Determination of energy barriers to rotation and absolute conformations of thermally interconvertible 5,5-dimethyl-3-(o-aryl)-2,4-oxazolidinedione enantiomers

Article abstract of DOI:10.1016/j.tetasy.2003.12.039

The activation barriers for the interconversion between the enantiomers of 5,5-dimethyl-3-(o-aryl)-2,4-oxazolidinediones (M ? P) have been determined by temperature dependent NMR and by enantioresolution on chiral sorbents via HPLC. The activation barriers were found to increase linearly with the size of the van der Waals radii of the ortho-halogen substituents. The enantiomers of the o-iodo derivative were micropreparatively enriched on a Chiralpak AD column, leading to the determination of its barrier to rotation via thermal racemization and resulting in the assignments of conformations in the presence of the optically active chiral auxiliary (S)-(+)-1-(9-anthryl)-2,2, 2-trifluoro ethanol [(S)-TFAE].

Full text of DOI:10.1016/j.tetasy.2003.12.039

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 925–933
Determination of energy barriers to rotation and absolute
conformations of thermally interconvertible 5,5-dimethyl-3-
(o-aryl)-2,4-oxazolidinedione enantiomers
*
€
_
ꢀ
Oznur Demir Ordu and Ilknur Dogan
ꢀ
Bogazißci University, Chemistry Department, Bebek, 34342 Istanbul, Turkey
Received 21 November 2003; accepted 22 December 2003
Abstract—The activation barriers for the interconversion between the enantiomers of 5,5-dimethyl-3-(o-aryl)-2,4-oxazolidinediones
(M ꢀ P) have been determined by temperature dependent NMR and by enantioresolution on chiral sorbents via HPLC. The
activation barriers were found to increase linearly with the size of the van der Waals radii of the ortho-halogen substituents. The
enantiomers of the o-iodo derivative were micropreparatively enriched on a Chiralpak AD column, leading to the determination of
its barrier to rotation via thermal racemization and resulting in the assignments of conformations in the presence of the optically
active chiral auxiliary (S)-(+)-1-(9-anthryl)-2,2,2-trifluoro ethanol [(S)-TFAE].
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
the two enantiomers by the help of the chiral auxiliary
(S)-(+)-1-(9-anthryl)-2,2,2-trifluoro ethanol [(S)-TFAE]
and by comparison of what we have found in the pre-
vious work for the (5S)-methyl-3-(o-aryl)-2,4-oxazo-
lidinediones.
Compounds bearing an oxazolidine ring are of consid-
erable interest due to their presence in a number of
biologically active synthetic products,1;2 and their utility
as directing groups in asymmetric synthesis.3;4 The 5,5-
dimethyl-2,4-oxazolidinedione ring is present in some
antidiabetic drugs,2 and anticonvulsants.5 Introducing
an o-substituted phenyl ring to the oxazolidine ring on
the nitrogen atom brings dissymmetry and makes these
compounds axially chiral due to restricted rotation
around the Caryl–Nsp2 single bond, which gives an
enantiomeric pair of atropisomers.6
CH
CH
5
CH
O
3
CH
O
3
3
3
O
O
1
5
1
2
4
2
4
3
3
N
O
O
R
N
H
R
H
In our previous work on (5S)-methyl-3-(o-aryl)-2,4-
oxazolidinediones, which were obtained in unequal
compositions of the two diastereomers,7 we found that
the 5-methyl group had enough spacial proximity to the
ortho methyl or hydrogen atom that enabled us to
determine the absolute conformation of the major and
the minor diastereomers. Starting from this set of
compounds, in this study we determine the barrier to
rotation of the analogous enantiomers (Scheme 1) either
by enantioseparation of the racemic compounds on
chiral sorbents via HPLC or by temperature dependent
NMR and comment on the absolute conformations of
(
(
(
(
(
)-1, R= CH
3
)-2, R= F
)-3, R= Cl
)-4, R= Br
)-5, R= I
M
P
Scheme 1. The synthesized compounds.
2. Results and discussion
2.1. H and 13C NMR
1
5,5-Dimethyl-3-(o-aryl)-2,4-oxazolidinedione derivatives
(Scheme 1) were synthesized by the reaction of o-aryl
isocyanates with ethyl a-hydroxyisobutyrate in the
* Corresponding author. Tel.: +90-212-358-1540; fax: +90-212-287-
2467; e-mail: dogan@boun.edu.tr
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2003.12.039

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
926
presence of sodium metal in toluene.7 The rotational
isomers of the synthesized compounds are enantiomeric
and expected to exhibit identical NMR spectra in an
achiral solvent. However, the methyl protons on the C-5
position of the heterocyclic ring, because of the presence
of a Caryl–Nsp2 chiral axis, are diastereotopically related,
and should be, in principle, magnetically nonequivalent
provided that the rate of internal rotation is slow on the
2.2. Barriers to rotation
2.2.1. Temperature dependent NMR. Activation para-
meters to hindered rotation around the C–N single bond
were determined by temperature dependent NMR
spectroscopy for all the synthesized compounds, except
for ( )-5 (for which thermal racemization after micro-
preparative enrichment of the enantiomers has been
applied). The two magnetically nonequivalent C-5
methyl protons, which were distinquished by NMR at
room temperature, became equivalent at higher tem-
peratures (Fig. 1). The kinetic data of the interconver-
sion process were determined using the Eyring
equation,8 and the results are listed in Table 2.
1
NMR time scale. Analysis of the H NMR spectra did
indicate the existence of diastereotopic protons. The two
methyl groups on C-5 of the heterocyclic moiety in
compounds ( )-1, ( )-3, ( )-4, and ( )-5 gave two sep-
arate singlets with chemical shift differences of 0.02,
0.09, 0.12, and 0.2 ppm, respectively, in C6D6 (Table 1).
The diastereotopic methyl groups also exhibited unequal
shifts of 13C nuclei in C6D6, the shift differences being
equal to, 0.7, 0.9, 0.9, and 0.5 ppm, respectively (Table
1). These findings show that the rates of internal rota-
tion are slow on the NMR time scale and confirm the
chirality of the compounds in their ground states. For
compound ( )-2, the ortho-fluoro derivative, on the
other hand, anisochronous carbon and proton nuclei for
the diastereotopic methyl groups were not detected at
ordinary probe temperature (30 ꢁC) by NMR. Lowering
the probe temperature to )55 ꢁC enabled the observa-
tion of these diastereotopic methyl protons.
Comparison of the energy barriers of the studied com-
pounds revealed the influence of the relative sizes of the
substituents on hindered rotation. As can be seen from
Figure 2, the DGz values for the ortho-halogen substi-
tuted compounds ( )-2, ( )-3, ( )-4, and ( )-5 were
found to increase linearly with the van der Waals radii
of the halogens.
Compound ( )-3 bearing an o-chlorine substituent
showed a barrier greater than compound ( )-1 having
an o-methyl group although the van der Waals radius of
Table 1. 1H and 13C NMR spectral data for the synthesized compounds in the presence and absence of the optically active auxiliary, (S)-TFAE
at 30 ꢁC
CH3
CH3
O
O
O
N
Ar
Compound
no
Ar
Medium
1H NMR, ppm, C-5
methyl
1H NMR, ppm,
o-methyl
13C NMR, ppm,
C-5 methyl
( )-1
( )-2
( )-3
( )-4
( )-5
o-Tolyl
C6D6
C6D6+(S)-TFAEa
1.15 and 1.17c
1.97
1.88,
1.89c
––
23.0, 23.7
1.06, 1.08, 1.09c
o-Fluorophenyl
o-Chlorophenyl
o-Bromophenyl
o-Iodophenyl
C6D6
1.11c
23.3
––
Acetone-d6
C6D6+(S)-TFAEb
C6D6
1.68 and 1.69d;e
1.05 and 1.06c
1.16 and 1.25c
1.64 and 1.69c
1.10, 1.11, 1.20, 1.21c
1.19 and 1.31c
1.71 and 1.75d
1.12, 1.13, 1.26c;f
1.21 and 1.41d
1.76d
––
––
––
22.9, 23.8
––
––
DMSO
––
––
C6D6+(S)-TFAEb
C6D6
DMSO
––
––
22.9, 23.8
––
––
––
C6D6+(S)-TFAEb
C6D6
––
21.8, 22.3
––
––
DMSO
CDCl3
––
24.1, 24.4
1.72 and 1.80d
2.27 and 2.32d
0.20, 0.21, 0.39, 0.40d
1.68, 1.69, 1.76, 1.77d
––
––
CDCl2CDCl2
C6D6+(S)-TFAEa
CDCl3+TFAEa
––
––
––
––
––
a 1:8 equivalents of (S)-TFAE were used.
b 1:6 equivalents of (S)-TFAE were used.
c 200 MHz 1H NMR spectral data.
d 400 MHz 1H NMR spectral data.
e At )55 ꢁC.
f A shoulder was observed for the signal at 1.26 ppm.

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
927
120
100
80
o-I
o-Br
o-Cl
60
o-F
40
20
0
1.4
1.5
1.6
1.7
1.8
1.9
2
2.1
°
van der Waals radii, A
Figure 2. The plot of activation barriers versus Van der Waals radii of
the o-substituted halogens.
hexane/ethanol a series of flow rate and temperature
dependent plateaus,9–11 resulting from on-column race-
mization were observed (Fig. 3). The interfering
ÔplateauÕ originated from molecules for which slow
racemization took place during elution on the column,
whereas the two terminal peaks essentially resulted from
molecules that kept their conformation as M or P during
the whole resolution process. Increasing the ratio of
ethanol in the ethanol/hexane eluent resulted in the
disappearance of the plateaus, but decreased the differ-
ence between the retention times. A series of flow rate
and temperature dependent plateaus were also observed
on an AD column for compounds ( )-3 and ( )-4 with
eluent composition of ethanol/hexane (1:1, v/v) at a flow
rate of 0.3 mL/min, at 7 2 ꢁC. On column racemiza-
tion can, in fact, be expected for interconvertible
enantiomers having barriers about these magnitudes
(Table 2) as has been observed before.10
Figure 1. The temperature-dependent 200 MHz 1H NMR spectrum of
the diastereotopic C-5 methyl protons of compound ( )-1.
chlorine is smaller than that for the methyl group. Thus,
this difference cannot be explained by consideration of
the relative sizes in the planar transition states. Elec-
trostatic repulsion between the lone pairs of carbonyl
oxygen and chlorine atom can be expected to make the
rotation around C–N single bond more hindered and
thus causing a larger barrier for this compound.
2.2.2. HPLC. Resolution of the racemic mixtures of
interconverting enantiomers was attempted on Chiralcel
OD-H and Chiralpak AD columns, packed with cellu-
lose tris-3,5-dimethylphenyl carbamate and amylose
tris-3,5-dimethylphenyl carbamate as chiral stationary
phases, respectively. On both columns at room tempera-
ture, separations of enantiomers were not achieved ex-
cept for ( )-4 and ( )-5 due to the fact that the barrier
to rotation was not sufficiently high to allow. The res-
olution of enantiomeric peaks was found to be better on
the AD column where ( )-4 was separated analytically
and ( )-5 micropreparatively in ethanol/hexane (9:1,
v/v) as eluent.
For ( )-5 the micropreparative enrichment of the first
eluted enantiomer using 90% ethanol/10% hexane as
eluent was performed for two purposes: to follow the
interconversion between enantiomers in order to obtain
the kinetic data (Table 2) of the interconversion process
(thermal racemization), which led to the determination
of the activation barrier, and, second, to determine
the absolute conformations of the enantiomers using
(S)-TFAE as an optically active auxiliary by NMR as
will be described later.
In the chromatographic separation of interconverting
enantiomers of ( )-5 on a chiral OD-H column, the
solvent was found to have an effect on the barrier as has
been observed before.7 When the eluent was 80:20 (v/v)
The ethanol solution containing the enriched enantiomer
of ( )-5 was kept at a constant temperature (20 ꢁC,
Table 2. The kinetic and thermodynamic data for the interconversion process shown in Scheme 1
¼e
Compound no
Solvent
T (K)
k (sꢀ1
)
DG (kJ/mol)
( )-1
( )-2
( )-3
( )-4
( )-5
C6D6
Acetone-d6
321c
247c
377c
395c
293f
6.91d
3.64d
73.49 0.05a
57.33 0.05b
83.24 0.05a
86.62 0.05b
94.16 1.31
DMSO-d6
DMSO-d6
Ethanol/hexane (9:1, v/v)
22.97d
28.82d
1 · 10ꢀ4g
a Determined by 200 MHz NMR instrument.
b Determined by 400 MHz NMR instrument.
c The coalescence temperature.
d Rate constant at coalescence temperature.
e Free energy of activation.
f The temperature at which the HPLC analysis has been done.
g The rate constant of interconversion.

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
928
Table 3. 1H NMR chemical shifts of N-o-tolyl substituted O, S, N
containing five membered heterocyclic systems in the presence of
(S)-TFAE. Ar ¼ o-tolyl
We have previously found that the diastereomers of (5S)-
methyl-3-(o-aryl)-2,4-oxazolidinediones were obtained in
unequal compositions.7 The conformations of the major
and the minor diastereomers were deduced from the
analysis of 1H NMR, NOESY, and HMQC spectra, and
it was found that the major diastereomer had a P
conformation. When (5S)-methyl-3-(o-iodophenyl)-2,4-
oxazolidinedione, which is analogous to the compound
( )-5, was injected onto the Chiralpak AD column it was
observed that the minor M conformer eluted first with
eluent composition of ethanol/hexane (9:1, v/v) at a flow
rate of 0.3 mL/min, at 240 nm. Since hydrogen bonding
between racemic compounds and the carbamate residues
of Chiralpak AD column plays an important role in
chiral recognition,12 and provided that these interactions
are the same for the enantiomers of 5,5-dimethyl-3-
(o-iodophenyl)-2,4-oxazolidinedione, ( )-5, and the dia-
stereomers of (5S)-methyl-3-(o-iodophenyl)-2,4-oxazo-
lidinedione, the elution order of these compounds may be
expected to be the same on the AD column, thus the first
eluted enantiomer of the compound ( )-5 may corre-
spond to the M conformer and the second to the P.
Compound
1H NMR,
ppm C-5
methyl
1H NMR,
Experimental
ppm ortho- conditions
methyl
CH3
O
CH3
O
0.80
0.79
0.78
1.59
1.60
8 equiv (S)-TFAE,
C6D6, 400 MHza
O
N
Ar
( )-1
CH3
CH3
1.01
1.00
0.99
1.67
8 equiv (S)-TFAE,
C6D6, 400 MHz
S
S
O
N
Ar
( )-6
1
2.3. H NMR in the presence of a chiral auxiliary
CH3
CH3
The chirality of compounds ( )-1–5 was also confirmed
1
by H NMR spectroscopy in the presence of the chiral
S
1.27
1.28
1.90
8 equiv (S)-TFAE,
C6D6, 200 MHz
O
O
N
auxiliary (S)-(+)-1-(9-anthryl)-2,2,2-trifluoro ethanol
[(S)-TFAE] (Table 1). Noncovalent interactions between
the chiral auxiliary, (S)-TFAE, and the enantiomeric
solute molecules led to the formation of diastereomeric
association complexes. Hydrogen bonding between the
hydroxyl group of (S)-TFAE and oxygen atoms of the
heterocyclic ring, and p–p interaction between anthryl
group and benzene ring have been thought to be
responsible for the formation of these association com-
plexes.6 A special type of complex formation has been
proposed between lactones and (S)-TFAE where (S)-
TFAE hydrogen bonds to the exocyclic oxygen with
the hydroxyl proton and to the ring oxygen with the
methine hydrogen.13–15 The proposed structure of the
complex led to the assignment of the absolute configu-
ration of the furanone based on the consideration of
different shielding effects of the anthryl group on the
lactone protons.13;15
Ar
( )-7
CH3
N
CH3
O
0.80
0.79
0.78
1.62
8 equiv (S)-TFAE,
C6D6, 400 MHz
O
S
Ar
( )-8
a Similar spectrum with 200 MHz NMR.
In the 1H NMR spectra of ( )-1, 3–5, the expected four
singlets were observed in the presence of (S)-TFAE. The
peaks observed with a chemical shift difference of 0.02,
0.09, 0.12, and 0.2 ppm, respectively, due to the diaste-
reotopic methyl groups in the absence of the optically
active auxiliary were found to further split by a chemical
shift difference of 0.01 ppm (Table 1). Only three singlets
(like triplet in appearance) were observed however for
( )-1, resulting from an accidental overlapping of two
inner singlets having the same chemical shifts. o-Methyl
protons of the two diastereomeric complexes were also
observed for ( )-1 with a chemical shift difference of
0.01 ppm.
Figure 3. The flow rate dependent plateaus observed for ( )-5 on OD-
H column, (a), flow rate ¼ 0.3 mL/min, (b), flow rate ¼ 0.2 mL/min, (c),
flow rate ¼ 0.1 mL/min. Eluent composition: 80:20, v/v, hexane/ethanol.
298 K) and its thermal racemization was followed by UV
detection at 240 nm (Fig. 4). The barrier to rotation for
the first-order interconversion process was determined
using the rate constant and the Eyring equation,8 and
found as 94.16 kJ/mol. Data pertinent to the chromato-
graphic separation of the enantiomers of compound ( )-
5 on a Chiralpak AD column are shown in Figure 4.
In the spectrum of ( )-2 taken in the presence of 6 equiv
of (S)-TFAE, two singlets were observed with a chemi-

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
929
Figure 4. The change in enantiomer composition versus time for compound ( )-5 after micropreparative enrichment of the first eluted enantiomer
followed by HPLC at 293 K. Eluent: ethanol/hexane (9:1, v/v), flow rate: 0.3 mL/min, t1: 12.85 min, t2: 13.89 min, k1: 0.33, k2: 0.44, a: 1.33, detection:
UV at 240 nm. Inset: The plot of ln(([M]0)[M]eq/[M])[M]eq)) versus time.
cal shift difference of 0.01 ppm due to the overlapping of
two singlets because of the fast rotation around C–N
single bond. In fact, when the temperature was lowered
to )55 ꢁC, the expected four singlets were observed.
A weaker p–p interaction between anthryl and the
o-substituted phenyl rings and a hydrogen bond between
anthryl hydroxyl and the amide carbonyl oxygen might
be the other possible interactions.13;14 With the knowl-
edge of the stereostructure of (S)-TFAE, in which the
hydroxyl group and the carbinyl hydrogen are aligned to
the same face,16;17 we proposed a solvation model rep-
resented in Figure 5. In this model, while one of the
phenyl rings of the anthryl system could interact with
the ortho-substituted phenyl ring of the compounds
studied the other two rings of the anthryl could not
(Fig. 5).
Determination of absolute stereochemistry of certain
enantiomers can be done on the basis of interactions
between enantiomers and (S)-TFAE.15 In order to get a
1
solvation model for our compounds, all the H NMR
signals of pure (S)-TFAE were compared with those of
the diastereomeric association complex. It has been
observed that all of the signals of the complexed anthryl,
except for the hydroxyl proton, have been shielded with
respect to pure (S)-TFAE. The downfield shift (0.1–
0.2 ppm) noted for O–H proton was thought to result
from hydrogen bonding with oxygens of the heterocyclic
ring. Among these groups, the stronger hydrogen
bonding was expected to occur between C-4 and C-2
carbonyl oxygens and the anthryl hydroxyl due to the
greater basicity of these groups compared to the ring
oxygen. When 1H NMR signals of enantiomers of ( )-5
in the presence of (S)-TFAE were examined in C6D6, all
protons were shifted upfield, probably due to the
anisotropic effect of the anthryl ring. The upfield shift
was also observed in CDCl3 (Table 1) ruling out the
shielding effect of the solvent C6D6. An increase in the
amount of (S)-TFAE also resulted in an enhancement of
the shielding effect. Since an upfield shift was observed
for both C-5 methyl protons of the enantiomers of ( )-5,
the two-point lactone model,13–15 alone cannot account
for the upfield shift of both C-5 methyl protons.
anthryl-
CH3
1
CH3
H
H
O
C
CF3
O
O
O
H
N
O
C
CF3
H
R
anthryl-
2
Figure 5. The proposed solvation model for (M,S) compounds ( )-1–5
(R ¼ CH3, F, Cl, Br, I) and (S)-TFAE.
In order to gain further insight for the solvation model,
NOESY spectra of compound ( )-5 was taken in the
presence of 2 equiv of (S)-TFAE. Close inspection of the
NOESY spectra revealed the presence of the through
space p–p interactions of two of the aromatic protons of

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
930
compound ( )-5 and two of the anthryl protons (either
protons 2, 3 or 6, 7) of (S)-TFAE.
The 5,5-dimethyl-3-(o-aryl)-2,4-oxazolidinedione enantio-
mers studied here are thus thought to interact with
(S)-TFAE in three ways: (1) A two-point interaction
complex is formed with the lactone part of the ring, (2)
an association of the (S)-TFAE hydroxyl proton with
the amide carbonyl oxygen and (3) p–p interactions exist
between the aromatic rings of the two. These interac-
tions are schematically shown in Figure 5. The impor-
tance of the two-point lactone model in differentiation
of C-5 methyl protons by NMR in the presence of
(S)-TFAE might be confirmed when the 1H spectrum of
( )-1 together with (S)-TFAE was compared with those
of 5,5-dimethyl-3-(o-tolyl)-rhodanine ( )-6,6 5,5-dimethyl-
3-(o-tolyl)-2,4-thiazolidinedione ( )-7,18 and 5,5-dimethyl-
3-(o-tolyl)-oxazolidinone-2-thione ( )-8,6 in the presence
of (S)-TFAE (Table 3). While differentiation of the C-5
methyl groups of the enantiomers of ( )-1 was achieved,
for ( )-6 and ( )-8 only one pair of the previously
enantiotopic protons was distinguished in the presence
of (S)-TFAE. The upfield pair was resolved with a
chemical shift difference of 0.01 ppm whereas the
downfield pair appeared as one singlet. On the other
hand, for ( )-7,18 only two singlets were observed
instead of four. Of the o-methyl groups of compounds
( )-1, 6, 7, 8, only those of ( )-1 could be differentiated
(Dd ¼ 0:01 ppm) in NMR. It can be deduced from these
results that the oxygen atoms in the oxazolidinedione
ring play an important role for enantiomeric discrimi-
nation in the presence of (S)-TFAE. As a result of lac-
tone type of association,13–15 with (S)-TFAE, together
with association from the C-4 carbonyl oxygen and the
p–p interactions (Fig. 5), enantiotopic groups of the
enantiomers ( )-1–5 became anisochronous, thus 5,5-
Figure 6. The 1H NMR signals in the presence of 8 equiv of (S)-TFAE
in C6D6 for the 5,5-dimethyl protons of the enantiomers ( )-5. (i) At
30 ꢁC, (ii) at 22 ꢁC, enriched in M conformer.
of anthryl groups. It was thought that the two anthryl
groups might have slightly different anisotropic shielding
effects with the closer one having the larger shielding effect.
Protons (a) and (b0) were exposed to the same shielding
effect by the anthryl1 (Fig. 5) and they were directly
included in the shielding region of this anthryl (Figs. 5
and 6). Since proton (a) unlike (b0) was also shielded by
the o-iodine it appeared as the most shielded proton.
The anthryl2 might have an important role in the
shielding of (a0) and (b). This anthryl group may have
had a tighter approach for the (P,S) solvate than for the
(M,S) solvate,19 owing to the steric interaction of the
lone pair electrons of the o-iodine in the M conformer
and the p electrons of the anthryl group,13 which pre-
vented the two rings being close to each other. There-
fore, by virtue of the greater proximity of the proton (b)
to anthryl2 group than the proton (b0) to anthryl1, the
proton (b) of the P conformer may experience a stronger
shielding effect than the proton (b0) of M conformer.
The proton (a0) was affected by o-iodine together with
anthryl2, therefore it was shielded to the upfield region.
On the other hand, shielding of (a0) by anthryl2 was not
as strong as that of (a) by anthryl1 because of the pre-
vention of a tight complex by o-iodine in the M con-
former.
1
dimethyl protons displayed four distinct H NMR sig-
nals.
1
When H NMR spectrum of a mixture enriched in the
M conformer of compound ( )-5 was taken in the
presence of (S)-TFAE, the chemical shifts of the dia-
stereotopic 5,5-dimethyl protons [(a0) and (b0) in Fig.
6(ii)] of the (M,S) solvate was observed downfield to
that of the (P,S) solvate [(a) and (b) in Fig. 6(i)]. Thus,
the diastereotopic protons of the M conformer were
both deshielded while those of P conformer were shiel-
ded in the presence of (S)-TFAE. That observation
could be explained when the proposed solvate model
(Fig. 5) was taken into account. Since the chemical shift
differences of the diastereotopic protons [(a), (b) and
(a0), (b0)] of ( )-5 was the same before and after the
complex formation, it was thought that the difference
resulted from the position of iodine atom with respect to
the C-5 methyl protons. In the previous work,7 we had
determined by NOESY and HMQC experiments, that
iodine, naphthyl and the other ortho-substituents, due to
their anisotropy, caused a shielding effect on the pro-
tons, which had a close proximity. Therefore the pro-
tons in the upfield pair may be the ones indicated as (a)
and (a0), since they were close to iodine in the P and M
conformations, respectively. The discriminations of (a),
(a0) and (b), (b0) were achieved via analysis of disposition
Since the ortho-substituents are known to have a shield-
ing effect on the syn-substituent at C-5,7 it can be argued
that of all the compounds studied, ( )-1–5, in the pres-
ence of (S)-TFAE the more deshielded of the diastereo-

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
931
topic 5-methyl pairs can be assigned to the complexed M
conformer and the more shielded to complexed P.
Tokyo, Japan). Thermohypersil-Keystone column
pocket was used for the temperature control of the
Chiralpak AD column. Elemental analyses were per-
formed on Carlo Erba 1106. Melting points were re-
corded using Electrothermal 9100 melting point
apparatus.
3. Conclusion
The activation barriers for 5,5-dimethyl-3-(o-tolyl)-2,4-
oxazolidinedione, 5,5-dimethyl-3-(o-fluorophenyl)-2,4-
oxazolidinedione, 5,5-dimethyl-3-(o-chlorophenyl)-2,4-
oxazolidinedione and 5,5-dimethyl-3-(o-bromophenyl)-
2,4-oxazolidinedione were determined as 73.49, 57.33,
83.06, and 86.44 kJ/mol, respectively, by temperature
dependent NMR.
4.1. Syntheses
The compounds, ( )-1–5, 5,5-dimethyl-3-(o-aryl)-2,4-
oxazolidinediones were synthesized by the reaction of
ethyl a-hydroxyisobutyrate and o-aryl isocyanates in the
presence of sodium metal in toluene (Fig. 7).7
The enantiomers of 5,5-dimethyl-3-(o-iodophenyl)-2,4-
oxazolidinedione were resolved micropreparatively on
Chiralpak AD. The thermal racemization of the
enriched enantiomer was followed to obtain the barrier
to rotation, which amounted to 94.16 kJ/mol.
4.1.1. General procedure. In a 100 mL three-necked flask,
fitted with a thermometer and reflux condenser, o-aryl
isocyanate and ethyl a-hydroxyisobutyrate were mixed
in toluene. Sodium metal was added prior to heating in
small pieces. After the addition of sodium, the mixture
was heated for 10 h at about 80 ꢁC then the temperature
was raised to 100–110 ꢁC for 1 h. At the end of the reflux
process, dark-orange crude products were obtained.
After dissolving in ethanol, 5,5-dimethyl-3-(o-aryl)-2,4-
oxazolidinediones separated as white crystals. The
compounds were further purified by recrystallization
The activation energies showed a linear increase with the
van der Waals radii of the ortho-halogen substituents.
A
diastereomeric association model between the
enantiomers of 5,5-dimethyl-3-(o-aryl)-2,4-oxazolidine-
dione and the chiral auxiliary (S)-TFAE has been
drawn. The conformational assignments of the enantio-
mers were achieved via this solvation model, on the basis
of the relative chemical shift values of the well resolved
four singlets of the C-5 methyl protons in the presence of
(S)-TFAE. This solvation model fitted well with the
absolute conformations predicted by comparison of the
elution orders in enantioselective HPLC of the struc-
turally related compounds whose absolute stereostruc-
tures had been determined before.
1
from ethanol and identified based on their H NMR
spectra and elemental analyses.
o-Aryl isocyanates used as starting materials were syn-
thesized by the reaction of the corresponding acid
chloride with sodium azide by the Curtius rearrange-
ment of the acyl azide.7 The isocyanates were identified
by their IR spectra.
4.1.2. 5,5-Dimethyl-3-(o-tolyl)-2,4-oxazolidinedione ( )-1.
The compound was prepared according to the general
procedure using 3.87 g (0.029 mol) ethyl a-hydroxy-
isobutyrate, 3.90 g (0.029 mol) o-tolyl isocyanate, 0.12 g
(0.0052 mol) sodium metal and 15 mL toluene. Yield:
4. Experimental
1H and 13C NMR spectra of all compounds were
recorded on a Varian-Mercury VX-400 MHz-BB (30 ꢁC)
or Bruker AC-200L (200 MHz, 22 ꢁC). IR analyses were
performed on a Mattson Genesis II FTIR using KBr
discs for 2,4-oxazolidinedione derivatives and NaCl
windows for o-aryl isocyanates. Chromatographic
analyses were done using Cecil 1100 pump (P ¼ 3–
3.5 Mpa), a Rheodyne, 7125 injector mode with a 20 lL
sample loop, a Cecil UV monitor (240 nm) and an
integrating recorder. The chiral columns used were
commercially available Chiralcel OD-H (4.6 mm
ID · 250 mm L, 5 lm, Daicel, Tokyo, Japan) and
Chiralpak AD (4,6 mm ID · 250 mm L, 5 lm, Daicel,
1
3.28 g, 51.6%. Mp 91.2–92.2 ꢁC. H NMR (200 MHz)
data in C6D6: diastereotopic methyl protons at C-5:
d ¼ 1:15 ppm (s, 3H), 1.17 ppm (s, 3H). o-Methyl pro-
tons: d ¼ 1:97 ppm (s, 3H). Aromatic protons: d ¼
6:88–7:03 ppm (m, 4H). 13C NMR data in C6D6: car-
bonyl carbons in the heterocyclic ring: 174.8, 153.1 ppm.
Diastereotopic methyl carbons: 23.0, 23.7 ppm. Methine
carbon (C-5) in the heterocyclic ring: 83.5 ppm. o-Methyl
carbon: 17.4 ppm. Aromatic carbons: 127.1, 128.54,
129.8, 130.5, 131.3, 136.2 ppm. Elemental analysis data:
Found: C, 64.61; H, 5.93; N, 5.57%. Calcd for
C12H13O3N: C, 65.74; H, 5.98; N, 6.39%. IR data: m–Ar–H
:
CH3
CH3
O
CH3
O
CH3
Na
C2H5OH
+
+
Ar
N
C
O
C
C
OC2H5
O
O
N
OH
Ar
Figure 7. The synthesis of 5,5-dimethyl-3-(o-aryl)-2,4-oxazolidinediones.

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
932
3066 cmꢀ1, aromatic C–H stretching; m–C–H: 2985, 2934,
22.9, 23.8 ppm. Methine carbon (C-5) in the heterocyclic
ring: 84.2 ppm. Aromatic carbons: 122.9, 128.6, 129.2,
130.5, 131.4, 133.5 ppm. Elemental analysis data:
Found: C, 46.78; H, 3.79; N, 4.61%. Calcd for
C11H10O3NBr: C, 46.5; H, 3.55; N, 4.93%. IR data:
2875 cmꢀ1, alkyl C–H stretching; mAC@O: 1749 cmꢀ1
,
carbonyl stretching; m–C–H: 1498, 1463, 1405 cmꢀ1, alkyl
C–H bending; m–C–H: 767 cmꢀ1, o-disubstituted aromatic
C–H out-of-plane bending; m–C–O: 1173 cmꢀ1, –C–O
stretching.
m
–Ar–H: 3067 cmꢀ1, aromatic C–H stretching; m–C–H: 2986,
2937 cmꢀ1, alkyl C–H stretching; m–C–O: 1749 cmꢀ1, car-
bonyl stretching; m–C–H: 1485, 1411 cmꢀ1, alkyl C–H
4.1.3. 5,5-Dimethyl-3-(o-fluorophenyl)-2,4-oxazolidinedi-
one ( )-2. The compound was prepared according to the
general procedure using 2.06 g (0.015 mol) o-fluoro-
phenyl isocyanate, 1.98 g (0.015 mol) ethyl a-hydroxy-
isobutyrate, 0.03 g (0.0013 mol) sodium metal and 15 mL
toluene. Yield: 1.17 g, 34.7%. Mp 114.1–114.8 ꢁC. 1H
NMR (200 MHz) data in C6D6: Diastereotopic methyl
protons at C-5: d ¼ 1:11 ppm (s, 6H). Aromatic protons:
d ¼ 6:60–7:48 ppm (m, 4H). 13C NMR data in C6D6:
Carbonyl carbons in the heterocyclic ring: 174.2, 160.5,
155.4, 152.4 ppm. Diastereotopic carbons at C-5:
23.3 ppm. Aromatic carbons: 124.8, 125.7, 129.2, 129.6,
131.3, 131.4, 116.9, 116.5 ppm. Elemental analysis data:
Found: C, 59.78; H, 4.84; N, 6.33%. Calcd for
C11H10O3NF: C, 59.19; H, 4.54; N, 6.28%. IR data:
bending; m–C–H: , o-disubstituted aromatic
763 cmꢀ1
C–H out-of-plane bending; m–C–O: 1180 cmꢀ1, –C–O
stretching.
4.1.6. 5,5-Dimethyl-3-(o-iodophenyl)-2,4-oxazolidinedi-
one ( )-5. The compound was prepared according to
the general procedure using 1.59 g (6.5 · 10ꢀ3 mol)
o-iodophenyl isocyanate, 0.86 g (6.5 · 10ꢀ3 mol) ethyl
a-hydroxyisobutyrate, 0.015 g (6.5 · 10ꢀ4 mol) sodium
metal and 15 mL toluene. Yield: 0.87 g, 40.4%. Mp
118.2–119.2 ꢁC. 1H NMR (400 MHz) data in C6D6:
diastereotopic methyl protons at C-5: d ¼ 1:2 ppm (s,
3H),
1.4 ppm
(s,
3H).
Aromatic
protons:
d ¼ 6:4–7:4 ppm (m, 4H). 13C NMR data in C6D6: car-
bonyl carbons in the heterocyclic ring: 172.1 ppm,
150.3 ppm. Diastereotopic methyl carbons at C-5: 21.8,
22.4 ppm. Methine carbon (C-5) in the heterocyclic ring:
82.4 ppm. Aromatic carbons: 127.7, 128.1, 129.5, 132.6,
138.2. Elemental analysis data: Found: C, 39.91; H, 2.99;
N, 3.86%. Calcd for C11H10O3NBr: C, 39.9; H, 3.00; N,
4.23%. IR data: m–Ar–H: 3070 cmꢀ1, aromatic C–H
stretching; m–C–H: 2933, 2984 cmꢀ1, alkyl C–H stretching;
m
–Ar–H: 3078 cmꢀ1, aromatic C–H stretching; m–C–H: 2992,
2942 cmꢀ1, alkyl C–H stretching; mAC@O: 1744 cmꢀ1
,
carbonyl stretching; m–C–H: 1508, 1410 cmꢀ1, alkyl C–H
bending; m–C–H: 774 cmꢀ1, o-disubstituted aromatic C–H
out-of-plane bending; m–C–O: 1170 cmꢀ1, –C–O stretching.
4.1.4. 5,5-Dimethyl-3-(o-chlorophenyl)-2,4-oxazolidinedi-
one ( )-3. The compound was prepared according to the
general procedure using 1.54 g (0.01 mol) o-chlorophenyl
isocyanate, 1.32 g (0.01 mol) ethyl a-hydroxyisobutyrate,
0.023 g (0.001 mol) sodium metal and 15 mL toluene.
Yield: 0.58 g, 24.3%. Mp 94.9–95.6 ꢁC. 1H NMR
(200 MHz) data in C6D6: diastereotopic methyl protons
at C-5: d ¼ 1:16 ppm (s, 3H), 1.25 ppm (s, 3H). Aro-
matic protons: d ¼ 6:60–7:00 ppm (m, 4H). 13C NMR
data in C6D6: carbonyl carbons in the heterocyclic ring:
174.2, 152.3 ppm. Diastereotopic methyl carbons at C-5:
22.9, 23.8 ppm. Methine carbon (C-5) in the heterocyclic
ring: 84.1 ppm. Aromatic carbons: 127.8, 129.2, 130.4,
131.1, 132.9 ppm. Elemental analysis data: Found: C,
53.94; H, 4.14; N, 4.99%. Calcd for C11H10O3NCl: C,
m
–C–O: 1746 cmꢀ1, carbonyl stretching; m–C–H: 1405,
alkyl C–H bending; m–C–H ,
1477 cmꢀ1 767 cmꢀ1
o-disubstituted aromatic C–H out-of-plane bending;
,
:
m
–C–O: 1176 cmꢀ1, –C–O stretching.
Acknowledgements
ꢀ
This project has been supported by Bogazißci University
research fund (BAP) with project number 03B503.
55.13; H, 4.21; N, 5.84%. IR data: m–C–H: 2987 cmꢀ1
alkyl C–H stretching; m–C–O carbonyl
1745 cmꢀ1
stretching; m–C–H: 1488, 1406 cmꢀ1, alkyl C–H bending;
,
References and notes
:
,
1. Walker, R. B.; Fitz, L. D.; Williams, L. M.; Linton, H.;
Smith, C. C. Gen. Pharmacol.: Vascular Syst. 1993, 24,
669–673.
m
–C–H: 772 cmꢀ1, o-disubstituted aromatic C–H out-of-
plane bending; m–C–O: 1171 cmꢀ1, –C–O stretching.
2. Momose, Y.; Maekawa, T.; Yamano, T.; Kawada, M.;
Odaka, H.; Ikeda, H.; Sohda, T. J. Med. Chem. 2002, 45,
1518–1534.
3. Lin, G. Q.; Li, Y. M.; Chan, A. S. C. Principles and
Applications of Asymmetric Synthesis; Wiley Interscience:
New York, 2001. pp 135–188.
4. Aitken, R. A.; Kilenyi, S. N. Asymmetric Synthesis;
Blackie Academic & Professional: London, 1992. pp 83–
140.
5. Shapiro, S. L.; Rose, I. M.; Testa, F. C.; Roskin, E.;
Freedman, L. J. Am. Chem. Soc. 1959, 81, 6498–6504.
4.1.5. 5,5-Dimethyl-3-(o-bromophenyl)-2,4-oxazolidinedi-
one( )-4. The compound was prepared according to the
general procedure using 1.98 g (0.01 mol) o-bromophe-
nyl isocyanate, 1.32 g (0.01 mol) ethyl a-hydroxyiso-
butyrate, 0.02 g (0.00087 mol) sodium metal and 15 mL
toluene. Yield: 1.24 g, 43.7%. Mp 91.7–92.4 ꢁC. 1H
NMR (200 MHz) data in C6D6: Diastereotopic methyl
protons at C-5: d ¼ 1:2 ppm (s, 3H), 1.3 ppm (s, 3H).
Aromatic protons: d ¼ 6:5–7:2 ppm (m, 4H). 13C NMR
data in C6D6: carbonyl carbons in the heterocyclic ring:
174.1, 152.3 ppm. Diastereotopic methyl carbons at C-5:
_
_
ꢀ
6. Dogan, I.; Burgemeister, T.; Ißcli, S.; Mannschreck, A.
Tetrahedron 1992, 48, 7157–7164.
€
7. Demir, O.; Dogan, I. Chirality 2003, 15, 242–250.
_
ꢀ

€
_
ꢀ
O. Demir Ordu, I. Dogan / Tetrahedron: Asymmetry 15 (2004) 925–933
933
8. Alberty, R. A.; Silbey, R. J. Physical Chemistry; Wiley:
New York, 1992.
14. Allinger, N. L.; Eliel, E. L.; Wilen, S. H. Topics in Stereochem-
istry; John Wiley & Sons: New York, 1982; pp 263–331.
15. Jullian, J. C.; Franck, X.; Latypov, S.; Hocquemiller, R.;
Figadere, B. Tetrahedron: Asymmetry 2003, 14, 963–
966.
€
9. Burkle, W.; Karfunkel, H.; Schurig, V. J. Chromatogr.
1984, 288, 1–14.
10. Mannschreck, A.; Kiebi, L. Chromatographia 1989, 28,
263–266.
11. Wolf, C.; Pirkle, W. H.; Welch, C. J.; Hochmuth, D. H.;
16. Pirkle, W. H.; Finn, J. M. J. Org. Chem. 1981, 46, 2935–
2938.
€
Konig, W. A.; Chee, G. L.; Charlton, J. L. J. Org. Chem.
1997, 62, 5208–5210.
17. Lipkowitz, K. B.; Demeter, D. A.; Parish, C. A. Anal.
Chem. 1987, 59, 1733–1736.
_
ꢀ
12. Okamoto, Y.; Kaida, Y. J. Chromatogr. A 1994, 666, 403–
419.
13. Pirkle, W. H.; Sikkenga, D. L. J. Org. Chem. 1977, 42,
1370–1374.
18. Karataßs, M.; Koni, S.; Dogan, I. Can. J. Chem. 1998, 76,
254–259.
19. Beaufour, M.; Merelli, B.; Menguy, L.; Cherton, J. C.
Chirality 2003, 15, 382–390.