Achiral deuterated derivatizing agent for enantiomeric analysis of carboxylic acids by NMR in a chiral liquid crystalline solvent

Article abstract of DOI:10.1016/S0957-4166(00)00321-9

The use of a deuterated 'probe' for the enantiomeric analysis of chiral carboxylic acids is proposed. The probe is the perdeuterobenzyl fragment which can be easily attached to the acid and removed from the corresponding ester. The analysis is performed t

Full text of DOI:10.1016/S0957-4166(00)00321-9

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 11 (2000) 3635–3644
Achiral deuterated derivatizing agent for enantiomeric
analysis of carboxylic acids by NMR in a chiral liquid
crystalline solvent
Abdelkrim Meddour* and Jacques Courtieu
Laboratoire de Chimie Structurale Organique, ICMO, ESA CNRS no 8074, Universit e´ Paris-Sud,
91405 Orsay cedex, France
Received 27 July 2000; accepted 16 August 2000
Abstract
The use of a deuterated ‘probe’ for the enantiomeric analysis of chiral carboxylic acids is proposed. The
probe is the perdeuterobenzyl fragment which can be easily attached to the acid and removed from the
corresponding ester. The analysis is performed through the measurement of the proton-decoupled
deuterium NMR spectrum of the chiral benzyl ester dissolved in poly-g-benzyl-L-glutamate/dimethylform-
amide liquid crystal. Enantiomeric discrimination was observed for all the studied compounds on the para
and/or a deuterons. There is no need for the deuterons to be located close to the stereogenic centre. Thus
enantiomers were correctly distinguished from the signal of a deuterium located 12 bonds away from the
asymmetric carbon. The major interest of this general technique is that no kinetic resolution should occur
during the derivatization process. © 2000 Elsevier Science Ltd. All rights reserved.
1
. Introduction
NMR spectroscopy in chiral liquid crystalline solvents has proven to be a powerful tool for
1
enantiomeric analysis. The best separations were obtained using liquid crystalline solutions of
poly-g-benzyl- -glutamate (PBLG) in various organic solvents, such as dichloromethane, tetra-
L
hydrofuran or dimethylformamide (DMF).
We have shown that, in such media, two enantiomers of a chiral solute are not oriented in the
same way. This difference in orientation can be observed in the NMR spectra through the
ij
orientation-dependent interactions such as the dipolar couplings, D , the chemical shift
i
i
2
anisotropies, D| , and the quadrupolar splittings, Dw , for spins >1/2. Various nuclei such as
Q
1
13
19
3–5
H, C or F may be used for enantiomeric analysis.
obtained through the use of deuterium (spin=1) NMR.
However, most results have been
Deuterium NMR is efficient for the
3,6,7
*
Corresponding author. E-mail: ameddour@icmo.u-psud.fr
0
957-4166/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00321-9

3636
A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
observation of enantiomers because the quadrupolar interaction is manifested through the
observation of large quadrupolar splittings (Dw ). A slight difference in the orientation of
Q
enantiomers gives rise to a noticeable difference in their Dw ’s.
Q
The main disadvantage in the use of deuterium NMR in liquid crystals for enantiomeric
analysis is the need for deuterium labelling which may require complex synthetic procedures.
The alternative, which consists of acquiring the deuterium NMR spectrum in natural abun-
8
dance, although efficient, requires the use of quite large amounts of chiral solutes. In order to
avoid these constraints, we proposed the use of deuterated ‘probes’ to attach to the solute to be
studied. These probes can be described as achiral deuterated derivatizing agents (ADDA). In
contrast with the classical chiral derivatizing agents (CDA’s), ADDA’s do not need to be chiral
as the enantiomeric discrimination is made by the oriented solvent. Thus, we showed that the
use of the perdeuterobenzyl fragment (C D ꢀCOꢀ) was a powerful method to obtain the
6
5
9
enantiomeric excess (ee) of alcohols, amines and aminoacids.
While a plethora of isotropic NMR methods exist for the enantiomeric analysis of alcohols
10
and amines, relatively few are reported for carboxylic acids. One of the most efficient consists
11
of using chiral oxazolidine-2-selones as CDA’s. However the enantiomers clearly react at
different rates with these CDA’s. This may induce partial kinetic resolution if the reaction is not
1
1
complete, as pointed out by the authors.
Herein, we report the use of the perdeuterobenzyl group as ADDA to determine the
enantiomeric purity of carboxylic acids. This probe is preferred for the following reasons:
“
The benzyl fragment can be easily attached to a carboxylic acid. Moreover, it can be readily
removed by simple hydrogenolysis. This may be important in the case where one needs to
recover the chiral starting material.
1
2
“
“
Heptadeuterobenzyl alcohol is achiral and therefore no kinetic resolution should occur in the
course of derivatization. Therefore, the esterification reaction does not need to be complete.
The heptadeuterobenzyl ester of a chiral acid possesses five magnetically different deuterium
atoms which are in the a , a , ortho, meta or para positions. This provides the possibility of
A
B
measuring the enantiomeric excess on different sites in a molecule.
2
. Results and discussion
2
.1. Synthesis
Benzyl alcohol-d , although commercially available, was readily prepared in two steps starting
7
from toluene-d (Scheme 1). A monobromination of toluene-d in the a position is achieved with
8
8
1
3
N-bromosuccinimide (NBS). Substitution of the bromine with a hydroxy group is realised in
a mixture of dioxane/water in the presence of calcium carbonate providing benzyl alcohol-d
with 47% overall yield.
7
1
4
Scheme 1.

A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
3637
Eleven a,a,2,3,4,5,6-heptadeuterobenzyl esters have been prepared by reacting the correspond-
15
ing carboxylic acids with benzyl alcohol-d using dicyclohexylcarbodiimide (Scheme 2). Esters
7
were obtained in the range of 59–89% yield (Table 1). Esters 3, 6, 9 and 12 were prepared in an
enantiomerically enriched mixture starting from non-racemic acids. The ee measured on the
esters are consistent with those of the related acids, which shows that no racemisation or kinetic
resolution occurs during the esterification reaction.
Scheme 2.
2
.2. NMR results
A proton-decoupled deuterium ( H-{ H}) NMR spectrum of one enantiomer of such esters in
2
1
PBLG liquid crystal contains ten signals which are to be seen as five doublets: two doublets with
intensity 2 for the ortho and meta deuterons, one doublet with intensity 1 for the para deuterium
and one doublet with intensity 1 for each of the diastereotopic a deuterons. In isotropic NMR,
the diastereotopic benzylic deuterons are isochronous when they are remote from the stereogenic
centre. In PBLG liquid crystalline solvent, it was found that, even when they have the same
chemical shift, these two deuterons generally give different quadrupolar splittings, which implies
that the order parameters of the C–D direction for the two benzylic deuterons are different.
Besides, while the ortho, meta and para doublets are centred on approximately the same
chemical shift, the benzylic deuterons give doublets with close to identical chemical shift, which
is about 2.2 ppm upfield from that of the aromatic deuterons. Consequently, a racemic mixture
of a heptadeuterobenzyl ester should contain up to ten quadrupolar doublets, six with intensity
1
and four with intensity 2, if there is no signal superposition and in the case where enantiomers
are discriminated on all sites.
2
1
Our first H-{ H} NMR experiments were performed using PBLG/CH Cl liquid crystalline
2
2
solvent. It was found that, although enantiomeric visualisation was effective, line separations
were usually too small to allow an accurate measurement of ee. We therefore used the liquid
crystalline solvent obtained with PBLG/DMF which led to larger line separations. A typical
spectrum, obtained with ester 3 prepared from (S)-2-methylbutyric acid (32% ee) is presented in
Fig. 1. The two outer doublets are due to the para deuterium: the larger is due to the
(
S)-enantiomer and the smaller is from the (R)-enantiomer. The next three doublets with a
relatively large linewidth correspond to the a deuterons. Among the latter, the two doublets with
the largest quadrupolar splittings belong to different enantiomers. The line separation is total
and an accurate integration can be made from these signals, which leads to 32% ee. Finally, the
inner doublets with large intensities are relative to the ortho and meta deuterons and show no
enantiomeric discrimination. This behaviour is general since only a small or even no discrimina-
tion was observed on the ortho and meta deuterons for all the compounds under study.
2
1
Table 2 summarises the data obtained with H-{ H} NMR in PBLG/DMF for all com-
pounds. This Table includes the quadrupolar splittings of the para (p) and the two a deuterons.

3638
A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
Table 1
Heptadeuterobenzyl esters

A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
3639
Table 1 (Continued)
1
Q
2
Q
It also includes the value ꢀDw −Dw ꢀ/2, which is the spectral line separation of enantiomers for
a given deuterium.
The NMR signals of the a deuterons are broad in comparison with those of the aromatic ring.
They also exhibit the structure of a non-resolved multiplet. This lineshape is due to dipolar
coupling between the benzylic nuclei. Besides, the assignment of the two doublets for the a
deuterons of each enantiomer is not obvious. These two nuclei are diastereotopic. However, as
they are remote from the stereogenic centre, they show the same chemical shift in isotropic
deuterium NMR except for esters 4, 6, 7 and 9, where a difference smaller than 0.07 ppm was
measured. The same behaviour of the chemical shift was observed in PBLG/DMF liquid crystal.
However, four doublets were obtained for these deuterons, i.e. two doublets for each enan-
tiomer. In the case where no difference in the chemical shift of the diastereotopic deuterons is
observed, there is no way to assign the two doublets for a given a deuterium in two enantiomers
1
Q
2
Q
from the spectrum. Therefore, the ꢀDw −Dw ꢀ/2 values could not be calculated for esters 5, 8, 10,
1
1 and 13. Consequently, we have reported the values of the four quadrupolar splittings
measured, which are not to be viewed as assigned.
Enantiomeric discrimination was observed for all the compounds under study. Noticeable
separations have been observed on the para deuterium signal for esters 5–7, 12 and 13, while the
a deuterium atoms are good discriminating sites for esters 3–12. Therefore, no correlation can
be drawn between the spectral discrimination and the position of the deuterium with regard to
the stereogenic centre.

3640
A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
The quadrupolar splitting for the para deuterium of the (S)-enantiomer is larger than that of
the (R)-enantiomer for esters 3, 6 and 12. In contrast, the quadrupolar splittings of the (R)- and
S)-enantiomers for this deuterium are equal in esters 9 and 10. This results implies that no
(
correlation can be made between the absolute configuration and the relative values of quadrupolar
splittings for these compounds.
The discrimination power of this method is illustrated with ester 13 (Fig. 2). A slight spectral
separation is observed on the a deuterons, although this part is not obvious to analyse. However,
for the para deuterium which is 12 bonds away from the asymmetric centre, a spectral separation
of 19.5 Hz is observed. This line splitting is sufficient for the measurement of an eventual ee.
2
1
Figure 1. H-{ H} NMR spectrum obtained with 3.6 mg of an enantiomerically enriched mixture of 3 in PBLG/DMF
34.8 wt% PBLG) at T=310 K. Enantiomeric discrimination is mainly observed on the para and a deuterons
(
2
1
Figure 2. H-{ H} NMR spectrum obtained with 53 mg of a racemic mixture of ester 13 in PBLG/DMF (31.1 wt%
1
2
PBLG) at T=310 K. In this spectrum, E represents one enantiomer and E the other one. Enantiomeric
discrimination is mainly observed on the para deuterium

A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
3641
Table 2
a
Quadrupolar splittings (Hz) of heptadeuterobenzyl esters in PBLG/DMF liquid crystal at T=310 K

3642
A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
Table 2 (Continued)
3
. Conclusion
We have shown that benzyl alcohol-d can be used as an achiral deuterated derivatizing agent
7
to achieve enantiomeric analysis of carboxylic acids through their ester derivatives by deuterium
NMR in PBLG/DMF liquid crystal. Large enantiomeric discriminations were obtained mainly
on the para and a deuterons. No racemisation or kinetic resolution was observed during the
synthetic procedure, which enables the measurement of reliable values of ee. Furthermore, the
deuterium probe does not need to be close to the stereogenic centre. Consequently, this
technique appears to be of a very general use for enantiomeric analysis of chiral carboxylic
acids.

A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
3643
4
. Experimental
4
.1. General
2
H NMR spectra were recorded on a Bruker AM 250 NMR spectrometer equipped with a
broad-band reverse 5 mm probe at 38.39 MHz. For isotropic spectra, the CHDCl residual
2
signal, used as internal reference, was assigned to 5.32 ppm. Enantiomeric excesses of the esters
2
1
were measured by H-{ H} NMR in PBLG/DMF liquid crystal.
4
4
.2. Synthesis
.2.1. Benzyl bromide-d 1
7
Benzoyl peroxide (50 mg, 0.21 mmol) was added to a solution of N-bromosuccinimide (17.80
g, 100.0 mmol) and toluene-d (10.00 g, 100.0 mmol) in carbon tetrachloride (250 ml). The
8
reaction was refluxed for 40 minutes under magnetic stirring. The precipitate was filtered off at
room temperature and the solvent was removed under reduced pressure. The residue was
2
chromatographed on silica gel, using pentane as eluant, to give 14.97 g (84%) of 1; H NMR
(
38.4 MHz, CH Cl ): l (ppm) 7.44 (s, 2D); 7.38 (s, 2D); 7.34 (s, 1D); 4.50 (s, 2D).
2
2
4
.2.2. Benzyl Alcohol-d 2
7
Water (230 ml) and CaCO (41.2 g, 412 mmol) were added to a solution of 1 (14.24 g, 80
3
mmol) in dioxane (230 ml) and the mixture was refluxed for 8 hours. The solution was cooled
and the dioxane was removed under reduced pressure. Methylene chloride (200 ml) was added,
followed by treatment with dilute HCl, at 0°C, until all solids had dissolved. Phases were
separated and the aqueous phase extracted twice with methylene chloride. The combined organic
phases were washed with an NaHCO solution, dried with MgSO and filtered. Removal of the
3
4
2
solvent left an oil which was distilled under reduce pressure to give 5.14 g (56%) of 2; H NMR
(
38.4 MHz, CH Cl ): l (ppm) 7.40 (s, 4D); 7.33 (s, 1D); 4.63 (s, 2D).
2
2
4
.2.3. General procedure for the synthesis of esters
An ether solution (10 ml) of carboxylic acid (1.82 mmol), benzyl alcohol-d 2 (2.00 mmol),
7
N,N-dicyclohexylcarbodiimide (2.73 mmol) and 4-pyrrolidino-pyridine (0.27 mmol) was stirred
at room temperature for 24–48 hours. The precipitate was filtered off and the solvent was
evaporated under reduced pressure. Heptadeuterobenzyl ester was isolated by chromatography
on silica gel.
4
.3. NMR sample preparation
PBLG (130 mg) (degree of polymerisation 562, mol. wt. 70 000–150 000 purchased from
Sigma) was weighted into a 5 mm o.d. NMR tube, dry DMF (250 ml) was added and the
mixture was heated to 70°C until complete dissolution of the polymer. Then heptadeuterobenzyl
ester (3.6–53 mg) was introduced into the NMR tube. In order to homogenize the viscous
mixture, the NMR tube was centrifuged back and forth about 20 times.

3644
A. Meddour, J. Courtieu / Tetrahedron: Asymmetry 11 (2000) 3635–3644
4
.4. NMR measurements in PBLG/DMF
Deuterium NMR measurements in PBLG/DMF were performed at 310 K using the Bruker
BVT-1000 temperature regulation system and samples were spun at about 20 Hz. Broad-band
proton decoupling was achieved by applying the WALTZ-16 composite pulse sequence. Trans-
ients (500–10 000) with 4 k of data points were acquired to obtain a good signal to noise ratio.
References
1
2
3
4
5
6
7
8
9
. Bayle, J. P.; Courtieu, J.; Gabetty, E.; Loewenstein, A.; P e´ chin e´ , J. M. New J. Chem. 1992, 16, 1837.
. Emsley, J. W.; Lindon, J. C. NMR Spectroscopy Using Liquid Crystal Solvents; Pergamon: Oxford, 1975.
. Canet, I.; Courtieu, J.; Loewenstein, A.; Meddour, A.; P e´ chin e´ , J. M. J. Am. Chem. Soc. 1995, 117, 6520.
. Meddour, A.; Berdagu e´ , P.; Hedli, A.; Courtieu, J.; Lesot, P. J. Am. Chem. Soc. 1997, 119, 4502.
. Jakubcova, M.; Meddour, A.; P e´ chin e´ , J. M.; Baklouti, A.; Courtieu, J. J. Fluorine Chem. 1997, 86, 149.
. Meddour, A.; Canet, I; Loewenstein, A.; P e´ chin e´ , J. M.; Courtieu, J. J. Am. Chem. Soc. 1994, 116, 9652.
. Meddour, A.; Atkinson, D.; Loewenstein, A.; Courtieu, J. Chem. Eur. J. 1998, 4, 1142.
. Merlet, D.; Ancian, B.; Courtieu, J.; Lesot, P. J. Am. Chem. Soc. 1999, 121, 5249.
. Meddour, A.; Loewenstein, A.; P e´ chin e´ , J. M.; Courtieu, J. Tetrahedron: Asymmetry 1997, 8, 485.
1
1
0. Parker, D. Chem. Rev. 1991, 91, 1441.
1. Peng, J.; Barr, M. E.; Ashburn, D. A.; Lebioda, L.; Garber, A. R.; Martinez, R. A.; Odom, J. D.; Dunlap, R.
B.; Silks, L. A. J. Org. Chem. 1995, 60, 5540.
1
1
1
1
2. Hartung, W. H.; Simonoff, R. Org. React. 1953, 7, 263.
3. Kalir, A. Org. Synth. 1966, 46, 81.
4. Smith, J. G.; Dibble, P. W; Sandborn, R. E. J. Org. Chem. 1986, 51, 3762.
5. Hassner, A.; Alexanian, V., Tetrahedron Lett. 1978, 4475.
.
.