High ?-Facial Selectivity Through Chelation of Magnesium Ions in the DMD Epoxidation of α,β-Unsaturated Imides with Chiral Pyrrolidinone Auxiliaries

Article abstract of DOI:10.1002/ejoc.200300544

High diastereoselectivity, nut of the opposite sense, is observed in the epoxidation (DMD or mCPBA) of α,β-unsaturated imides equipped with pyrrolidinone-type chiral auxiliaries that bear either a hydroxymethyl or trityloxymethyl side chain. This unprecedented reversed ?-facial differentiation is promoted by chelation of a magnesium ion, which results in conformational control over the essential steric interactions.

Full text of DOI:10.1002/ejoc.200300544

FULL PAPER
High π-Facial Selectivity Through Chelation of Magnesium Ions in the DMD
Epoxidation of α,β-Unsaturated Imides with Chiral Pyrrolidinone Auxiliaries
Waldemar Adam[a] and Aimin Zhang*[a]
Keywords: Chiral auxiliaries / Dioxirane / Epoxidation / Mechanism / Metal chelation
High diastereoselectivity, but of the opposite sense, is ob-
served in the epoxidation (DMD or mCPBA) of α,β-unsatur-
ated imides equipped with pyrrolidinone-type chiral auxiliar-
ies that bear either a hydroxymethyl or trityloxymethyl side
promoted by chelation of a magnesium ion, which results in
conformational control over the essential steric interactions.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
chain. This unprecedented reversed π-facial differentiation is Germany, 2004)
Introduction
Optically active epoxides are key building blocks in or-
ganic synthesis,[1] and are usually prepared through the oxi-
dation of the corresponding olefins.[2] Stereoselectivity is
achieved through substrate manipulation,[2a] or the use of
chiral oxidants,[2b] in which steric,[2a] electronic,[2bϪ2d]
stereoelectronic,[2eϪ2f] and conformational[3] effects play a
decisive role. An alternative approach utilizes chiral auxili-
aries to effect the π-facial selectivity in the oxygen-transfer
process. A recent application of this methodology is the
highly diastereoselective epoxidation of enecarbamates[4]
that are equipped with oxazolidinone-type chiral auxiliaries.
Through the synergistic interplay of conformational and
steric effects, the attack by DMD and mCPBA may be di-
rected at will to the desired π face of the substrate. In the
present study, we show that metal chelation of the chiral
auxiliary and olefinic functionality can result in effective
diastereoselective epoxidation.
Scheme 1. Chelation of a magnesium ion to the imide 3b
one π face is severely obstructed by the trityloxymethyl sub-
stituent and, consequently, the reagent attacks from the
other face. So far, this concept of diastereoselective control
has not been employed in epoxidations; thus, we became
interested in applying it to the oxidation of the chiral
pyrrolidinone imides 3. Indeed, we demonstrate herein that
metal chelation is a productive method to induce stereo-
selective oxygen transfer in the epoxidation of the olefinic
substrates 3 with mCPBA and DMD.
Such advantageous metal chelation has been reported in
the asymmetric conjugate addition[5a] and DielsϪAlder re-
action[5b] of the triphenylmethyl ether 3b derived from the
chiral pyrrolidinone imide 3a. With Mg(ClO4)2 as the chel-
ating agent, it was shown through NMR spectroscopy stud-
ies[5c] that the preferred 3b(Mg) conformer results from co-
ordination of a magnesium ion to the two carbonyl func-
tionalities (Scheme 1).
For the best conjugation of the CϭC and the CϭO
groups, the s-cis conformation is favored over the s-trans
form by ca. 5 kcal/mol as a result of steric interactions be-
tween the olefinic methylene unit and the lactam carbonyl
group.[5c,5d] Inspection of molecular models discloses that
Results and Discussion
To study this stereochemical concept in the epoxidations
mediated by m-chloroperbenzoic acid (mCPBA) and di-
methyldioxirane (DMD), the imides 3aϪc were prepared
from the commercially available (S)-5-hydroxymethyl-2-pyr-
rolidinone (Scheme 2). Protection of the hydroxy group as
its silyl ether 1, followed by acylation[6] and subsequent de-
protection (TBAF, THF), provided the imide 3a in an over-
all yield of 91%. The cinnamoyl and crotonoyl derivatives
of Koga’s chiral auxiliaries 3b and 3c were synthesized as re-
ported.[7]
[a]
Institut für Organische Chemie, Universität Würzburg,
Am Hubland, 97074 Würzburg, Germany
Fax: (internat.) ϩ 1-787-8070389
E-mail: wadam@chemie.uni-wuerzburg.de
Internet: http://www.organik.chemie.uni-wuerzburg.de
Eur. J. Org. Chem. 2004, 147Ϫ152
DOI: 10.1002/ejoc.200300544
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
147

W. Adam, A. Zhang
FULL PAPER
conversions are impressive because electron-poor olefins
usually exhibit rather low reactivity toward electrophilic
oxidants.[8] In contrast to the results obtained when using
mCPBA (Entries 1 and 2), the π-facial selectivity for the
DMD epoxidations (Entries 3 and 4) is significantly lower;
in fact, for the hydroxy imide 3a, the oxygen transfer takes
place unselectively. Nevertheless, with the triphenylmethyl
derivatives 3b and 3c, the same major diastereoisomers, na-
mely the (2S,3R)-epoxides 4b and 4c, were produced in the
epoxidations with both mCPBA (Entry 2) and DMD (En-
try 4). The higher dr values for mCPBA versus DMD may
be reconciled in terms of hydrogen bonding between the
carbonyl group of the imide and the proton of the per-
acid.[5]
Scheme 2. Preparation of the imides 3aϪc. a. tBuMe2SiCl, imidaz-
ole, DMAP, 97%; b. i. nBuLi, THF, (E)-PhCHCHCOCl, Ϫ78 °C;
ii. TBAF, THF, 94%; c. Ph3CCl, Et3N, DMAP, CH2Cl2, 73%; d. i.
nBuLi, THF, Ϫ78 °C. ii. (E)-PhCHϭCHCOCl or (E)-CH3CHϭ
CHCOCl
Chelation effects[9] were assessed by the addition of mag-
nesium perchlorate during the DMD oxidation of the tri-
phenylmethyl ether imide 3b (Entries 5Ϫ9). Maximum dia-
stereoselectivity (Entry 7) was obtained when 0.5 equiv. of
magnesium perchlorate was employed. Larger amounts of
magnesium perchlorate (Entries 8 and 9) did not improve
the dr values; instead, at high salt loading the yield of the
epoxide 4b was substantially lowered. We attempted aque-
ous workup, but, unfortunately, the epoxides are too labile
and deteriorated on workup, especially when purification
was attempted by chromatography on silica gel. Several
other purification methods were also tested, e.g., Al2O3
chromatography and HPLC, but the epoxides were too sen-
sitive to survive.
The results of the epoxidations are summarized in
Table 1. With mCPBA (1.5 equiv.), the epoxidation was
conducted at 20 °C for 8 h, and the corresponding isomeric
epoxides 4 were obtained in good diastereoselectivities (En-
tries 1 and 2). The low yield of epoxide in Entry 2 is due
presumably to hydrolysis of the trityl functionality in ether
3b on account of the acidic conditions experienced during
the epoxidation with mCPBA.
It is mechanistically significant to note the reversed sense
in the π-facial selectivity between the imides 3a and 3b; that
is, for 3a, which features a free hydroxy group, the major
epoxide is the diastereoisomer (2R,3S)-4a, but for the tri-
phenylmethyl ether derivative 3b it is (2S,3R)-4b. As ex-
pected, the mCPBA epoxidation of the triphenylmethyl
ethers derived from cinnamoyl 3b (Entry 2) and crotonyl 3c
We note that the major epoxide diastereoisomer again is
(data not shown) gave the same (2S,3R)-configured epox- the (2S,3R)-configured one (Entries 5Ϫ9), the same one
ides 4b and 4c with similar dr values. that formed in the absence of the chelating agent
The DMD epoxidation of the imides (S)-3 furnished the Mg(ClO4)2 (Entry 4). The fact that only 0.5 equiv. of
same epoxides 4 in high yields (Entries 3 and 4). The high Mg(ClO4)2 is necessary for optimal diastereoselectivity sug-
Table 1. Data for the products of the diastereoselective epoxidation of α,β-unsaturated imides 3
Entry
Substrate
R1
Oxidant
Additive (equiv.)
Conversion (%)[c]
m.b.(%)[c][d]
Yield (%)[c][e]
dr[c]
Config.[f]
1
2
3
4
5
6
7
8
9
3a
3b
3a
3b
3b
3b
3b
3b
3b
3a
H
Ph3C
H
Ph3C
Ph3C
Ph3C
Ph3C
Ph3C
Ph3C
H
mCPBA
mCPBA
DMD
DMD
DMD
DMD
DMD
DMD
DMD
DMD
Ϫ
Ϫ
Ϫ
Ϫ
91
23
95
96
95
95
95
92
83
85
86
89
95
95
96
92
90
85
80
69
86 (78)
52 (12)[g]
95 (90)
95 (91)
96 (91)
92 (87)
89 (85)
84 (77)
76 (63)
63 (54)
13:87
84:16
48:52
73:27
72:28
87:13
91:09
90:10
91:09
14:86
(2R,3S)
(2S,3R)
(2R,3S)
(2S,3R)
(2S,3R)
(2S,3R)
(2S,3R)
(2S,3R)
(2S,3R)
(2R,3S)
Mg(ClO4)2 (0.1)
Mg(ClO4)2 (0.3)
Mg(ClO4)2 (0.5)
Mg(ClO4)2 (1.0)
Mg(ClO4)2 (3.0)
Mg(ClO4)2 (0.5)
10
[a]
[b]
[c]
In acetone.
In dichloromethane. Determined by 1H NMR spectroscopic analysis directly on the crude reaction mixture, using
dimethyl isophthalate as an internal standard; error Ϯ 5% of the stated value. [d] m.b. ϭ material balance. The material balance represents
the sum of isolated products and recovered starting material. Yield is based on 100% conversion; in parenthesis is given the yield of
isolated epoxide. Configuration of major product.
[e]
[f]
[g]
The low yield of epoxide is presumably due to hydrolysis of the trityl ether on
account of the acidic conditions experienced during the oxidation with mCPBA.
148
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 147Ϫ152

High π-Facial Selectivity Through Chelation of Magnesium Ions
FULL PAPER
gests that the stoichiometry of chelation is two molecules tion,[12] its configuration must be (3R) to comply with the
of the imide 3b substrate per Mg(ClO4)2 chelating agent.[5] observed (3S) assignment in the diol 5. Consequently, the
The epoxidation of the tert-butyldimethylsilyl (TBDMS)- configuration of the C2 position in the epoxide 4b must be
protected ether was also examined in the presence of the (2S), that is, the same as in the diol 5, since this stereogenic
Mg2ϩ salt. We observed a similar trend, but the trityloxy- center is not involved in the process of ring-opening of the
methyl ether is more effective in controlling the diastereosel- epoxide. Thus, by means of this chemical correlation, we
ectivity. For this reason, all subsequent epoxidations were assign the absolute (2S,3R) configuration to epoxide 4b and
conducted using only the trityl ether 3b.
infer that the remaining epoxides 4 have the same structure.
Additionally, the (2S,3R)-4b configuration was confirmed
Tomioka[5c] reported a detailed study, by means of theor-
etical work and NMR spectroscopy, on the preferred con- by 2D-NOESY analysis (Figure 1).
1
formation of imide 3 upon metal ion chelation. H NMR
spectra disclosed that under such conditions the olefinic
methine proton H5Ј is shifted to low field in the presence
of magnesium perchlorate; thus, an NOE effect operates be-
tween protons H2 and H5Ј. This observation was explained
in terms of a shielding effect by the trityloxymethyl group,
which becomes positioned favorably for such interactions
upon the chelation of the magnesium ion between the two
Figure 1. 2D-NOE effects observed for the diastereoisomeric epox-
carbonyl functionalities.
ides (2S,3R)-4b and (2R,3S)-4b
In a manner similar to that observed for substrate 3b, we
found that the diastereoselectivity is significantly increased
in the DMD oxidation of imides 3a and 3c when 0.5 equiv.
Strong NOE effects were found between the protons H3
of magnesium perchlorate were added. Thus, the diastereo- with HaЈ (one of the methylene protons of the trityloxyme-
selectivity increased from 48:52 to 14:86 for epoxide 4a (En- thyl group) and with H2 for the (2S,3R)-4b diastereoisomer
tries 3 and 10) and from 73:27 to 85:15 for epoxide 4c (data (left side), while for the isomer (2R,3S)-4b (right side), a
not shown).
strong effect was observed between the protons H3 and Ha
The assignment of the epoxide configuration is exem- (the other methylene proton), which concords with the con-
plified for (2S,3R)-4b, based on the chemical correlation figurational assignment from the chemical correlation
outlined in Scheme 3. For this purpose, the epoxide 4b was (Scheme 3). Moreover, the NOE effects between the proton
treated with aqueous hydrochloric acid, followed by diazo- H2 with the protons Ha and H3, as well as between the
methane, to give the diol 5. In the absence of magnesium protons HaЈ and H3 and the methine proton H5Ј, corrobor-
perchlorate, the diastereoisomeric ratio (dr) of 4b was 75:25, ate the proposed structure of (2S,3R)-4b.
as determined by 1H NMR spectroscopy, which is the same
To rationalize the opposite sense in the π-facial selectivity
as the enantiomeric ratio (er, 76:24) of the diol 5, as deter- observed for the epoxidation of the imide 3a versus its ether
mined by optical rotation. Analogously, in the presence of derivative 3b, and the substantially higher diastereoselec-
0.5 equiv. of magnesium perchlorate, the enantiomeric ratio tivity in the presence of magnesium perchlorate, we believe
of diol ester 5 (89:11) is in accordance with the diastereoiso- that conformational control of steric effects seems to be at
meric ratio of epoxide 4b (90:10), within experimental error. play (Scheme 4). In the imide 3a, the conformational equi-
librium between 3a and 3aЈ should lie to the side of 3a
because of the intramolecular hydrogen bonding that exists
between the hydroxy group and the carbonyl functionality
of the cinnamoyl imide (in Scheme 4, the five-membered
ring of the pyrrolidinone is perpendicular, whereas the
carbonϪcarbon double bond is parallel, to the plane of the
page). For the epoxidation of substrate 3a by DMD (the
spiro transition structure is assumed to apply[13]) in acetone
solution in the absence of Mg2ϩ ions, the intermolecular
hydrogen bonding with acetone overrides any intramolecu-
lar interaction. Accordingly, no π-facial selectivity through
conformational control is observed (see Entry 3); in the
presence of magnesium ions, however, chelation between
the hydroxy and carbonyl functionalities should afford the
Scheme 3. Transformation of the epoxide 4b to the diol 5
conformationally fixed chelate 3a(Mg). Since the back side
The absolute configuration of the diol 5 was assigned to of the double bond is sterically obstructed towards DMD
be (2S,3S) by comparison of the sign of rotation with an attack, the major epoxide diastereoisomer (2R,3S)-4a is
authentic sample of optically pure diol 5, which was pre- formed by way of the favored transition structure
pared as reported.[11] Since the acid-catalyzed ring opening 3a‡(front), in which the DMD attack occurs from the front
of the epoxide 4b takes place with inversion at the C3 posi- side, i.e., above the plane of the page. Back-side attack
Eur. J. Org. Chem. 2004, 147Ϫ152
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
149

W. Adam, A. Zhang
FULL PAPER
methyl group; therefore, the (2S,3R)-4b diastereoisomer is
the major epoxide product, formed by way of the 3b‡(back)
transition structure, in which the DMD attack occurs from
the back side, i.e., below the plane of the page. Oxygen atom
transfer from the front side through the transition structure
3b‡(front) would afford the minor epoxide diastereoisomer
(2R,3S)-4b, but again this pathway is sterically obstructed.
Conclusion
In summary, we have shown that either one of the two
diastereoisomeric epoxides may be obtained diastereoselec-
tively by using 5-hydroxymethyl-2-pyrrolidinone or its tri-
phenylmethyl ether derivative as a chiral auxiliary. Chel-
ation of a magnesium ion between the chiral auxiliary and
the imide functionality provides the necessary confor-
mational control to induce steric obstruction towards the
attacking oxidant. Since the acid-catalyzed deprotection of
the chiral auxiliary in the resulting epoxy imides affords the
respective diols 5 in high enantiomeric ratio, our present
results may have preparative applications in oxidation
chemistry, in that useful methodologies may be developed
for the syntheses of valuable building blocks bearing oxy-
gen functionalities.
Scheme 4. Proposed mechanism for the diastereoselective DMD
epoxidation of imide 3a through chelation of a magnesium ion with
the hydroxy/carbonyl functionalities
through transition structure 3a‡(back) would afford the
minor (2S,3R)-4a epoxide, but is disfavored in view of
steric impediments.
In the case of the triphenylmethyl ether derivative 3b, the
greater steric hindrance between the trityloxymethyl group
and the carbonyl group of the cinnamoyl moiety in the 3bЈ
versus 3b conformers should favor the latter. The 3b con-
former is predestined for coordination of a magnesium ion
with the two carbonyl functionalities, to afford preferen-
tially the 3b(Mg) chelate (Scheme 5). The front-side attack
of DMD is then sterically obstructed by the large trityloxy-
Experimental Section
1
General Remarks: H and 13C NMR spectra were recorded with a
Bruker AC 200 (1H: 200 MHz, 13C: 50 MHz) or with a Bruker
Avance 400 (1H: 400 MHz, 13C: 101 MHz) spectrometer. IR spectra
were measured with a PerkinϪElmer 1600 FT-IR spectrophoto-
meter. Silica gel (20Ϫ63 mm, Woelm) was used for flash chroma-
tography. TLC analysis was conducted on precoated silica gel foils
60 F254 (20 ϫ 20 cm) obtained from Merck, Darmstadt, Germany.
The spots were visualized either by UV (254 nm) irradiation or by
staining with a 5% solution of polymolybdic acid in ethanol. Silica
gel (32Ϫ62 mm) from Woelm, Erlangen, Germany was used for
column chromatography. The solvents were dried by standard
methods and purified by distillation before use. Melting points (un-
corrected) were determined on a Buchi B-545. Elemental analyses
were carried out by the Microanalytical Division of the Institute
of Inorganic Chemistry, University of Würzburg. All commercial
reagents were used without further purification.
(5S)-5-(tert-Butyldimethylsiloxymethyl)-2-pyrrolidinone
(1):[6b]
Imidazole (2.35 g, 39.0 mmol), tert-butyldimethylsilyl chloride
(3.00 g, 20.0 mmol), and DMAP (200 mg, 1.78 mmol) was added
to
a solution of (S)-5-hydroxymethyl-2-pyrrolidinone (1.50 g,
13.0 mmol) in dry THF (60 mL). The reaction mixture was stirred
magnetically at room temperature (ca. 20 °C) for 24 h and then the
solvent was evaporated (40 °C, 20 Torr). The resultant oil was taken
up in ethyl acetate (200 mL) and washed with water (3 ϫ 30 mL)
and brine (3 ϫ 30 mL), and then dried with Na2SO4. Evaporation
of the solvent (40 °C, 20 Torr) afforded pyrrodinone 1 (2.90 g, 97%)
1
as a colorless oil. H NMR (400 MHz, CDCl3): δ ؍ 0.05 (s, 6 H),
0.85 (s, 9 H), 1.7 (m, 1 H), 2.15 (m, 1 H), 2.3 (m, 2 H), 3.42 (dd,
J ϭ 10.1, 6.9 Hz, 1 H), 3.6 (dd, J ϭ 10.1, 4.1 Hz, 1 H), 3.74
(m, 1 H), 6.59 (br. s, 1 H) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ
Ϫ5.0, Ϫ4.8, Ϫ3.2, 18.5, 23.2, 25.8, 26.1, 30.0, 56.2, 66.4, 178.8 (s,
CϭO) ppm.
Scheme 5. Proposed mechanism for the diastereoselective DMD
epoxidation of imide 3b through chelation of a magnesium ion to
the carbonyl/carbonyl functionalities
150
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 147Ϫ152

High π-Facial Selectivity Through Chelation of Magnesium Ions
(5S)-1-Cinnamoyl-5-(hydroxymethyl)-2-pyrrolidinone (3a): A solu-
FULL PAPER
3.65 (dd, J ϭ 9.8, 3.7 Hz, 1 H), 4.61 (m, 1 H, H5Ј), 7.2Ϫ7.7 (m,
tion of nBuLi (5.7 mL, 1.6  in THF, 9.1 mmol) was added drop- 20 H, aryl H), 7.8 (d, J ϭ 15.6 Hz, 1 H, H3), 8.05 (d, J ϭ 15.6 Hz,
wise over a period of 3 min to a magnetically well-stirred solution
of pyrrolidinone 1 (1.40 g, 6.10 mmol) in THF (50 mL) at Ϫ78 °C
(dry ice/acetone). After magnetic stirring for 30 min at this tem-
1 H, H2) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 14.6, 21.4, 33.9,
57.3, 60.8, 64.5, 87.5, 119.9, 127.5, 128.1, 128.2, 128.4, 128.9, 129,
129.3, 130, 130.7, 135.4, 144, 145.9, 166 (s, CϭO), 176 (s, CϭO)
perature, trans-cinnamoyl chloride (1.53 g, 9.1 mmol) in THF ppm.
(10 mL) was administered over a period of 5 min. The resulting
(5S)-1-Crotonoyl-5-(trityloxymethyl)-2-pyrrolidinone
(3c):[7]
A
mixture was stirred magnetically at Ϫ78 °C for 2 h and warmed to
room temperature (ca. 20 °C). A saturated, aqueous solution of
NH4Cl (50 mL) was added and the organic layer was extracted with
ethyl acetate (3 ϫ 50 mL). The combined organic layers were dried
(Na2SO4) and evaporated (40 °C, 20 Torr) and then the residue
was purified by chromatography on silica gel (petroleum ether/ethyl
acetate, 5:1), to provide (5S)-5-(tert-butyldimethylsiloxymethyl)-1-
cinnamoyl-2-pyrrolidinone (2.10 g, 96%) as a colorless oil.[7] 1H
NMR (200 MHz, CDCl3): δ ؍ 0.03 (s, 6 H), 0.88 (s, 9 H), 2.1Ϫ2.9
(m, 4 H), 3.72 (dd, J ϭ 6.0, 2.1 Hz, 1 H), 4.0 (dd, J ϭ 10.5, 3.5 Hz,
1 H), 4.55 (m, 1 H, H5Ј), 7.35Ϫ7.6 (m, 5 H, aryl H), 7.8 (d, J ϭ
15.6 Hz, 1 H, H3), 8.0 (d, J ϭ 15.6 Hz, 1 H, H2) ppm. The residue
(670 mg, 1.87 mmol) was dissolved in THF (10 mL) and then
TBAF (2 mL, 1  in THF) was added. The resulting mixture was
stirred magnetically at room temperature (ca. 20 °C) for 4 h. Satu-
rated, aqueous sodium hydrogen carbonate (20 mL) was added and
the mixture was extracted with ethyl acetate (3 ϫ 50 mL). The com-
bined organic layers were dried (Na2SO4) and evaporated (40 °C,
20 Torr) and then the residue was purified by chromatography on
silica gel (CH2Cl2/CH3OH, 10:1) to provide the imide 3a (430 mg,
solution of nBuLi (5.3 mL, 1.6  in THF, 8.5 mmol) was added
dropwise over a period of 3 min to a magnetically well-stirred solu-
tion of pyrrolidinone 2 (2.00 g, 5.60 mmol) in THF (50 mL) at Ϫ78
°C (dry ice/acetone). After magnetic stirring for 30 min at this tem-
perature, trans-crotonoyl chloride (0.9 mL, 8.4 mmol) in THF
(10 mL) was administered over a period of 5 min. The resulting
mixture was stirred at Ϫ78 °C for 2 h and then warmed to room
temperature (ca. 20 °C). Saturated aqueous NH4Cl (50 mL) was
added and then the organic layer was extracted with ethyl acetate
(3 ϫ 50 mL). The combined organic layers were dried (Na2SO4)
and evaporated (40 °C, 20 Torr) and then the residue was purified
by chromatography on silica gel (petroleum ether/ethyl acetate, 5:1)
to provide the imide 3c (2.21 g, 93%). 1H NMR (200 MHz, CDCl3):
δ ؍ 1.9Ϫ2.15 (m, 5 H), 2.52 (ddd, J ϭ 17.8, 9.3, 2.4 Hz, 1 H), 3.0
(m, 1 H), 3.15 (dd, J ϭ 9.6, 2.6 Hz, 1 H), 5.57 (dd, J ϭ 9.6, 3.8 Hz,
1 H), 4.52 (m, 1 H), 7.0Ϫ7.2 (m, 15 H, aryl H) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 18.8, 21.5, 3.8, 57.2, 64.5, 87.4, 124.4,
127.6, 128.3, 129, 144, 146, 166 (s, CϭO), 176.9 (s, CϭO) ppm.
General Procedure for the Epoxidation by meta-Chloroperbenzoic
Acid (mCPBA): A sample of the olefin (100 mg, 0.24 mmol) was
dissolved in dichloromethene (20 mL) and then mCPBA (1.5
equiv.) was added. After magnetic stirring at 20 °C for 8 h, the
solid material was removed by filtration, potassium carbonate (ca.
200 mg) was added, and the mixture was stirred magnetically at 20
°C for 30 min. Again the solid materials were removed by filtration
1
94%) as colorless needles, m.p. 123Ϫ125 °C. H NMR (200 MHz,
CDCl3): δ ؍ 1.9Ϫ2.7 (m, 4 H), 4.05 (m, 2 H), 4.32 (m, 1 H), 6.43
(d, J ϭ 16 Hz, 1 H), 6.59 (br. s, 1 H, OH), 7.26Ϫ7.55 (m, 5 H, aryl
H), 7.73 (d, J ϭ 16.0 Hz, 1 H, H3) ppm. IR (KBr): ν˜ ϭ 3350, 2966,
1760, 1670 cmϪ1. HRMS (EI): m/z calcd for C14H15N2O3 [M ϩ
NH4]ϩ 263.1396; found 263.1402.
1
and the product distribution was determined quantitatively by H
(5S)-5-(Trityloxymethyl)-2-pyrrolidinone (2):[6b] Triphenylmethyl
chloride (8.40 g, 30.0 mmol) was added over 5 min at room tem-
perature (ca. 20 °C) to a mixture of S-5-hydroxymethyl-2-pyrrolidi-
none (2.30 g, 20.0 mmol), triethylamine (4.2 mL, 30.0 mmol), and
DMAP (300 mg, 2.67 mmol) in dry CH2Cl2 (200 mL). The reaction
mixture was stirred magnetically overnight and then washed with
water (3 ϫ 30 mL) and brine (3 ϫ 30 mL); the combined extracts
were dried (Na2SO4) and the solvent was evaporated (40 °C,
20 Torr). The crude product was purified by chromatography on
silica gel (dichloromethane/methanol, 5:1) to afford the ether 2
(5.21 g, 73%). 1H NMR (400 MHz, CDCl3): δ ؍ 1.65 (m, 1 H),
2.15 (m, 1 H), 2.31 (m, 2 H), 3.0 (t, J ϭ 8.1 Hz, 1 H), 3.2 (dd, J ϭ
9.3, 3.8 Hz, 1 H), 3.88 (m, 1 H), 5.98 (br., 1 H), 7.26Ϫ7.43 (m, 15
H) ppm.
NMR spectroscopy, using dimethyl isophthalate as the internal
standard.
General Procedure for the Epoxidation by Dimethyldioxirane
(DMD) in the Absence and Presence of Magnesium Perchlorate: A
sample of the olefin (100 mg, 0.24 mmol), and, when necessary,
magnesium perchlorate (0.1Ϫ3 equiv.), was dissolved in acetone
(15 mL) and then a solution of DMD (0.05 ഠ 0.08 , 2 equiv.) in
acetone was added. After magnetic stirring at 20 °C for 8 h, the
solvent was evaporated (20 °C, 20 Torr) and the product distri-
bution was determined quantitatively by 1H NMR spectroscopy
using dimethyl isophthalate as the internal standard.
Epoxide 4a: A colorless, viscous oil was isolated (95% yield) as a
mixture of the (2R,3S) (main) and (2S,3R) (minor) diastereoiso-
(5S)-1-Cinnamoyl-5-(trityloxymethyl)-2-pyrrolidinone (3b):[7]
A
1
mers, which were too labile to be purified by chromatography. H
solution of nBuLi (5.3 mL, 1.6  in THF, 8.5 mmol) was added
dropwise over a period of 3 min to a magnetically well-stirred solu-
tion of pyrrolidinone 2 (2.00 g, 5.60 mmol) in THF (50 mL) at Ϫ78
°C (dry ice/acetone). After magnetic stirring for 30 min at this tem-
NMR (200 MHz, CDCl3): δ ؍ 1.8Ϫ2.7 (m, 4 H), 3.55 (d, J ϭ
1.7 Hz, 1 H, H3), 4.0Ϫ4.4 (m, 4 H), 7.27Ϫ7.37 (m, 5 H, aryl H)
ppm.
perature (ca. 20 °C), trans-cinnamoyl chloride (1.40 g, 8.40 mmol) Epoxide 4b: A colorless, viscous oil was isolated (89% yield) as a
in THF (10 mL) was administered over a period of 5 min. The re-
sulting mixture was stirred at Ϫ78 °C for 2 h and then warmed to
room temperature. Saturated aqueous NH4Cl (50 mL) was added
and the organic layer was extracted with ethyl acetate (3 ϫ 50 mL).
mixture of (2S,3R) (main) and (2R,3S) (minor) diastereoisomers
after purification by chromatography on silica gel (petroleum ether/
1
ethyl acetate, 5:1). H NMR (200 MHz, CDCl3): δ ؍ 1.9Ϫ2.24 (m,
2 H), 2.5 (m, 1 H), 2.96 (m, 1 H), 3.22 (dd, J ϭ 9.8, 2.7 Hz, 1 H),
The combined organic layers were dried (Na2SO4) and evaporated 3.62 (dd, J ϭ 9.8, 3.8 Hz, 1 H), 3.99 (d, J ϭ 1.5 Hz, 1 H, H2), 4.50
(40 °C, 20 Torr) and then the residue was purified by chromatogra-
(m, 1 H, H5Ј), 4.76 (d, J ϭ 1.7 Hz, 1 H, H3), 7.26Ϫ7.37 (m, 20 H,
phy on silica gel (petroleum ether/ethyl acetate, 5:1), to provide the
aryl H) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 22.3, 32.8, 57.1,
1
imide 3b (2.48 g, 91%). H NMR (200 MHz, CDCl3): δ ؍ 2.1 (m, 57.8, 59.6, 61.2, 87.5, 126.7, 126.7, 127.7, 128.3, 128.3, 128.4, 128.9,
1 H), 2.52 (ddd, J ϭ 10.7, 8.5, 2.1 Hz, 1 H), 2.94Ϫ3.23 (m, 2 H),
129, 129, 129.3, 135.6, 143.9, 168.6 (s, CϭO), 177.2 (s, CϭO) ppm.
Eur. J. Org. Chem. 2004, 147Ϫ152
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
151

W. Adam, A. Zhang
FULL PAPER
IR (KBr): ν˜ ϭ 2966, 1757, 1670 cmϪ1. HRMS (EI): m/z calcd for
C33H29 NO4 [Mϩ] 504.2175; found 504.2170.
[1] [1a]
[1b]
R. G. Harvey, Synthesis 1986, 605Ϫ619.
K. A.
Jørgensen, Chem. Rev. 1989, 89, 431Ϫ458.
Epoxide 4c: A colorless, viscous oil was isolated (75% yield) as a
mixture of (2S,3R) (main) and (2R,3S) (minor) diastereoisomers,
after purification by chromatography on silica gel (petroleum ether/
[2]
[2a] P. Bovicelli, P. Lupatelli, E. Mincione, T. Prencipe, R. Curci,
[2b]
J. Org. Chem. 1992, 57, 2182Ϫ2184.
W. Adam, R. T. Fell,
C. R. Saha-Möller, C. G. Zhao, Tetrahedron: Asymmetry 1998,
1
[2c]
ethyl acetate, 5:1). H NMR (200 MHz, CDCl3): δ ؍ 1.5 (d, J ϭ
9, 397Ϫ401.
R. W. Murray, M. Singh, B. L. Williams, H.
M. Moncrieff, J. Org. Chem. 1996, 61, 1830Ϫ1841. [2d] D. Yang,
8.0 Hz, 3 H, CH3), 1.9Ϫ2.23 (m, 2 H), 2.54 (m, 1 H), 2.91Ϫ3.2 (m,
3 H), 3.58 (dd, J ϭ 9.8, 4.0 Hz, 1 H), 4.3 (d, J ϭ 1.7 Hz, 1 H, H3),
4.43 (m, 1 H, H5Ј), 7.26Ϫ7.37 (m, 15 H, aryl H) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 17.9, 22.4, 32.9, 56.2, 56.7, 57.1, 64.2, 87.5,
127.6, 127.6, 128.3, 128.4, 128.9, 143.9, 147.3, 169.5 (s, CϭO),
G.-S. Jiao, Y. C. Yip, M. K. Wong, J. Org. Chem. 1999, 64,
1635Ϫ1639.
Perkin. Trans. 1 2001, 2861Ϫ2873.
khovtsev, C. Gonzales, J. Am. Chem. Soc. 1998, 120,
[2e]
A. Armstrong, A. G. Draffan, J. Chem. Soc.,
[2f]
R. D. Bach, M. N. Glu-
9902Ϫ9910.
177.3 (s, CϭO) ppm. IR (KBr): ν˜ ϭ 2966, 1750, 1670 cmϪ1
.
[3]
[4]
M. J. Lucero, K. N. Houk, J. Org. Chem. 1998, 63, 6973Ϫ6977.
W. Adam, S. G. Bosio, B. T. Wolff, Org. Lett. 2003, 5, 819Ϫ822.
C28H27NO4 (441.5): calcd. C 76.17, H 6.16, N 3.17; found C 76.54,
H 6.07, N 2.67.
[5] [5a]
K. Tomioka, T. Suenaga, Tetrahedron Lett. 1986, 27,
[5b]
Methyl (2S,3S)-2,3-Dihydroxy-3-phenylpropanoate (5):[11] DMD
(7.5 mL, 0.08  in acetone, 0.60 mmol) was added to a mixture of
imide 3b (100.0 mg, 0.21 mmol) and magnesium perchlorate
(23.0 mg, 0.10 mmol) in acetone (15 mL) and then the resulting
mixture was stirred magnetically for 8 h at room temperature (ca.
20 °C). The solvent was evaporated (40 °C, 20 Torr), the residue
was dissolved in water (20 mL), and then 37% aqueous hydro-
chloric acid (1 mL) was added. The reaction mixture was stirred
magnetically overnight and then the solvent was evaporated (40 °C,
20 Torr). The resultant acid was treated at room temperature with
an ethereal solution of diazomethane (20 mL). After evaporation
of the solvent (40 °C, 20 Torr), the residue was purified by chroma-
tography on silica gel (petroleum ether/ethyl acetate, 2:1) to provide
the diol 5 (30.0 mg, 75%) as a colorless oil. [α]2D0 ϭ 32.3 (c ϭ 0.65,
CHCl3), {ref.[12] for (2R,3R)-5, [α]2D0 ϭ Ϫ41.3 (c ϭ 0.48, CHCl3)}.
1H NMR (400 MHz, CDCl3): δ ؍ 2.68 (br., 1 H, OH), 3.61 (s, 3
H, CH3), 4.42 (d, J ϭ 4.2 Hz, 1 H, H2), 4.93 (d, J ϭ 4.2 Hz, 1 H,
H3), 5.22 (br., 1 H, OH), 7.26Ϫ7.29 (m, 5 H, aryl H) ppm. 13C
NMR (100 MHz, CDCl3): δ ϭ 52.8(q, CH3), 75.2 (d, C-2), 75.4 (d,
C-3), 126.7 (d), 128.6 (d), 128.7 (d, 139.0 (s9, 172.8 (s, CϭO) ppm.
369Ϫ372.
J. Chem. Soc., Perkin Trans. 1 1990, 426Ϫ428.
A. Muraoka, T. Tanaka, M. Sawada, N. Ikota, K. Tomioka,
K. Tomioka, N. Hamada, T. Suenaga, K. Koga,
[5c]
M. Kanai,
[5d]
Tetrahedron Lett. 1995, 36, 9349Ϫ9352.
J. P. Nandy, E. N.
Prabhakaran, S. K. Kumar, A. C. Kunwar, J. Iqbal, J. Org.
Chem. 2003, 68, 1679Ϫ1692.
[6]
[6a] D. A. Evans, J. Bartroli, T. L. Shih, J. Am. Chem. Soc. 1981,
103, 2127Ϫ2129. [6b] D. W. Konas, J. K. Coward, J. Org. Chem.
2001, 66, 8831Ϫ8842.
[7]
[8]
[9]
R. A. N. Crump, I. Fleming, C. Urch, J. Chem. Soc., Perkin
Trans. 1 1994, 701Ϫ706.
W. Adam, A. Smerz, Bull. Soc. Chim. Belg. 1996, 105,
581Ϫ599.
We also examined (data not shown) various other salts, e.g.,
LiClO4 and MgCl2, and Lewis acids [Ti(OiPr)4, TiCl4, SnCl4,
BF3·Et2O)], but either no effect was observed on the dr values
or extensive decomposition of the substrate and/or epoxide oc-
curred to give complex, intractable reaction mixtures.
[10] [10a]
D. Yang, Y. L. Yan, B. Lui, J. Org. Chem. 2002, 67,
7429Ϫ7431. [10b] I. Fleming, N. D. Kindon, J. Chem. Soc., Per-
kin Trans. 1 1995, 303Ϫ315.
B. R. Mattews, W. R. Jackson, H. A. Jacobs, K. G. Watson,
[11]
[12]
Aust. J. Chem. 1990, 43, 1195Ϫ1214.
H. Riechers, H.-P. Albrecht, W. Amberg, E. Baumann, H.
Bernard, H.-J. Boehm, D. Klinge, A. Kling, S. Mueller, J. Med.
Chem. 1996, 39, 2123Ϫ2128.
A. L. Baumstark, P. C. Vasquez, J. Org. Chem. 1988, 53,
3437Ϫ3439.
Acknowledgments
Generous financial support by the Deutsche Forschungsgemein-
schaft and the Fonds der Chemischen Industrie is appreciated.
Aimin Zhang is grateful to the Alexander-von-Humbolt Foun-
dation for a postdoctoral fellowship (2001Ϫ2002).
[13]
Received August 29, 2003
152
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 147Ϫ152