9 (CH2), 26.2 (CH2), 26.6 (CH2), 30.7 (CH2), 31.8 (CH2),
38.3 (CH3), 59.7 (CH2), 68.3 (CH), 112.5 (C), 157.3 (C), 168.5 (C)
ppm.
OCH2CH3), 1.53 (s,
3
H, HOCCH3), 1.58 (m,
1
H,
CH3CHCH2CH3), 2.95 (m, 1 H, CHCO), 3.68 (m, 1 H, HNCH),
4.25 (m, 2 H, OCH2CH3) ppm. 13C NMR (CDCl3, 300 MHz): δ =
11.8 (CH3), 13.6 (CH3), 18.2 (CH3), 23.1 (CH), 26.6 (CH2), 40.1
(CH3), 60.3 (CH), 61.9 (CH2), 70.1 (CH), 105.0 (C), 167.2 (C) ppm.
C11H21NO4 (231.29): calcd. C 57.12, H 9.15, N 6.06; found C
57.11, H 9.19, N 6.08.
3d: Yield Ͼ95%, colourless clear oil. 1H NMR (CDCl3, 300 MHz):
δ = 0.21 [s, 9 H, OSi(CH3)3], 0.30 [s, 9 H, OSi(CH3)3], 1.08 (t, J =
7.2 Hz, 3 H, OCH2CH3), 2.05 (s, 3 H, OCCH3), 4.05 (q, J = 6.9 Hz,
2 H, OCH2CH3), 4.89 (d, J = 11.4 Hz, 1 H, HNCH) 6.27 (d, J =
11.4 Hz, 1 H, NH), 7.29–7.35 (m, 5 H, Ph) ppm. 13C NMR
(CDCl3, 300 MHz): δ = –0.9 (3 CH3), 0.5 (3 CH3), 14.0 (CH3), 20.9
(CH3), 59.7 (CH2), 60.9 (CH), 113.13 (C), 126.8 (2 CH), 128.0 (2
CH), 128.6 (CH), 139.9 (C), 153.4 (C), 167.5 (C) ppm.
4c: Yield 95%, yellow oil. Major anomer: 1H NMR (CDCl3,
300 MHz): δ = 0.92–1.40 (m, 6 H, cyclohexyl), 1.29 (t, J = 7.2 Hz,
3 H, OCH2CH3), 1.52–1.78 (m, 4 H, cyclohexyl), 1.67 (s, 3 H,
OCCH3), 2.00 (m, 1 H, cyclohexyl), 3.00 (d, J = 5.7 Hz, 1 H,
CHCO), 3.56 (dd, J = 5.7, 8.1 Hz, 1 H, HNCH), 4.24 (m, 2 H,
OCH2CH3) ppm. 13C NMR (CDCl3, 300 MHz): δ = 14.1 (CH3),
23.9 (CH), 25.8 (CH2), 26.0 (CH2), 26.3 (CH2), 29.4 (CH2), 29.9
(CH2), 42.3 (CH3), 60.1 (CH), 61.4 (CH2), 67.2 (CH), 106.3 (C),
General Procedure for the Cyclisation of 2a–d or 3a–d to 5-Hydroxy-
isoxazolidine-4-carboxylates 4a–d: To a stirred solution of 2a–d or
3a–d (0.5 mmol) in CH2Cl2 (5 mL), wet silica gel or Amberlist H15
(100 mg) was added, and the reaction mixture was stirred at room
temperature for 2 h. The mixture was filtered through a Gouch
filter, the silica gel was washed with methanol (10 mL), and the
filtrate was concentrated under reduced pressure to afford 4a–d.
1
170.8 (C) ppm. Minor anomer: H NMR (CDCl3, 300 MHz): δ =
0.92–1.40 (m, 6 H, cyclohexyl), 1.33 (t, J = 7.2 Hz, 3 H,
OCH2CH3), 1.52–1.78 (m, 4 H, cyclohexyl), 1.64 (s, 3 H, OCCH3),
1.85 (m, 1 H, cyclohexyl), 2.96 (d, J = 7.5 Hz, 1 H, CHCO), 3.56
(t, J = 7.5 Hz, 1 H, HNCH), 4.30 (m, 2 H, OCH2CH3) ppm. 13C
NMR (CDCl3, 300 MHz): δ = 14.1 (CH3), 22.0 (CH), 25.1 (CH2),
25.6 (CH2), 26.9 (CH2), 30.9 (CH2), 31.0 (CH2), 40.0 (CH3), 61.0
(CH), 62.1 (CH2), 70.2 (CH), 108.2 (C), 171.4 (C) ppm. C13H23NO4
(257.33): calcd. C 60.68, H 9.01, N 5.44; found C 60.75, H 9.00, N
5.41.
4a (Methyl Ester): Yield Ͼ95%, colourless clear oil. Major anomer:
1H NMR (CDCl3, 300 MHz): δ = 0.88 (d, J = 6.6 Hz, 3 H,
CHCH3), 0.97 (d, J = 6.6 Hz, 3 H, CHCH3), 1.67 (s, 3 H,
HOCCH3), 1.69 (m, 1 H, CH3CHCH3), 2.96 (d, J = 5.4 Hz, 1 H,
CHCO), 3.56 (dd, J = 5.4, 7.5 Hz, 1 H, HNCH), 3.76 (s, 3 H,
OCH3) ppm. 13C NMR (CDCl3, 300 MHz): δ = 18.7 (CH3), 19.2
(CH3), 23.8 (CH), 32.5 (CH3), 52.4 (CH), 60.2 (CH3), 67.9 (CH),
106.5 (C), 171.0 (C) ppm. Minor anomer: 1H NMR (CDCl3,
300 MHz): δ = 0.92 (d, J = 6.9 Hz 3 H, CHCH3), 1.00 (d, J =
6.6 Hz, 3 H, CHCH3), 1.50 (s, 3 H, HOCCH3), 1.70 (m, 1 H,
CH3CHCH3), 2.95 (d, J = 7.6 Hz 1 H, CHCO), 3.58 (m, 1 H,
HNCH), 3.74 (s, 3 H, OCH3) ppm. 13C NMR (CDCl3, 300 MHz):
δ = 19.6 (CH3), 20.3 (CH3), 22.2 (CH), 30.1 (CH3), 52.2 (CH), 62.0
(CH3), 71.3 (CH), 108.4 (C), 172.0 (C) ppm. C9H17NO4 (203.24):
calcd. C 53.19, H 8.43, N 6.89; found C 53.14, H 8.45, N 6.92.
4d: Yield 80%, yellow oil; Major anomer: 1H NMR (CDCl3,
300 MHz): δ = 1.31 (t, J = 7.2 Hz, 3 H, OCH2CH3), 1.74 (s, 3 H,
OCCH3), 3.24 (d, J = 5.4 Hz, 1 H, CHCO), 4.27 (q, J = 7.2 Hz, 2
H, OCH2CH3), 5.09 (d, J = 5.4 Hz, 1 H, HNCH), 7.28–7.57 (m, 5
H, Ph) ppm. 13C NMR (CDCl3, 300 MHz): δ = 14.4 (CH3), 24.4
(CH3), 61.9 (CH2), 65.0 (CH), 70.2 (CH), 107.0 (C), 126.3 (2 CH),
127.3 (2 CH), 128.7 (CH), 134.1 (C), 178.0 (C) ppm. Minor an-
1
omer: H NMR (CDCl3, 300 MHz): δ = 1.27 (t, J = 7.2 Hz, 3 H,
OCH2CH3), 1.64 (s, 3 H, OCCH3), 3.41 (d, J = 7.2 Hz, 1 H,
CHCO), 4.15 (q, J = 7.2 Hz, 2 H, OCH2CH3), 4.94 (d, J = 7.2 Hz,
1 H, HNCH), 7.28–7.60 (m, 5 H, Ph) ppm. 13C NMR (CDCl3,
300 MHz): δ = 14.1 (CH3), 25.1 (CH3), 60.9 (CH2), 64.2 (CH), 69.3
(CH), 105.8 (C), 126.5 (2 CH), 127.1 (2 CH), 128.9 (CH), 136.1
(C), 169.4 (C) ppm. C13H17NO4 (251.28): calcd. C 62.14, H 6.82,
N 5.57; found C 62.19, H 6.80, N 5.53.
1
4a (Ethyl Ester): Yield 95%, yellow oil. Major anomer: H NMR
(C6D6, 600 MHz): δ = 0.82 (d, J = 6.6 Hz, 3 H, CHCH3), 0.93 (t,
J = 7.2 Hz, 3 H, OCH2CH3), 0.97 (d, J = 6.6 Hz, 3 H, CHCH3),
1.51 (s, 3 H, HOCCH3), 1.64 (m, 1 H, CH3CHCH3), 2.83 (d, J =
5.4 Hz, 1 H, CHCO), 3.58 (dd, J = 5.4, 7.2 Hz, 1 H, HNCH), 3.93
(m, 1 H, OCH2CH3), 3.98 (m, 1 H, OCH2CH3) ppm. 13C NMR
General Procedure for the Conversion of 4a–c to 5a–c: To a stirred
(CDCl3, 300 MHz): δ = 14.1 (CH3), 18.7 (CH3), 19.4 (CH3), 23.8 solution of 4a–c (0.5 mmol) in CH2Cl2 (2.5 mL) at 0 °C, Et3N
Eur. J. Org. Chem. 2008, 6119–6127
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6125

G. Cardillo, M. Stenta, et al.
FULL PAPER
(1.1 equiv., 0.55 mmol) and 3,5-dinitrobenzoyl chloride (1.1 equiv.,
0.55 mmol) were added. The solution was stirred for 1 h, and then
the mixture was quenched with H2O. The residue was diluted with
CH2Cl2 (5 mL), and the biphasic mixture was washed with H2O
(2ϫ5 mL). The organic layer was dried with Na2SO4, and the sol-
vent was removed under reduced pressure. Compounds 5a–c were
purified by ash chromatography with silica gel (cyclohexane/ethyl
acetate, 85:15).
B3LYP DFT functional (referred to in TURBOMOLE as
“b3 lyp_Gaussian”[26]) with the “m3” grid size for the density fit-
ting and an SCF convergence criterion of 1ϫ10–7°Eh. A balanced
double-ζ split valence (SP) basis set with polarisation functions (P)
[referred to in TURBOMOLE as def2-SV(P) basis][27] was adopted
to describe the H, C, N, and O atoms, whereas for the Cu atom, a
triple-ζ split valence (SP) basis set with polarisation functions (P)
(def2 TZVP basis in TURBOMOLE) was employed.[22] The proper
use of the RI and MARI approximations required the introduction
of an auxiliary basis set.[28] The nature of the various critical points
located on the PES was ascertained by numerical frequency calcu-
lations on the optimised structures. The standard GAUSSIAN 03
convergence criteria were adopted. All geometry optimizations
were carried out in the presence of solvent effects, which were
evaluated with the COSMO[29] solvent continuous model approach
as implemented in the TURBOMOLE package.[14] The dielectric
constant for CH2Cl2 (ε = 8.930) and a solvent radius of r = 2.27 Å
were used. In both cases an “optimised” radius[24] was assigned to
the H, C, N, and O atoms, whereas a “bondii” radius[24] was used
for Si and Cu, according to the TURBOMOLE manual.[14]
5a: Yield 80%, white solid; m.p. 148–152 °C. 1H NMR (CDCl3,
300 MHz): δ = 1.07 (d, J = 6.6 Hz, 6 H, CH3CHCH3), 1.34 (t, J =
7.2 Hz, 3 H, OCH2CH3), 1.65 (s, 3 H, OCCH3), 2.05–2.16 (m, 1
H, CH3CHCH3) 3.16 (d, J = 6.9 Hz, 1 H, HCCO2), 4.29 (q, J =
7.2 Hz, 2 H, OCH2CH3), 4.94 (t, J = 6.9 Hz, 1 H, NCH), 8.97 (s,
2 H, Ph), 9.14 (s, 1 H, Ph) ppm. 13C NMR (CDCl3, 300 MHz): δ
= 14.0 (CH3), 18.2 (CH3), 18.7 (CH3), 23.0 (CH3), 32.4 (CH), 57.7
(CH), 62.2 (CH2), 65.9 (CH) 106.5 (C), 120.5 (CH), 129.6 (2 CH),
137.6 (C), 148.0 (2 C), 151.0 (C), 169.9 (C) ppm. C17H21NO4
(303.36): calcd. C 49.64, H 5.15, N 10.21; found C 49.55, H 5.13,
N 10.26.
5b: Yield 70%, pale yellow solid; m.p. 120–122 °C. 1H NMR
(CDCl3, 300 MHz): δ = 0.95–1.05 (m, 6 H, CH3CHCH2CH3), 1.31
X-ray Crystallography: The diffraction experiments for 5a were car-
ried out at room temperature with a Bruker AXS Apex II CCD
diffractometer with graphite-monochromated Mo-Kα radiation (λ
= 0.71073 Å). Intensity data were measured over full diffraction
spheres with 0.3°-wide ω-scans. The SMART[30] software package
was used for collecting frames of data, indexing reflections, and the
determination of lattice parameters. The collected frames were then
processed for integration with SAINT,[27] and an empirical absorp-
tion correction was applied with SADABS.[31] The structures were
solved by direct methods (SIR 97)[32] and subsequent Fourier syn-
theses and refined by full-matrix least-squares calculations on F2
(SHELXTL)[33] with anisotropic thermal parameters for the non-H
atoms. The methyl and aromatic H atoms were placed in calculated
positions and refined with idealized geometry, whereas the other H
atoms were located in the Fourier map and refined isotropically.
CCDC-694159 for 5a contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif.
(t,
J = 7.2 Hz, 3 H, OCH2CH3), 1.37–1.43 (m, 2 H,
CH3CHCH2CH3), 1.51–1.60 (m, 1 H, CH3CHCH2CH3), 1.63 (s, 3
H, OCCH3), 3.15 (d, J = 6.9 Hz, 1 H, HCCO2), 4.26 (q, J = 7.2 Hz,
2 H, OCH2CH3), 5.03 (q, J = 6.3 Hz, 1 H, NCH), 5.84 (br. s, OH),
8.93 (s, 2 H, Ph), 9.19 (s, 1 H, Ph) ppm. 13C NMR (CDCl3,
300 MHz): δ = 12.9 (CH3), 14.0 (CH3), 15.2 (CH3), 22.7 (CH3),
25.8 (CH2), 39.0 (CH), 58.3 (CH), 62.1 (CH2), 64.6 (CH) 106.5 (C),
120.4 (CH), 129.6 (2 CH), 137.4 (C), 147.9 (2 C), 158.6 (C), 169.4
(C) ppm. C18H23NO4 (317.38): calcd. C 50.82, H 5.45, N 9.88;
found C 50.85, H 5.42, N 9.90.
5c: Yield 68%, white solid; m.p. 145–148 °C. 1H NMR (CDCl3,
300 MHz): δ = 1.02–1.31 (m, 6 H, cyclohexyl), 1.29 (t, J = 7.2 Hz,
3 H, OCH2CH3), 1.59 (s, 3 H, OCCH3), 1.62–1.85 (m, 5 H, cyclo-
hexyl), 2.14 (d, J = 6.9 Hz, 1 H, HCCO2), 4.24 (q, J = 6.9 Hz, 2
H, OCH2CH3), 4.88 (t, J = 6.3 Hz, 1 H, NCH), 8.92 (s, 2 H, Ph),
9.09 (s, 1 H, Ph) ppm. 13C NMR (CDCl3, 300 MHz): δ = 14.1
(CH3), 23.0 (CH3), 25.8 (CH2), 26.1 (CH2), 26.9 (CH2), 28.8 (CH2),
29.3 (CH2), 42.2 (CH), 58.0 (CH), 62.3 (CH2), 65.2 (CH) 106.5 (C),
120.5 (CH), 129.6 (2 CH), 137.6 (C), 148.0 (3 C), 169.9 (C) ppm.
C20H25NO4 (343.42): calcd. C 53.21, H 5.58, N 9.31; found C
53.09, H 5.59, N 9.28.
Supporting Information (see footnote on the first page of this arti-
cle): Colour representation of Figures 1, 4 and 5 of the article; com-
putational details, including Wiberg bond order for the molecules
IntF and IntC, total energy (E), Gibbs energy (G) and optimised
geometries of the molecules discussed in the paper.
Computational Details: All calculations were carried out with the
COBRAMM[18] suite of programs, which was used as an interface
between the TURBOMOLE[19] and the GAUSSIAN 03 pack-
ages.[20] This approach allowed us to combine the accurate optimis-
ation algorithms implemented in GAUSSIAN 03 and the TUR-
BOMOLE approximated DFT[21] potential. In particular, the use
of the “resolution of identity” (RI)[22] and the “multipole ac-
celerated resolution of identity” (MARI)[23] approximations, avail-
able in TURBOMOLE for the most common DFT functionals,
has proven to speed up the calculation (with a negligible loss of
accuracy) by about one order of magnitude relative to non-approxi-
mated DFT approaches. In particular, we chose the GAUSSIAN 03
optimisation driver in place of that available in the TURBOMOLE
package because of its faster convergence in locating minima (equi-
librium structures) and saddle-points (transition states) on the po-
tential energy surface (PES). The efficiency of this optimization
algorithm is due to the employment of a redundant internal coordi-
nates system[24] with the powerful BFGS algorithm.[25] The details
of the COBRAMM interface have been extensively described in
our previous paper.[13] In all computations, we have used the
Acknowledgments
We thank the Ministero dell’Università e della Ricerca (MIUR)
(PRIN 2006 prot. n. 2006030449_003), MAE (Italia-Messico) and
University of Bologna (Strategic project ID 450) for financial sup-
port.
[1] a) P. Perlmutter, Conjugate Addition Reaction in Organic Syn-
thesis, Pergamon Press, Oxford, 1992; b) M. P. Sibi, S. Man-
yem, Tetrahedron 2000, 56, 8033–8061; c) J. Leonard, E. Diez-
Barra, S. Merino, Eur. J. Org. Chem. 1998, 2051–2061; d) S. G.
Davies, O. J. Ichihara, Synth. Org. Chem. Jpn. 1997, 55, 26–34.
[2] a) X. L. Jin, H. Sugihara, K. Daikai, H. Tateishi, Y. Z. Jin, H.
Furuno, J. Inanaga, Tetrahedron 2002, 58, 8321–8329; b) D. J.
Guerin, S. J. Miller, J. Am. Chem. Soc. 2002, 124, 2134–2136;
c) J. K. Myers, E. N. Jacobsen, J. Am. Chem. Soc. 1999, 121,
8959–8960; d) N. Yamagiwa, H. Qin, S. Matsunaga, M. Shiba-
saki, J. Am. Chem. Soc. 2005, 127, 13419–13427; e) M. P. Sibi,
6126
www.eurjoc.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2008, 6119–6127

Synthesis of Ethyl 5-Hydroxyisoxazolidine-4-carboxylates
N. Prabagaran, S. G. Ghorpade, C. P. Jasperse, J. Am. Chem.
Soc. 2003, 125, 11796–11797; f) M. P. Sibi, J. J. Shay, M. Liu,
C. P. Jasperse, J. Am. Chem. Soc. 1998, 120, 6615–6616; g) W.
Zhuang, R. G. Hazell, K. A. Jorgensen, Chem. Commun. 2001,
1240–1241; h) C. Palomo, M. Oiarbide, R. Halder, M. Kelso,
E. Gomez-Bengoa, J. M. Garcia, J. Am. Chem. Soc. 2004, 126,
9188–9189.
[14]
[15]
D. Niu, K. Zhao, J. Am. Chem. Soc. 1999, 121, 2456–2459.
Experiments involving the direct formation of 4 from alkylid-
eneacetoacetates and TMSONHTMS without the isolation of
2 or 3 resulted in more complex mixtures and afforded 4 in
lower yield.
Detailed pictures and computational data for the eight isomers
are reported in the Supporting Information.
F. Benfatti, G. Cardillo, L. Gentilucci, E. Mosconi, A. Tolo-
melli, Tetrahedron: Asymmetry 2007, 18, 2227–2232.
P. Altoè, M. Stenta, A. Bottoni, M. Garavelli, Theor. Chem.
Acc. 2007, 118, 219–240.
[16]
[17]
[18]
[19]
[20]
[3] a) P. I. Dalko, L. Moisan, Angew. Chem. Int. Ed. 2004, 43,
5138–5175; b) Y. K. Chen, M. Yoshida, D. W. C. MacMillan,
J. Am. Chem. Soc. 2006, 128, 9328–9329; c) M. S. Taylor, E. N.
Jacobsen, Angew. Chem. Int. Ed. 2006, 45, 1520–1543; d) M. J.
Gaunt, C. C. C. Johansson, A. McNally, N. T. Vo, Drug Disc.
Tod. 2007, 12, 8–27; e) C. L. Cao, X. L. Sun, J. L. Zhou, Y.
Tang, J. Org. Chem. 2007, 72, 4073–4076; f) J. Vesely, I. Ib-
rahem, G.-L. Zhao, R. Rios, A. Córdova, Angew. Chem. Int.
Ed. 2007, 46, 778–781; g) C. Palomo, A. Mielgo, Angew. Chem.
Int. Ed. 2006, 45, 7876–7880.
[4] a) F. Berthiol, R. Matsubara, N. Kawai, S. Kobayashi, Angew.
Chem. Int. Ed. 2007, 46, 7803–7805; b) J. Zhou, Y. Tang, Chem.
Commun. 2004, 432–433; c) J. Zhou, Y. Tang, J. Am. Chem.
Soc. 2002, 124, 9030–9031; d) V. Annamalai, E. F. DiMauro,
P. J. Carroll, M. C. Kozlowski, J. Org. Chem. 2003, 68, 1973–
1981; e) R. Rasappan, M. Hager, A. Gissibl, O. Reiser, Org.
Lett. 2006, 8, 6099–6102; f) Yasuda, K. M. Shindo, K. Koga,
Tetrahedron Lett. 1996, 37, 6343–6346; g) J. M. Betancort, K.
Sakthivel, R. Thayumanavan, C. F. Barbas, Tetrahedron Lett.
2001, 42, 4441–4444; h) A. Pelter, J. M. Rao, Tetrahedron Lett.
1981, 22, 797–800; i) C. Simon, T. Constantieux, J. Rodriguez,
Eur. J. Org. Chem. 2004, 4957–4980.
[5] G. Cardillo, L. Gentilucci, M. Gianotti, H. Kim, R. Perciac-
cante, A. Tolomelli, Tetrahedron: Asymmetry 2001, 12, 2395–
2398.
[6] a) United States Patent 5122610; b) United States Patent
7026282.
[7] a) U. Chiacchio, L. Borrello, D. Iannazzo, P. Merino, A. Pip-
erno, A. Rescifina, B. Richichi, G. Romeo, Tetrahedron: Asym-
metry 2003, 14, 2419–2425; b) P. Merino, S. Franco, F. L. Mer-
chan, T. Tejero, J. Org. Chem. 2000, 65, 5575–5589; c) G. V. M.
Sharma, I. S. Reddy, V. G. Reddy, A. V. R. Rao, Tetrahedron:
Asymmetry 1999, 10, 229–235.
[8] a) M. P. Sadashiva, H. Mallesha, N. A. Hitesh, K. S. Ran-
gappa, Bioorg. Med. Chem. 2004, 12, 6389–6395; b) J. Kaffy,
R. Pontikis, D. Carrez, A. Croisy, C. Monneret, J.-C. Florent,
Bioorg. Med. Chem. 2006, 14, 4067–4077; c) G. A. Bennett,
G. B. Mullen, T. R. Decory, D. M. Maryniak, W. E. Jones, J. T.
Mitchell, S. D. Allen, V. S. Georgiev, Arch. Pharm. 1989, 322,
531–534; d) R. Kodagahally, H. M. K. S. R. Ravi Kumar, Arch.
Pharm. 2003, 336, 159–164; e) M. I. Torres-Sanchez, P. Borra-
chero, F. Cabrera-Escribano, M. Gomez-Guillen, M. Angulo-
Alvarez, M. J. Dianez, M. D. Estrada, A. Lopez-Castro, S.
Perez-Garrido, Tetrahedron: Asymmetry 2005, 16, 3897–3907;
f) M. J. Green, R. L. Tiberi, R. Friary, B. N. Lutsky, J. Berken-
koph, X. Fernandez, M. Monahan, J. Med. Chem. 1982, 25,
1492–1495; g) R. Plate, P. H. H. Hermkens, H. Behm, H. C. J.
Ottenheijm, J. Org. Chem. 1987, 52, 560–564; h) SmithKline_
Beecham, WO 02/30929 ed., 2002.
[9] I. Ibrahem, R. Rios, J. Vesely, G.-L. Zhao, A. Cordova, Chem.
Commun. 2007, 849–851.
[10] G. Cardillo, S. Fabbroni, L. Gentilucci, M. Gianotti, A. Tolo-
melli, Synt. Commun. 2003, 33, 1587–1594.
[11] F. Benfatti, G. Cardillo, L. Gentilucci, E. Mosconi, A. Tolo-
melli, Synlett, accepted for publication.
[12] a) I. N. Rao, E. N. Prabhakaran, S. K. I. J. Das, J. Org. Chem.
2003, 68, 4079–4082; b) R. Ghosh, S. Maiti, S. Ghosh, A. Mu-
kherjee, Synthesis 2007, 190–196.
R. Ahlrichs, M. Bar, M. Haser, H. Horn, C. Kolmel, Chem.
Phys. Lett. 1989, 162, 165–169.
M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, J. J. A. Montgomery, 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, V. Bakken, C. Adamo, J. Jaramillo, R. Gom-
perts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A.
Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dap-
prich, 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. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayak-
kara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
M. W. Wong, C. Gonzalez, J. A. Pople, GAUSSIAN 03, Re-
vision C.02, Gaussian, Inc., Wallingford, CT, 2004.
P. Geerlings, F. De Proft, W. Langenaeker, Chem. Rev. 2003,
103, 1793–1873.
a) J. L. Whitten, J. Chem. Phys. 1973, 58, 4496–4501; b) B. I.
Dunlap, J. W. D. Connolly, J. R. Sabin, J. Chem. Phys. 1979,
71, 3396–3402.
M. Sierka, A. Hogekamp, R. Ahlrichs, J. Chem. Phys. 2003,
118, 9136–9148.
C. Peng, P. Y. S. Ayala, H. Bernhard, M. J. Frisch, J. Comput.
Chem. 1996, 17, 49–56.
a) C. G. Broyden, J. Inst. Math. Applic. 1970, 6, 76–90; b) R.
Fletcher, Comput. J. 1970, 13, 317–322; c) D. Goldfarb, Math.
Comput. 1970, 24, 23–26; d) D. F. Shanno, Math. Comput.
1970, 24, 647–656.
O. Treutler, R. Ahlrichs, J. Chem. Phys. 1995, 102, 346–354.
F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7,
3297–3305.
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
F. Weigend, Phys. Chem. Chem. Phys. 2006, 8, 1057–1065.
a) A. Klamt, J. Phys. Chem. 1995, 99, 2224–2235; b) A. Jan,
K. Christoph, K. Andreas, J. Chem. Phys. 1995, 103, 9312–
9320; c) A. Schafer, A. Klamt, D. Sattel, J. C. W. Lohrenz, F.
Eckert, Phys. Chem. Chem. Phys. 2000, 2, 2187–2193.
SMART & SAINT Software Reference Manuals, version 5.051
(Windows NT Version), Bruker Analytical X-ray Instruments
Inc., Madison, WI, 1998.
G. M. Sheldrick, SADABS, program for empirical absorption
correction, University of Göttingen, Germany, 1996.
A. Altomare, M. C. Burla, M. Camalli, G. L. Cascarano, C.
Giacovazzo, A. Guagliardi, A. G. G. Moliterni, G. Polidori, R.
Spagna, J. Appl. Crystallogr. 1999, 32, 115–119.
[30]
[31]
[32]
[33]
G. M. Sheldrick, SHELXTLplus (Windows NT Version) Struc-
ture Determination Package, Version 5.1, Bruker Analytical X-
ray Instruments Inc., Madison, WI, USA, 1998.
[13] The characterization of the hydroxyisoxazolidine methyl ester
isomer of 4a is reported in the Experimental Section.
Received: July 18, 2008
Published Online: November 12, 2008
Eur. J. Org. Chem. 2008, 6119–6127
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
6127