Preparation, characterization and application of a stationary chromatographic phase from a new (+)-tartaric acid derivative

Article abstract of DOI:10.1016/j.tetlet.2010.02.101

The preparation, characterization and application of a new stationary phase derived from 1,4-cyclohexanedione and diethyl (+)-tartrate are described. A suitable TADDOL for immobilization has been synthesized and grafted to a γ-mercaptopropylsilylated silica gel. The resulting modified stationary phase has been characterized and its ability to separate enantiomers has been studied. While the free TADDOL in solution was able to resolve a range of enantiomers, the resolving properties were lost on immobilization. Solid state ¹³C CPMAS NMR of the new stationary phase was used to explain the lack of stereoselective recognition.

Full text of DOI:10.1016/j.tetlet.2010.02.101

Tetrahedron Letters 51 (2010) 2258–2261
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Preparation, characterization and application of a stationary
chromatographic phase from a new (+)-tartaric acid derivative
a
b
c
d
b,e
Sacha Legrand a, , Harri Heikkinen , Ian A. Nicholls , Andrew Root , Johan Svenson , C. Rikard Unelius
*
a VTT Technical Research Centre of Finland, Biologinkuja 7, Espoo, PO Box 1000, 02044 VTT, Finland
b School of Natural Sciences, Linnaeus University, SE-391 82 Kalmar, Sweden
c MagSol, Tuhkanummenkuja 2, 00970 Helsinki, Finland
d Department of Chemistry, University of Tromsø, N-9037 Tromsø, Norway
e The New Zealand Institute for Plant & Food Research Limited, Canterbury Research Centre, Lincoln 7608, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history:
The preparation, characterization and application of a new stationary phase derived from 1,4-cyclohex-
anedione and diethyl (+)-tartrate are described. A suitable TADDOL for immobilization has been synthe-
sized and grafted to a c-mercaptopropylsilylated silica gel. The resulting modified stationary phase has
been characterized and its ability to separate enantiomers has been studied. While the free TADDOL in
solution was able to resolve a range of enantiomers, the resolving properties were lost on immobilization.
Solid state 13C CPMAS NMR of the new stationary phase was used to explain the lack of stereoselective
recognition.
Received 18 December 2009
Revised 5 February 2010
Accepted 19 February 2010
Available online 23 February 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Chiral stationary phases (CSPs) are of fundamental importance
in liquid chromatography for the resolution of racemates into their
pure enantiomers.1 Partial resolution of racemic camphor with a
lactose-based stationary phase was initially demonstrated by Hen-
derson and Rule as early as 1938.2 Since then, numerous examples
of chiral stationary phases have been reported in the literature.3–5
TADDOLs, (+)-tartaric acid derivatives, have been shown to be
useful as host molecules for the resolution of racemic mixtures.6,7
Tartaric acid-based solid phases have been shown to be promising
chiral selectors for enantioselective liquid chromatography.8–13 Re-
cently, we prepared a new generation of TADDOLs 1a–d derived
from 1,4-cyclohexanedione and diethyl (+)-tartrate (Fig. 1).14
By using a series of 1H NMR studies, we demonstrated the abil-
ity of these new TADDOLs to interact stereoselectively with the
guest alcohols menthol and glycidol.14 Only a few other studies
of this new TADDOL family have been reported so far.15,16 Based
on our promising results and the observed enantioselectivities in
solution, we decided to study the possibility of using this new gen-
eration of TADDOLs as chiral selectors for liquid chromatography.
Herein, we report the synthesis of a new vinyl-containing TAD-
DOL 4, which is suitable for immobilization on a solid phase. The
TADDOL 4 was further coupled to a silica gel which was character-
ized by a series of analytical techniques. Finally, its performance as
a stationary phase was evaluated by chromatography using a range
of non-chiral and chiral analytes.
was possible to determine apparent dissociation constants (app.
Kd) (Table 1) for the complex formation between 1a and the pure
enantiomers of menthol and glycidol (Fig. 2). The mode of interac-
tion was mainly due to hydrogen bond formation between the hy-
droxy groups of TADDOL 1a and the alcohol guests.
Importantly, we demonstrated the ability of the phenyl-con-
taining TADDOL 1a to recognize, with enantioselectivity, both
menthol and glycidol. The difference in recognition between the
enantiomers was not general throughout the study and the TADD-
Ar
Ar
Ar
Ar
1a: Ar = Ph
Ar
HO
HO
Ar
Ar
O
O
O
O
OH
1b: Ar = 1-naphthyl
1c: Ar = 2-naphthyl
1d: Ar = 2-thiophenyl
OH
Ar
Figure 1. Structures of TADDOLs 1a–d.
Table 1
Dissociation constants [Kd
(l
M)]a for complex formation between the phenyl TADDOL
1a and guests (chiral alcohols)
a
Entry
Guest
Kd (lM)
1
2
3
4
(ꢀ)-Menthol
(+)-Menthol
(ꢀ)-Glycidol
(+)-Glycidol
550 30
100 30
190 60
630 20
A series of 1H NMR titration experiments was previously per-
formed in our laboratory.14 These were performed in CDCl3 and it
a
Apparent dissociation constants were calculated with non-linear line fitting to a
one-site model with the software package Prism (version 3.03, GraphPad Software,
USA).
* Corresponding author. Tel.: +358 (0)401763138; fax: +358 (0)207227026.
E-mail address: sacha.legrand@vtt.fi (S. Legrand).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.02.101

S. Legrand et al. / Tetrahedron Letters 51 (2010) 2258–2261
2259
O
O
OH
OH
HO
OH
OH
OH
O
O
O
(-)-menthol
(+)-menthol
(-)-glycidol (+)-glycidol
O OMe
Si
O
4
O OMe
Si
O
Figure 2. Structures of the tested chiral alcohols.
AIBN, CHCl3
O
SH
S
HO
OH
OLs 1b and 1d did not discriminate at all.14 The observed enanti-
oselectivities and differences in dissociation constants (Table 1)
between enantiomers encouraged us to use this new generation
of TADDOLs as chiral selectors for liquid chromatography, espe-
cially for the resolution of small chiral alcohols.
5
6
Scheme 2. Preparation of the chiral solid phase 6.
The design of a suitable TADDOL for immobilization on a solid
support was based on the work described by Degni et al.,17 where
a vinyl-containing TADDOL was immobilized on silica gel. In addi-
tion to an anchoring group, a suitable TADDOL should possess a
similar chemical structure to 1a, with sufficiently bulky pendant
side chains to generate a chiral ‘cavity’ to host the target analytes.
The critical effect of the steric bulk in enantioselective recognition
of TADDOLs has previously been highlighted.14 Therefore, we pre-
pared the vinyl derivative TADDOL 4 in two steps from the 1,4-
cyclohexanedione (2) and (2R,3R)-(+)-tartaric ethyl ester (Scheme
1). Reaction between the tetraester 3 and the Grignard reagent de-
rived from 4-bromostyrene afforded the target molecule (50% iso-
lated yield after purification).
Table 2
Elemental analysis, nitrogen adsorption isotherm measurements and average pore
diameter evaluation values for the silica gels 5 and 6
Entry
Experiment
5
6
1
2
3
4
5
6
% C found
% H found
% O found
10.25
2.30
2.00
6.75
487.8129
67.1932
31.70
3.50
4.70
4.35
866.7478
91.4067
% S found
BET surface area (m2 gꢀ1
)
Average pore diameter (Å)
The TADDOL 4 was attached to the c-mercaptopropylsilylated
silica 518 under radical conditions (AIBN) to afford the silica-bound
derivative 6 (Scheme 2).
CH2 and CH from
reacted vinyl and
reacted sulfur chain
C arom, CH arom
and CH vinyl
Carbons from
The silica gels 5 and 6 were analyzed by Raman spectroscopy,
which is a more sensitive method for silica-bound derivatives than
FT-IR spectroscopy. The Raman spectrum of the c-mercaptopropyl-
unreacted sulfur chain
OCH, C(OH)
and ssba
Si
HS
CH2
CH2
CH2
silylated silica 5 exhibited a very strong band at 2581 cmꢀ1 due to
the S–H bond. This band was found to be very weak in the Raman
spectrum of 6 which clearly indicated that the S–H moieties of 5
had reacted with the double bonds of TADDOL 4, or potentially
with each other to form disulfide bonds.19 Additionally, bands cor-
responding to different vibration modes of the immobilized TAD-
DOL were observed at ca. 3060–3007 (CH arom) on the Raman
spectra of 6. Other characterization of the silica gels 5 and 6 in-
cluded elemental analysis, nitrogen adsorption isotherm measure-
ments and average pore diameter determinations (Table 2).
The 13C NMR spectrum of the TADDOL 4 was partially resolved by
using a DEPT experiment (h = 135°). It was noticed that the chemical
shifts at 114.2, 114.0 and 113.8 ppm correspond exclusively to the
H2C@CH moieties present in 4. Solid state NMR (13C CPMAS NMR)
of the derivatized silica gel 6 was also used as a characterization
method. The 13C CPMAS NMR spectrum of the solid phase 6 is de-
picted in Figure 3. The region at 30–50 ppm, which corresponded
to the CH2 groups from the cyclohexane ring, displayed significantly
O-C-O
and
CH=CH2
ssba
CH2-Si
(reacted and
unreacted)
ssba
250
200
150
100
50
0
ppm (t1)
a ssb = spinning sideband
Figure 3. 13C CPMAS NMR spectrum and assignments of the derivatized silica gel 6.
greater intensity than the region at 112–115 ppm (vinyl H2C@CH).
This observation indicated that the vast majority of the vinyl groups
present in the TADDOL 4 had reacted during the polymerization pro-
cess. The solid state NMR spectrum of 6 also indicated that most of
the SH groups had reacted, since the peak at 27 ppm, due to the S–
OH
HO
EtO2C
O
CO2Et
EtO2C CO2Et
HO OH
O
O
O
O
O
MgBr
THF
BF3.Et2O
EtOAc
O
O
O
O
- 40 o
C
10 o
C
EtO2C
CO2Et
HO
OH
0
oC
2
3
4
Scheme 1. Preparation of the vinyl-containing TADDOL 4.

2260
S. Legrand et al. / Tetrahedron Letters 51 (2010) 2258–2261
CH2–CH2 groups, wasfoundtobemuchsmallerthantheSi–CH2 peak
at 12 ppm. This result supported our previous interpretation based
upon Raman spectroscopy analysis. Collectively, these observations
allowed us to conclude that the TADDOL 4 had been immobilized on
the silica gel 5.
Due to incompatibility issues with reversed-phase LC–MS and
UV detection, it was not possible to use chloroform as the mobile
phase for the liquid chromatographic experiments. In terms of
dielectric constants (DEs), hydrogen bond donor parameters
(HBDPs) and hydrogen bond acceptor parameters (HBAPs), it was
concluded that the binary solvent mixture, 2-propanol (IPA, 10%
in volume) in n-hexane should display properties sufficiently sim-
ilar to chloroform.20 Therefore, we decided to evaluate the new
TADDOL-grafted solid phase in the binary solvent 2-propanol/n-
hexane. However, the TADDOL 4 was not soluble in this mixture
and comparative NMR-titrations could not be performed in this
solvent mixture.
phase 6 and the various non-chiral compounds, we decided to
study the chiral recognition ability of the TADDOL-grafted silica
gel 6. The changes in retention factors (Table 4, entries 1–4) clearly
indicated that the chiral analytes interacted with the derivatized
silica gel. Unfortunately, no enantiomeric recognition of either
menthol or glycidol by the grafted silica gel 6 occurred.
While the solution behaviour of the free TADDOLs, previously
shown to display enantioselectivity,14 greatly differed from an-
chored 6, it was still surprising to observe that the resolving capac-
ity vanished upon immobilization. The LC experiments showed,
beyond doubt, that the analytes had free access to the hydroxy
groups inside the solid support 6 indicating that the loss of selec-
tivity was not caused by poor access. This was further supported
by the large pore size of the derivatized silica (Table 2). A more
likely explanation was instead loss of rotational freedom of the
pendant phenyl groups. These were no longer free to rotate in
the solid state. It may be that the nearly complete cross-linking
in 6, as shown by both solid state NMR and Raman spectroscopy,
disrupted the chiral surroundings around the hydroxy groups. In
solution, these groups were essential for enantioselectivity14 and
it is very likely that they were shifted from their stable chiral equi-
librium state during the immobilization process. The free radical
polymerization is a strongly exothermic and reactive process and
it generated a highly rigid stationary phase in this case. We could
therefore not rule out radical disulfide formation which would fur-
ther add to the rigidification of the system.
A solution to this challenge would be to design a mixed TADDOL
in which only one or two anchoring groups are attached to the struc-
ture. That would yield a stationary phase with recognition elements
that have more freedom to adopt the correct conformation.
In conclusion, a new (+)-tartaric acid derivative has been syn-
thesized and immobilized on silica gel to yield a novel TADDOL-
grafted solid phase. This new derivatized silica gel has been char-
acterized using various chemical, physical and spectroscopic meth-
ods. Interactions between the solid phase and several analytes
have been demonstrated. Although no chiral separation was ob-
served, the loss of the ability of the TADDOL to engage in enantio-
selective recognition could be explained by the loss of freedom of
the system. Alternative coupling strategies and more suitable TAD-
DOL derivatives are suggested improvements for further develop-
ment of these stationary phases.
In an initial investigation of the binding properties of the mod-
ified solid phase, retention factors (k0) were obtained for toluene, 2-
phenylphenol and 2-methoxyphenol in a range of different mix-
tures of n-hexane and IPA (Table 3). A very weak binding was ob-
served for toluene (Table 3, entry 1). This result was probably the
consequence of very weak
p–p interactions between the solid
phase 6 and toluene. On the other hand, 2-phenylphenol and 2-
methoxyphenol displayed significantly longer retention times than
toluene. Their retention times could be further increased by lower-
ing the polarity of the mobile phase. This behaviour supported the
earlier studies on these TADDOLs where the enantioselective inter-
actions were mainly believed to be electrostatic in nature for this
type of analytes. The weakly acidic phenyl hydroxy group in both
the tested phenols readily formed hydrogen bonds with the hydro-
xy groups present in 6, resulting in higher retention factors (Table
3, entries 2 and 3) in a less polar solvent. These initial results fur-
ther indicated that the recognition cavities of the TADDOL 6 were
available for binding after immobilization. A reference column con-
taining only non-derivatized silica gel 5 with free thiol groups dis-
played significantly lower retention factors in control experiments
in apolar solvents.
Based on these preliminary binding experiments, which illus-
trated strong electrostatic interactions between the stationary
Table 3
Acknowledgements
Retention factors found for toluene, 2-phenylphenol and 2-methoxyphenol in various
mobile phasesa
We thank Professor Roland Isaksson (University of Kalmar,
Sweden) for fruitful discussions and Björn Bohman (University of
Kalmar, Sweden) for assistance with the packing of the columns
and the preliminary chromatographic experiments. We also thank
Dr. Lykke Ryelund (University of Copenhagen, Denmark) for Raman
spectroscopy measurements, the Swedish Research Council (grant
2006-6041 to Ian Nicholls), the University of Kalmar (Sweden) and
the Technical Research Centre of Finland (VTT, Finland) for finan-
cial support.
Entry
Analyte
(k0)
1
2
3
Toluene
2-Phenylphenol
2-Methoxyphenol
0.09b
0.32b
0.44b
0.09c
0.64c
0.89c
0.11d
1.70d
2.17d
a
b
c
Flow: 0.8 ml/min; injection volume: 20
Mobile phase: n-hexane/IPA: 60/40.
Mobile phase: n-hexane/IPA: 80/20.
Mobile phase: n-hexane/IPA: 95/5.
ll; detection: UV 254 nm.
d
Supplementary data
Table 4
Retention factors found for (ꢀ)-menthol, (+)-menthol, (ꢀ)-glycidol and (+)-glycidola
Supplementary data (synthesis and characterization of the TAD-
DOL 4, preparation of the silica gels 5 and 6, chromatographic and
solid state NMR experimental procedures) associated with this
article can be found, in the online version, at doi:10.1016/
j.tetlet.2010.02.101.
Entry
Mobile phase
Analyte
(k0)
Selectivity factor (a)
1
2
3
4
5
6
n-Hexane/IPA: 95/5
n-Hexane/IPA: 95/5
n-Hexane/IPA: 99/1
n-Hexane/IPA: 99/1
n-Hexane/IPA: 90/10
n-Hexane/IPA: 90/10
(ꢀ)-Menthol
(+)-Menthol
(ꢀ)-Menthol
(+)-Menthol
(ꢀ)-Glycidol
(+)-Glycidol
1.08
1.08
2.62
2.62
2.48
2.48
n.s.b
n.s.b
n.s.b
n.s.b
n.s.b
n.s.b
References and notes
a
Flow: 0.8 ml/min; injection volume: 20
n.s. = no separation was observed under these experimental conditions.
l
l; detection: UV 254 nm.
1. Allenmark, S. G. Chromatographic Enantioseparation: Methods and Applications;
b
Ellis Horwood Ltd: Chichester, 1988.

S. Legrand et al. / Tetrahedron Letters 51 (2010) 2258–2261
2261
2. Henderson, G. M.; Rule, H. G. Nature 1938, 141, 917.
3. Lämmerhofer, M.; Lindner, W.. In Recent Developments in Liquid Chromatographic
Enantioseparation; Elsevier Eds.: Amsterdam, 2000; Vol. 1, pp 337–437.
12. Ôi, N.; Kitahara, H.; Aoki, F. J. Chromatogr., A 1995, 707, 380–383.
13. Dobashi, Y.; Hara, S. J. Org. Chem. 1987, 52, 2490–2496.
14. Legrand, S.; Luukinen, H.; Isaksson, R.; Kilpeläinen, I.; Lindström, M.; Nicholls, I.
A.; Unelius, C. R. Tetrahedron: Asymmetry 2005, 16, 635–640.
15. Gerard, B.; Sangji, S.; O’Leary, D. J.; Porco, J. A., Jr. J. Am. Chem. Soc. 2006, 128,
7754–7755.
16. Porco, J. A. Jr., Gerard, B. PCT Int. Appl. WO 2007139749, 2007; Chem. Abstr.
2007, 148, 33544.
17. Degni, S.; Wilén, C. E.; Leino, R. Org. Lett. 2001, 3, 2551–2554.
18. Rosini, C.; Altemura, P.; Pini, D.; Bertucci, C.; Zullino, G.; Salvadori, P.; Cinchona,
P. J. Chromatogr., A 1985, 348, 79–87.
4. Subramanian, G.
A Practical Approach to Chiral Separations by Liquid
Chromatography; VCH: Weinheim, 1994.
5. Ikai, T.; Okamoto, Y. Chem. Rev. 2009, 109, 6077–6101.
6. Seebach, D.; Beck, A. K.; Heckel, A. Angew. Chem., Int. Ed. 2001, 40, 92–138.
7. Tanaka, K.; Honke, S.; Urbanczyk-Lipkowska, Z.; Toda, F. Eur. J. Org. Chem. 2000,
3171–3176.
8. Weng, W.; Wang, Q. H.; Yao, B. X.; Zeng, Q. L. J. Chromatogr., A 2004, 1042, 81–87.
9. Oxelbark, J.; Sellén, I. B. Chirality 2003, 15, 787–793.
10. Oxelbark, J.; Claeson, S. Tetrahedron: Asymmetry 2002, 13, 2235–2240.
11. Monser, L. I.; Greeway, G. M.; Erwing, D. F. Tetrahedron: Asymmetry 1996, 7,
1189–1198.
19. Harrisson, S. Macromolecules 2009, 42, 897–898.
20. Allender, C. J.; Heard, C. M.; Brain, K. R. Chirality 1997, 9, 238–242.