Chiral separation of D,L-mandelic acid through cellulose membranes

Article abstract of DOI:10.1002/chir.20935

This work reports the chiral separation of D,L-mandelic acid with cellulose membranes. Cellulose was chosen as membrane material because it possesses multichiral carbon atoms in its molecular structure unit. The flux and permselective properties of membra

Full text of DOI:10.1002/chir.20935

CHIRALITY 23:379–382 (2011)
Chiral Separation of D,L-Mandelic Acid Through Cellulose Membranes
CHAO MA, XIAO-LIN XU, PING AI, SHENG-MING XIE, YING-CHUN LV, HAI-QIN SHAN, AND LI-MING YUAN*
Department of Chemistry, Yunnan Normal University, Kunming, People’s Republic of China
ABSTRACT
This work reports the chiral separation of D,L-mandelic acid with cellulose mem-
branes. Cellulose was chosen as membrane material because it possesses multichiral carbon
atoms in its molecular structure unit. The flux and permselective properties of membrane using
aqueous solutions of D,L-mandelic acid as feed solution was studied. The top surface and cross-
section morphology of the resulting membrane were examined by scanning electron micros-
copy. When the membrane was prepared with 8.1 wt % cellulose and 8.1 wt % LiCl in the DMA
casting solution, and the operating pressure and feed concentration of racemate were 0.0125
MPa and 0.5 mg/ml, respectively, over 90% of enantiomeric excess could be obtained. This is
the first report that the cellulose membrane is used for isolating the optical isomers of D,L-man-
delic acid. Chirality 23:379–382, 2011. VV C 2011 Wiley-Liss, Inc.
KEY WORDS: D,L-mandelic acid; cellulose; chirality; separation; enantionselective membrane
INTRODUCTION
EXPERIMENTAL
Many drugs, agrochemicals, food additives and fragrances
are racemic mixtures, having nearly identical physical and
Materials
chemical properties. The enantiomers often exhibit signifi-
cant differences in biological and pharmacological activity.
This phenomenon was discovered in the early 1960s that
only the (R)-enantiomer of thalidomide was used to treat
morning sickness while the (S)-enantiomer was the cause of
The cellulose was purchased from Fluka. D,L-mandelic acid and lithium
chloride was obtained from Acros (Belgium). All the reagents were of
analytical grade and used without any further purification. Pure water
was used as the solvent of feed solutions.
Membrane Preparation
1
–3
serious deformities in unborn children.
To develop an
A total of 3 g of cellulose was swollen in 33 ml of DMA (N,N-dimethy-
lacetamide) at 1008C for 20 h, 3 g of LiCl was then added and stirred for
3 h. The mixture was kept in an ultrasonic bath for 2 h to remove the air
bubbles. The resulting solution was cast on the surface of a glass plate
using a casting knife. After evaporating the liquid cellulose membrane
effective method for preparative scale resolution of racemic
mixtures is very important, especially in production of phar-
maceuticals and food products.
There are many optical resolution methods for racemic
4
–14
1
1
3 h at room temperature, the nascent membrane was dried at 508C for
1 h, then immersed into pure water at 108C. The membrane was
compounds,
but those methods often have many draw-
backs, such as many process steps, high cost, and in particu-
lar the small amounts of optical active compounds in one
operation. The membrane process has many potential advan-
tages in comparison, for example, it has high throughput,
energy saving, more economical and very easy to be scaled
up. Some enantioselective membranes prepared with dense
polymer immobilized DNA, BSA, enzymes, chiral selector or
optical polymer or molecularly imprinted polymers have
been published for the optical resolution of racemic com-
detached from the glass plate and washed in pure water for 24 h. Finally,
the membrane was stored in pure water until use.
Scanning electron microscopy (SEM) images of the cellulose mem-
brane were obtained by using a XL30ES-EM-TMP microscope (Hol-
land).
Permeation Experiment
The permeation experiments were performed at room temperature by
2
using membrane cell whose effective membrane area is 7.0 cm and vol-
1
5–30
16
pounds.
ume is 100 ml. A aqueous solution of 50 ml of 0.5 mg/ml D,L-mandelic
Historically, cellulose was one of the earlier materials used
as chromatographic chiral selectors due to its inherent chiral
nature and ready availability. Cellulose derivative mem-
branes have been widely applied to dialysis, ultrafiltration,
reverse osmosis separation technologies because of the
chemical stability, biological compatibility, high tenacity in
acid was used as feed solutions. Constant pressure was applied to the
membrane cell using nitrogen gas input through a knob located on the
top of the cell. 0.0125 MPa of operating pressure was controlled by adjust-
ing the regulator attached to a gas container and the operation time reached
120 h. All membranes were used once only.
HPLC Analysis
3
1–35
the wet state.
Mandelic acid enantiomer is an important
The concentration of D,L-mandelic acid in feed solution and permeate
were measured by HPLC. The HPLC system was equipped with a Lab-
intermediate reagent for synthesis of chiral compounds and
3
6
chiral selector in solvent extraction. To the best of our
knowledge, until now most enantiomers separated by mem-
brane are amino acid, and there is no example described of
enantiomer separations by cellulose membrane. This is a
report, for the first time, that the membrane prepared
from native cellulose is used as an optical resolution mem-
brane material for isolating the optical isomers of D,L-man-
delic acid.
Contract grant sponsor: National Natural Science Foundation, China;
Contract grant numbers: 20775066, 21075109
*Correspondence to: Li-Ming Yuan, Department of Chemistry, Yunnan Nor-
mal University, Kunming 650092, People’s Republic of China.
E-mail: yuan_limingpd@yahoo.com.cn or yuan_liming@hotmail.com
Received for publication 4 April 2010; accepted in revised form 30 September
2010; Accepted 22 October 2010
DOI: 10.1002/chir.20935
Published online 5 January 2011 in Wiley Online Library
(wileyonlinelibrary.com).
VV C 2011 Wiley-Liss, Inc.

3
80
MA ET AL.
RESULTS AND DISCUSSION
Morphological Structure of Cellulose Membrane
SEM was used to characterize the cellulose enantioselec-
tive membrane. Before scanning analysis, the wet membrane
was treated with propanol and heptane to retain its original
structures, and then snapped in liquid nitrogen to give a gen-
erally clean break for the cross-section scan. The resulting
membrane was coated with gold.
Figure 1 shows the morphologies of the cross-section and
top surface of membrane obtained from the SEM analysis. It
can be seen that no macrovoid was in the cross-section. The
top surface was uniform, and no pore was visible in this reso-
lution. The thickness of membrane is about 50 lm.
Effect of Cellulose Concentration on the
Properties of the Membrane
The native cellulose is not dissolved in generic solvent.
However, cellulose can be dissolved in LiCl/DMA solvent
system because the complex compounds are formed
between cellulose and LiCl/DMA. The cellulose concentra-
tion in the casting solution of DMA depends on the LiCl con-
tent. With an increase of cellulose/LiCl > 1, cellulose did not
dissolve completely in the DMA solvent. By decreasing the
cellulose/LiCl < 1 in membrane casting solution, the mem-
branes resulted an increase in flux but decreased in enantio-
meric excess value of mandelic acid. The best weight ratio of
cellulose over LiCl is 1.0. When the cellulose concentration
decreased from 9.4 wt % to 8.1 wt % in the membrane casting
solution (Fig. 2), the flux through the membrane decreased
and the enantioselectivity increased. The reason was the bet-
ter dissolution of the cellulose for higher DMA content. How-
ever, when the cellulose concentration decreased to 6.3 wt %
in the same case, the flux through the membrane increased
and the selectivity of the membrane reduced due to the
looser membrane structure.
Fig. 1. SEM images of cellulose membrane prepared with 8.1% wt % cellu-
lose (cellulose/LiCl, 1/1, wt/wt) in DMA solvent.
Effect of Drying Time on the Properties of the Membrane
After evaporating the liquid cellulose membrane 13 h at
room temperature, the nascent membrane was dried at 508C
with different drying time. Figure 3 showed that increasing
the drying time from 7 to 19 h, caused the flux to reduce, but
the selectivity of the membrane increased. If the drying time
was more than 11 h, over 90%e.e. could be obtained. This
Tech LC600 liquid delivery pump and UV-vis detector. A personal com-
puter equipped with a LabTech HPLC Workstation system was used to
process the chromatographic data. The chiral column was CHIRALPAK
OD (4.6 mm i.d. 3 250 mm, Daicel, Japan) and n-hexane/isopropanol/tri-
fluoracetic acid (90:10:0.2, v:v:v) was used as the mobile phase with 0.5
ml/min of flow rate at 258C. The detection wavelength was 256 nm. Each
injection volume was 20 ll.
The flux of D,L-mandelic acid was measured according to the equation
as follows:
Q
2
Fluxðmg=m hÞ ¼
At
where Q represents the mass of the solute permeated for a given time, t
the permeation time, and A means the effective membrane area.
The selectivity for enantioseparation was calculated in terms of the
percentage enantiomeric excess (%e.e.) of permeates. The equation is as
follows:
AR ꢀ AS
e:e ð%Þ ¼
3100
AS þ AR
Fig. 2. Q and %e.e. in enantioseparation of D,L-mandelic acid through the
membranes fabricated with 6.3 wt %, 8.1 wt %, and 9.4 wt % cellulose (cellu-
lose/LiCl 5 1). The operating pressure and feed concentration were 0.0125
MPa and 0.5 mg/ml, respectively.
S R
where A is the S-enantiomer peak area and A is the R-enantiomer peak
area obtained from HPLC in the permeation.
Chirality DOI 10.1002/chir

CELLULOSE OPTICAL RESOLUTION MEMBRANE
381
Fig. 3. Q and %e.e. in enantioseparation of D,L-mandelic acid through
cellulose membranes prepared with different drying time at 508C. The operat-
ing pressure and feed concentration were 0.0125 MPa and 0.5 mg/ml,
respectively.
Fig. 5. Q and %e.e. in the chiral separation of D,L-mandelic acid through
the cellulose membranes at different racematic feed concentration. The dry-
ing time and operating pressure was 11 h and 0.0125 MPa, respectively.
behavior could be explained by the evaporation of the solvent
of the liquid membrane, resulting in a tighter membrane.
in the ‘‘elution’’ rates of two isomers became less and
‘‘led’’ to a low enantioselectivity. When the feed concentra-
tion was 0.5 mg/ml, the best enantioselectivity was
achieved (Fig. 6).
Effect of Operating Pressure on Properties
of the Membrane
The effect of operating pressure on the separation pro-
cess for D,L-mandelic acid was studied. As shown in Figure
CONCLUSIONS
Cellulose can be dissolved in LiCl/DMA solvent system
because complex compounds are formed between cellulose
and LiCl/DMA. When the membrane was prepared with
4
0
, when the operating pressure increased from 0.0125 to
.0375 MPa, the flux increased and the %e.e. decreased
quickly. It is believed that the elution time decreased
because of the increased operating pressure and weakened
sorption of D,L-mandelic acid to the membrane. Therefore,
the enantioselectivity was decreased. The best pressure
was 0.0125 MPa.
8
.1 wt % cellulose and 8.1 wt % LiCl in the DMA casting
solution, over 90% of enantiomeric excess of D,L-mandelic
acid could be obtained. This is the first time use of native
cellulose in the membrane to separate a chiral compound.
3
7
The chiral recognition was a result of steric fit of the
enantiomers conformation in the chiral space of cellulose
membrane, and of hydrogen bonding, hydrophobic interac-
tion between racemate and cellulose membrane. This study
paves the way by cellulose to develop the optical resolution
membrane.
Effect of Feed Concentration of Racemate on
Properties of the Membrane
Figure 5 demonstrates that the effect of feed concentra-
tion on the optical resolution of D,L-mandelic acid. With the
feed concentration raised from 0.5 to 1.5 mg/ml, the flux
was remarkably increased; however, the enantiomeric
excess decreased. It is possible because of the higher feed
concentration, much of active points were covered by the
isomer whether it could permeate through the membrane,
and then the one isomer which could not permeate
through the membrane retarding the diffusion of the other
isomer. As a result, the flux increased, but the difference
Fig. 4. Q and %e.e. in the chiral separation of D,L-mandelic acid through
cellulose membranes at different operating pressure. The cellulose mem-
branes were prepared with 8.1 wt % cellulose and 8.1 wt % LiCl. The oven
drying time and concentration of feed solution was 11 h and 0.5 mg/ml,
respectively.
Fig. 6. The enantioseparation chromatograms of mandelic acid by OD col-
umn. A: feed solution; B: permeation solution [Color figure can be viewed in
the online issue, which is available at wileyonlinelibrary.com.]
Chirality DOI 10.1002/chir

3
82
MA ET AL.
19. Higuchi A, Hayashi A, Kanda N, Sanui K, Kitamura H. Chiral separation
LITERATURE CITED
of amino acids in ultrafiltration through DNA-immobilized cellulose mem-
branes. J Mol Struct 2005;739:145–152.
1
2
3
. Rikken GLA, Raupach E. Enantioselective magnetochiral photochemis-
try. Nature 2000;405:932–935.
2
0. Kiyohara S, Nakamura M, Saito K, Sugita K, Sugo T. Binding of dl-trypto-
phan to BSA adsorbed in multilayers by polymer chains grafted onto a
porous hollow-fiber membrane in a permeation mode. J Membr Sci
1999;152:143–149.
. Zhang LJ, Song MF, Tian Q, Ming SG. Chiral separation of l,d-tyrosine
and l,d-tryptophan by ct DNA. Sep Purif Technol. 2007;55:388–391.
. Hofstetter O, Hertweck JK, Hofstetter H. Detection of enantiomeric
impurities in a simple membrane-based optical immunosensor. J Bio-
chem Biophys Meth 2005;63:91–99.
2
2
1. Su CL, Dai RJ, Tong B, Deng YL. Preparation and chiral selectivity of
BSA modified ceramic membrane. Chin Chem Lett 2006;17:649–652.
4
5
. Okamoto Y, Yashima E. Polysaccharide derivatives for chromatographic
separation of enantiomers. Angew Chem Int Ed Engl 1998;37:1020–1043.
2. Tone S, Nakamura H. Pressure-driven optical resolution of amino acid
using anion exchange-enzyme binding membrane. Kagaku Kogaku Ron-
bunshu 1988;14:347–353.
. Okamato Y, Senoh T, Nakane H, Hatada K. Optical resolution of betalac-
tams by chiral HPLC on tris(phenylcarbamate)s of cellulose and amylose.
Chirality 1989;1:216–222.
2
3. Lopez JL, Matson SL. A multiphase/extractive enzyme membrane reac-
tor for production of diltiazem chiral intermediate.
1997;125:189.
J Membr Sci
6
7
. Nohira H, Watanabe K, Kurokawa M. Optical resolution of N-benzoyl-cis-
2
Lett 1976;299–300.
-aminocyclohe-xanecarboxylic acid by preferential crystallization. Chem
2
2
4. Lakshmi BB, Martin CR. Enantioseparation using apoenzymes immobi-
lized in a porous polymeric membrane. Nature 1997;388:758–760.
. Sharp VS, Letts MN, Risley DS, Rose AJP. Enantiomeric separation of
dansyl amino acids using macrocyclic antibiotics as chiral mobile phase
additives by narrow-bore high-performance liquid chromatography. Chir-
ality 2004;16:153–161.
5. Wang Y, Hu Y, Xu J, Luo GS, Dai YY. Immobilization of lipase with a spe-
cial microstructure in composite hydrophilic CA/hydrophobic PTFE
membrane for the chiral separation of racemic ibuprofen. J Membr Sci
2
007;293:133–141.
8
9
. Millot MC. Separation of drug enantiomers by liquid chromatography
and capillary electrophoresis, using immobilized proteins as chiral selec-
tors. J Chromatogr B 2003;797:131–159.
2
6. Bruin TJMD, Marcelis ATM, Zuilhof H, Rodenburg LM, Niederlander
HAG, Koudijs A, Overdevest PEM, Padt AVD, Sudholter EJR. Separation
of amino acid enantiomers by micelle-enhanced ultrafiltration. Chirality
. Armstrong DW, Rundlett KL, Chen JR. Evaluation of the macrocyclic an-
tibiotic vancomycin as a chiral selector for capillary electrophoresis. Chir-
ality 1994;6:496–509.
2
000;12:627–636.
2
2
7. Hazarika S. Enantioselective permeation of racemic alcohol through poly-
meric membrane. J Membr Sci 2008;310:174–183.
1
1
0. Ward TJ, Dann C III, Brown AP. Separation of enantiomers using vanco-
mycin in a countercurrent process by suppression of electroosmosis.
Chirality 1996;8:77–83.
8. Yang M, Xie SM, Wang JY, Yuan LM. Optical resolution of (R,S)-2-phe-
nyl-1-propanol through enantioselective ethycellulose membranes. J Appl
Polym Sci 2008;20:935–938.
1. Reeve TB, Cros JP, Gennari C, Piarulli U, de Vries JG. A practical
approach to the resolution of racemic N-benzyl a-amino acids by liquid–
liquid extraction with a lipophilic chiral salen-cobalt (III) complex. Angew
Chem Int Ed Engl 2006;45:2449–2453.
2
9. Wang WF, Xiong WW, Zhao M, Sun WZ, Li FR, Yuan LM, Chiral separa-
tion of trans-stilbene oxide through cellulose acetate butyrate membrane.
Tetrahedron Asymmetry 2009;20:1052–1056.
3
3
0. Dzgoev A, Haupt K. Enantioselective molecularly imprinted polymer
membranes. Chirality 1999;11:465–469.
1
1
1
1
1
1
1
2. Sharp VS, Risley DS. Evaluation of the macrocyclic antibiotic LY333328
as a chiral selector when used as a mobile phase additive in narrow bore
HPLC. Chirality 1999;11:75–81.
1. Yun S, Chen Y, Nayak JN, Kim J. Effect of solvent mixture on properties
and performance of electro-active paper made with regenerated cellulose.
Sens Actuators B Chem 2008;129:652–658.
3. Rundlett KL, Gasper MP, Zhou EY, Armstrong DW. Capillary electropho-
retic enantiomeric separations using the glycopeptides antibiotic, teico-
planin. Chirality 1996;8:88–107.
3
3
2. Zhou JP, Zhang LN, Cai J, Shu H. Cellulose microporous membranes pre-
pared from NaOH/urea aqueous solution. J Membr Sci 2002;210:77–90.
4. Ekborg-Ott KH, Kullman JP, Wang X, Gahm K, He L, Armstrong DW.
Evaluation of the macrocyclic antibiotic avoparcin as a new chiral selector
for HPLC. Chirality 1998;10:627–660.
3. Gao Li, Tang B, Wu P. An experimental investigation of evaporation time
and the relative humidity on a novel positively charged ultrafiltration
membrane via dry-wet phase inversion. J Membr Sci 2009;326:168–177.
5. Newcomb M, Helgeson RC, Cram DJ. Enantiomer differentiation in
transport through bulk liquid membranes. J Am Chem Soc 1974;96:7367–
34. Seman MNA, Khayet M, Hilal N. Nanofiltration thin-film composite poly-
ester polyethersulfone-based membranes prepared by interfacial poly-
merization. J Membr Sci 2010;348:109–116.
7
369.
6. Xie SM, Wang WF, Ai P, Yuan LM. Chiral separation of (R,S)-2-phenyl-1-
propanol through cellulose acetate butyrate membranes. J Membr Sci
35. Sarkar A, Carver PI, Zhang T, Merrington A, Bruza KJ, Rousseau JL, Kei-
nath SE, Dvornic PR. Derdrimer-based coatings for surface modification of
polyamide reverse osmosis membranes. J Membr Sci 2010;349:421–428.
2
008;321:293–298.
7. Xiong WW, Wang WF, Zhao L, Song Q, Yuan LM. Chiral separation of
R,S)-2-phenyl-1-propanol through glutaraldehyde-crosslinked chitosan
(
36. Choi WJ, Lee KY, Kang SH, Lee SB. Biocatalytic enantioconvergent sepa-
ration of racemic mandelic acid. Sep Purif Technol 2007;53:178–182.
membranes. J Membr Sci 2009;328:268–272.
8. Higuchi, Ishida Y, Nakagawa T. Surface modified polysulfone mem-
branes: separation of mixed proteins and optical resolution of tryptophan.
Desalination 1993;90:127–136.
37. Ikai T, Okamoto Y. Structure control of polysaccharide derivatives for
efficient separation of enantiomers by chromatography. Chem Rev
2009;109:6077–6101.
Chirality DOI 10.1002/chir