Optical resolution and mechanism using enantioselective cellulose, sodium alginate and hydroxypropyl-β-cyclodextrin membranes

Article abstract of DOI:10.1002/chir.22693

Chiral solid membranes of cellulose, sodium alginate, and hydroxypropyl-β-cyclodextrin were prepared for chiral dialysis separations. After optimizing the membrane material concentrations, the membrane preparation conditions and the feed concentrations, enantiomeric excesses of 89.1%, 42.6%, and 59.1% were obtained for mandelic acid on the cellulose membrane, p-hydroxy phenylglycine on the sodium alginate membrane, and p-hydroxy phenylglycine on the hydroxypropyl-β-cyclodextrin membrane, respectively. To study the optical resolution mechanism, chiral discrimination by membrane adsorption, solid phase extraction, membrane chromatography, high-pressure liquid chromatography ultrafiltration were performed. All of the experimental results showed that the first adsorbed enantiomer was not the enantiomer that first permeated the membrane. The crystal structures of mandelic acid and p-hydroxy phenylglycine are the racematic compounds. We suggest that the chiral separation mechanism of the solid membrane is “adsorption – association – diffusion,” which is able to explain the optical resolution of the enantioselective membrane. This is also the first report in which solid membranes of sodium alginate and hydroxypropyl-β-cyclodextrin were used in the chiral separation of p-hydroxy phenylglycine.

Full text of DOI:10.1002/chir.22693

Received: 8 September 2016
DOI: 10.1002/chir.22693
Revised: 24 December 2016
Accepted: 7 January 2017
R E G U L A R A R T I C L E
Optical resolution and mechanism using enantioselective cellulose,
sodium alginate and hydroxypropyl‐β‐cyclodextrin membranes
Li‐Ming Yuan
| Wei Ma† | Mei Xu† | Hui‐Lin Zhao† | Yuan‐Yuan Li‡ | Rui‐Lin Wang‡ |
Ai‐Hong Duan | Ping Ai | Xue‐Xian Chen
School of Chemistry and Chemical
Abstract
Engineering, Yunnan Normal University,
Chiral solid membranes of cellulose, sodium alginate, and hydroxypropyl‐β‐
Kunming, People's Republic of China
cyclodextrin were prepared for chiral dialysis separations. After optimizing the
membrane material concentrations, the membrane preparation conditions and the
feed concentrations, enantiomeric excesses of 89.1%, 42.6%, and 59.1% were
obtained for mandelic acid on the cellulose membrane, p‐hydroxy phenylglycine
on the sodium alginate membrane, and p‐hydroxy phenylglycine on the
hydroxypropyl‐β‐cyclodextrin membrane, respectively. To study the optical resolu-
tion mechanism, chiral discrimination by membrane adsorption, solid phase
extraction, membrane chromatography, high‐pressure liquid chromatography
ultrafiltration were performed. All of the experimental results showed that the first
adsorbed enantiomer was not the enantiomer that first permeated the membrane.
The crystal structures of mandelic acid and p‐hydroxy phenylglycine are the
racematic compounds. We suggest that the chiral separation mechanism of the solid
membrane is “adsorption – association – diffusion,” which is able to explain the opti-
cal resolution of the enantioselective membrane. This is also the first report in which
solid membranes of sodium alginate and hydroxypropyl‐β‐cyclodextrin were used in
the chiral separation of p‐hydroxy phenylglycine.
Correspondence
Li‐Ming Yuan, School of Chemistry and
Chemical Engineering, Yunnan Normal
University, Kunming 650500, P.R. China.
Email: yuan_limingpd@126.com
Funding Information
National Nature Science Foundation of
China, Grant/Award Number: 21365025,
21265026 and 21165022
KEYWORDS
cellulose, chiral separation, enantioselective membrane, hydroxypropyl‐β‐cyclodextrin, optical resolution
mechanism of membrane, sodium alginate
1 | INTRODUCTION
desalination of seawater, wastewater treatment, chemical
engineering, the medical field, metallurgy, etc. After the first
There are many preparation methods for chiral compounds,1-6
but these methods often have some drawbacks, such as many
processing steps, high cost, and low capacities. Since the
advantages of solid membranes are high‐throughput,
energy‐saving, pollution‐free, low‐cost, and easy operation,
solid membrane technologies have been widely used in the
article on chiral solid membranes was published in 1986,7
various enantioselective membranes have been studied.8-15
Recently, some new optical resolution membranes have also
been reported.16-29
To the best of our knowledge, only one original
research article on chiral‐separation‐mechanism‐based solid
enantioselective membranes appeared, in 2001,30 which has
since been introduced and referenced often.9,10 The article
was a theoretical study including some mathematical
†Contributed equally to this work.
‡Contributed equally to this work.
Chirality. 2017;1–10.
wileyonlinelibrary.com/journal/chir
© 2017 Wiley Periodicals, Inc.
1

2
YUAN ET AL.
deductions, which were based on the hypothesis that perme-
ation through a homogeneous film is determined by the solu-
bility and the diffusivity of the molecules in the polymer
matrix. The so‐called solution–diffusion mechanism comes
from gas separation theory.31 The permeability for the enan-
tiomers was expressed as:
The high‐pressure liquid chromatography (HPLC) system
was equipped with a Waters 515 liquid delivery pump and a
Waters 2489 ultraviolet–visible detector (Waters, Milford,
MA). A personal computer equipped with an N2000 HPLC
Workstation system was used to process the chromatographic
data. Fourier transform infrared (FTIR) spectroscopy (Bruker
Tensor 27, Germany) was used to characterize the mem-
branes, and scanning electron microscopy (SEM) images of
the enantioselective membranes were obtained (XL30ES‐
EM‐TMP, Holland). Single‐crystal XRD data for the
compounds were collected on a Rigaku R‐AXIS Spider
X‐ray diffractometer (Rigaku International, Japan).
Pe ¼ SeDe
(1)
where P is the permeability, S the sorption coefficient, and D
the diffusion coefficient. The subscript “e” indicates the D‐ or
L‐enantiomer, respectively.30
Cellulose is a good chromatographic chiral selector due to
its inherent chiral nature and ready availability.1 Cellulose
derivative membranes have been widely applied to dialysis,
ultrafiltration, and reverse osmosis separation technologies
due to its chemical stability, biological compatibility, and high
strength in the wet state. Sodium alginate possesses a high con-
centration of chiral active sites to form a chiral environment in
the membrane, which is critical for the separation of optical
isomers.10 hydroxypropyl‐β‐cyclodextri (HCD) has been
widely used as a mobile phase additive for chromatography
and capillary electrophoresis because of its 3D selectivities.9
In this study, chiral solid membranes of cellulose, sodium
alginate, and hydroxypropyl‐β‐cyclodextrin were prepared to
discriminate the enantiomers of mandelic acid or p‐hydroxy
phenylglycine, which are indispensable reagents in the synthe-
ses of medicines. During the chiral separation, all of the exper-
imental results showed that the first‐adsorbed enantiomer was
not the enantiomer that permeated the membrane first. The
above solution–diffusion mechanism cannot explain the chiral
separations of enantioselective membranes. We suggest that
the chiral separation mechanism of the membranes includes
three steps of “adsorption – association – diffusion,” which
can explain the optical resolution of the enantioselective
membranes. This is also the first report in which solid mem-
branes of sodium alginate and hydroxypropyl‐β‐cyclodextrin
were used in the chiral separation of p‐hydroxy phenylglycine.
2.2 | Chromatographic analysis
The concentrations of the enantiomers in the feed solution
and permeate were measured using HPLC. The chiral column
Crownpak CR(−) (4.6 mm i.d. × 150 mm, 5 μm, Daicel,
Japan) was used to separate p‐hydroxy phenylglycine with a
mobile phase of perchloric acid (pH = 1.0), and Chiralpak
OD (4.6 mm i.d. × 250 mm, 5 μm, Daicel) was used to sep-
arate mandelic acid with a mobile phase hexane:isopropanol:
trifluoroacetic acid =90:10:0.2. The detection wavelength
was 254 nm, and the volume of each injection was 20 μL.
The selectivity for enantioseparation was calculated in
terms of the percentage enantiomeric excess (% ee) of the
permeates:
AD−AL
AD þ AL
e:e:ð%Þ ¼
×100
where AD is the D‐enantiomer peak area and AL is the
L‐enantiomer peak area obtained from the HPLC during the
permeation.
2.3 | Membrane preparation
The cellulose membrane was prepared following a literature
procedure.32 Briefly, after dissolving 3.00 g of cellulose in
33 mL of N,N‐dimethylacetamide (DMAc) and 3.00 g of
LiCl, the resulting solution was cast on the surface of a
glass plate using a casting knife. After drying the liquid cel-
lulose membrane for 13 h at room temperature, the nascent
membrane was dried at 50 °C for 11 h and then immersed
in pure water at 10 °C. The membrane was washed and
stored in pure water.
Next, 6.28 g of sodium alginate was dissolved in 120 mL
of 2 wt.% aqueous acetic acid solution with an ultrasonic bath
for 30 min to break the aggregated molecules and to remove
the air bubbles. The resulting solution was cast on the surface
of a glass plate using a casting knife and dried for 4 days at
room temperature. The resulting dry films were immersed
in an acetone solution containing 5.0 wt.% GA and
2 | MATERIALS AND METHODS
2.1 | Materials and apparatus
D,L‐p‐hydroxy phenylglycine, D‐p‐hydroxy phenylglycine,
L‐p‐hydroxy phenylglycine, D,L‐mandelic acid, L‐mandelic
acid, and D‐mandelic acid were purchased from Acros
(Belgium). 1,6‐Diisocyanatohexane (1,6‐DCH), glutaralde-
hyde (GA), cellulose, sodium alginate, and hydroxy-
propyl‐β‐cyclodextrin (HCD) were purchased from Fluka
(Buchs, Switzerland). Polysulfone ultrafiltration mem-
branes (PS) were obtained from Vontron Membrane Tech-
nology (China). All reagents were of analytical grade and
used without any further purification. Pure water was used
as the solvent in the feed solutions.

YUAN ET AL.
3
1.0 wt.% HCl to crosslink the films (Figure 1). Then they
were removed from the solution and washed with excess ace-
tone and water until they were neutralized. Finally, the
upper and lower frits remained at each end of the cartridge
to hold the packed coated silica gels in place. The entire
SPE assembly was carefully washed with a sufficient amount
of methanol.
33,34
sodium alginate membranes were obtained.
To prepare the hydroxypropyl‐β‐cyclodextrin mem-
brane,35 a flat PS membrane was first immersed in an
aqueous solution containing 20 g/L of hydroxypropyl‐β‐
cyclodextrin for 3 h, then the membrane was removed and
the excess solution was drained off at room temperature in a
vertical position. The coated hydroxypropyl‐β‐cyclodextrin
surface was then immersed in a 1,6‐DCH–heptane solution
in which the crosslinking reaction immediately occurred at
the interface (Figure 2). Amine was used as a catalyst to pro-
mote the reaction. After 10 s of reaction time, the composite
membrane was removed and the membrane surface was
washed with heptane. After the composite membrane was
heated for 20 min at 110 °C, the membrane was kept in
deionized water at 4 °C until use.
2.5 | Column preparation
Stainless steel tubes (25 × 0.20 cm i.d.) (Alltech, Deerfield,
IL) were packed with the above packing materials for chiral
solid phase extraction using a conventional high‐pressure
slurry packing technique with a 1/3 HP liquid pump
(Alltech).
3 | RESULTS AND DISCUSSION
3.1 | FTIR analysis of the membranes
Figure 3 shows the FTIR spectra of the cellulose, sodium
alginate and hydroxypropyl‐β‐cyclodextrin membranes. The
peaks at ~3300 cm−1, ~2900 cm−1, ~1650 cm−1, and
~1060 cm−1 indicated hydroxyl, alkyl, ketone, and ether
groups, respectively. These membranes possessed their char-
acteristic chemical structures.
2.4 | Preparation of chiral solid extraction
cartridges
Silica gels were coated using the above solutions of cellulose,
sodium alginate, and hydroxypropyl‐β‐cyclodextrin. The
same crosslinking reactions for sodium alginate and
hydroxypropyl‐β‐cyclodextrin were carried out with GA or
1,6‐DCH. After filtering through a mesh screen with 75 μm
holes, 0.2000 g of coated silica gels were packed into 3 mL
empty solid extraction cartridges. The 20 μm polyethylene
FIGURE 1 Schematic drawing of the chemical structure of
crosslinked sodium alginate
FIGURE 2 Schematic drawing of the structure of crosslinked
hydroxypropyl‐β‐cyclodextrin
FIGURE 3 FTIR spectra of the membranes: A, cellulose, B, sodium
alginate, C, hydroxypropyl‐β‐cyclodextrin

4
YUAN ET AL.
SEM analysis. Before the scanning analysis, the wet mem-
branes were treated with propanol and heptane to retain their
original structures, and broken under liquid nitrogen to give a
generally clean break. Then the surfaces and cross‐sections of
the resulting membranes were coated with gold. As shown in
Figure 4, the top surfaces of the membranes were tight and
smooth. The thicknesses of the cellulose and sodium alginate
membranes were 75 μm (Figure 4B) and 84 μm (Figure 4D),
respectively, which are symmetrical solid membrane struc-
ture. There were slight differences between the top surfaces
of the hydroxypropyl‐β‐cyclodextrin (Figure 4G) and PS
(Figure 4E) membranes. Compared with the cross‐section
of the PS membrane (Figure 4F), the thin active surface
layers (Figure 4H) were formed through the polymerization
process on the PS membranes. The hydroxypropyl‐β‐cyclo-
dextrin membrane is a composite membrane.
3.2 | Morphologies of the membranes
Figure 4 shows the morphologies of the top surfaces and
cross‐sections of the solid membranes obtained from the
3.3 | Optical resolution
The optical resolutions of racemates were performed via dial-
ysis experiments at room temperature. The effective mem-
brane area of the dialysis cell was 3.3 cm2. The volume of
each side was 10 mL. An aqueous solution of 10 mL of race-
mate was used as the feed solution and the accept phase was
pure water, with the exception of the hydroxypropyl‐β‐cyclo-
dextrin membrane, which used 30% acetonitrile in water
solution as the feed and accepting phases.
3.4 | Effect of the membrane material
concentrations on the properties of the optical
resolution membranes
Some enantioselective membranes were prepared with differ-
ent membrane material concentrations using three prepara-
tion methods.29-32 As shown in Figure 5A,B, when the
cellulose and sodium alginate concentrations increased in
the membrane casting solutions, the enantiomer excess (ee)
values gradually increased and then decreased; there was an
optimal mole ratio of 1,6‐DCH to HCD for the preparation
of the interfacial polymerization membrane of
hydroxypropyl‐β‐cyclodextrin, as shown in Figure 5C. A
higher membrane material concentration should result in a
tighter membrane structure.
3.5 | Effect of membrane preparation on the
properties of optical membranes
After drying the liquid cellulose membrane for 13 h at room
temperature, the nascent membrane was dried at 50 °C for
2–19 h. If the drying time was more than 11 h, over 89% ee
could be obtained (Figure 6A). Figure 6B shows the chiral
separation properties of p‐hydroxy phenylglycine through
the sodium alginate membrane according to the swelling
FIGURE 4 Scanning electron micrographs of the membranes.
Cellulose A,B, sodium alginate C,D, polysulfone E,F,
hydroxypropyl‐β‐cyclodextrin G,H

YUAN ET AL.
5
FIGURE 5 %ee in the chiral separations
through optical membranes fabricated using
different membrane material concentrations.
A, cellulose for mandelic acid, B, sodium
alginate for p‐hydroxy phenylglycine, C,
hydroxypropyl‐β‐cyclodextrin for p‐hydroxy
phenylglycine
index of the membrane. There is usually a close relationship
between the swelling index and the degree of crosslinking.
In the preparation of the hydroxypropyl‐β‐cyclodextrin mem-
brane (Figure 6C), the interfacial polymerization reaction of
1,6‐DCH and HCD on the supported membrane surface was
catalyzed by the triethylamine, and 5 mL of triethylamine
was suitable.
%ee value of 23.4%. The elute order through the sodium algi-
nate and hydroxypropyl‐β‐cyclodextrin membranes was L‐
configuration first and then D‐configuration, but there was
a reverse elute order for the cellulose membrane.
3.7 | Chiral separation mechanism
Chiral discrimination by membrane adsorption, solid phase
extraction, membrane chromatography, HPLC, and ultrafil-
tration were performed to study the optical resolution
mechanism.
3.6 | Effect of the feed concentrations on the
optical resolutions
Figure 7 shows the effect of the feed concentration on the
optical resolutions of the racemic mixtures using the above
membranes. When the concentrations of the feed solutions
for the cellulose, sodium alginate, and hydroxypropyl‐β‐
cyclodextrin membranes were 0.5 mg/mL of mandelic acid,
0.8 mg/mL of p‐hydroxy phenylglycine and 0.5 mg/mL of
hydroxy phenylglycine, %ee of 89.1%, 42.6%, and 59.1.4%
were obtained. Higher concentrations of the racemates
resulted in decreased enantioselectivity of the membrane.
As shown in Figure 7, 30% acetonitrile in water was used
for the accept phase and the feed solution for the
hydroxypropyl‐β‐cyclodextrin membrane. If water was used
as the solvent in the feed solution and accept phase, the
hydroxypropyl‐β‐cyclodextrin membrane only achieved a
3.8 | Membrane adsorption
After finishing the dialysis experiments, the membranes were
removed and drained. The surfaces of the enantioselective
membranes were washed with ethanol, then the enantiomers
in the eluent were determined via chiral HPLC. Compared
to the dialysis permeate, surprisingly, all of the experiment
results revealed that the first‐adsorbed enantiomer was not
the enantiomer that first permeated the membrane. The %ee
for the optical membranes of cellulose, sodium alginate,
and hydroxypropyl‐β‐cyclodextrin were 48.9%, 47.2%, and
55.4%, respectively.

6
YUAN ET AL.
FIGURE 6 %ee in the chiral separations
through optical membranes fabricated using
different membrane preparation methods. A,
cellulose for mandelic acid, B, sodium
alginate for p‐hydroxy phenylglycine, C,
hydroxypropyl‐β‐cyclodextrin for p‐hydroxy
phenylglycine
same as in the membrane dialysis using cellulose, sodium
alginate, and hydroxypropyl‐β‐cyclodextrin.
3.9 | Solid phase extraction
To further verify the enantioselective adsorptions by the
membranes, solid phase extraction cartridges were prepared
according to the membrane preparation method (section
2.5). Prior to SPE, cartridges were washed with methanol
and dried under negative pressure. After loading 2.0 mL of
racemate in water (0.2 mg/mL), the SPE tubes were dried
under a vacuum. In the elution step, the enantiomers were
extracted by methanol (2 mL) that was passed through the
SPE tube at a flow rate of ca.1.0 mL/min. The eluted metha-
nol solutions were collected and analyzed. Compared to opti-
cal membranes of cellulose, sodium alginate, and
hydroxypropyl‐β‐cyclodextrin for the adsorption of mandelic
acid and p‐hydroxy phenylglycine, the same first‐adsorbed
enantiomers were obtained on the SPE tubes, and their ee's
were 32.5%, 31.9%, and 48.2%, respectively.
3.11 | Chiral HPLC
We
prepared
cellulose,
sodium
alginate,
and
hydroxypropyl‐β‐cyclodextrin chiral columns to separate
mandelic acid or p‐hydroxy phenylglycine using water as
the mobile phase with a flow rate of 0.1 mL/min at 25 °C.
Chiral separations were achieved except for with the
hydroxypropyl‐β‐cyclodextrin chiral column. However, when
the mobile phase was methanol/water (1:1), good separation
was realized with the hydroxypropyl‐β‐cyclodextrin column.
The elution order of enantiomers is also the same as in the
membrane chromatography (Figure 9). It is possible that
3.10 | Membrane chromatography
We studied the membrane chromatography using the optical
resolution membranes as the columns (effective area
3.3 cm2) and water as the mobile phase in HPLC. Optical res-
olution of the racemates was achieved at 25 °C and the mem-
brane chromatograms are shown in Figure 8. The elution
orders of the D‐ and L‐configurations in Figure 8 were the
FIGURE 7 %ee in the chiral separations through the optical
membranes vs the concentrations of the feed solution

YUAN ET AL.
7
FIGURE 8 Membrane chromatograms of
A, mandelic acid on cellulose with a flow
rate of 0.3 mL/min; B, p‐hydroxy
phenylglycine on sodium alginate with a
flow rate of 0.1 mL/min; C, p‐hydroxy
phenylglycine on hydroxypropyl‐β‐
cyclodextrin with a flow rate of 0.1 mL/min
FIGURE 9 HPLC chromatograms from
A, mamdelic acid on a cellulose column; B,
p‐hydroxy phenylglycine on a sodium
alginate column; C, p‐hydroxy phenylglycine
on a hydroxypropyl‐β‐cyclodextrin column

8
YUAN ET AL.
Figure 8 is the first enantioselective membrane chromato-
gram, and Figure 9A,B are the first chiral HPLC chromato-
grams with only water as the mobile phase. Further studies
on these aspects will be reported later.
single‐crystal structure analysis. Nevertheless, we determined
the solubilities of the enantiomers and racemate of p‐hydroxy
phenylglycine, which are 17.7 g/L for the L‐configuration,
18.0 g/L for the D‐configuration, and 5.0 g/L for the race-
mate, and confirmed that the p‐hydroxy phenylglycine race-
mate is the compound crystal structure because the
compound crystal structures of the racemates have lower sol-
ubilities than their enantiomers. Therefore, in the racemate
solutions of mandelic acid and p‐hydroxy phenylglycine,
there must be association between the L‐configuration and
the D‐configuration.
3.12 | Ultrafiltration experiments
Using pressure as the driving force,34,36 ultrafiltrations using
the membranes was studied to explore the chiral separation of
the racemates. The selective permeations of the cellulose
(0.2 MPa pressure), sodium alginate (0.3 MPa pressure),
and hydroxypropyl‐β‐cyclodextrin (0.5 MPa pressure) mem-
branes by mandelic acid (0.3 mg/mL), p‐hydroxy
phenylglycine (0.3 mg/mL) and p‐hydroxy phenylglycine
(0.2 mg/mL), respectively, were attained with 80.9%,
16.6%, and 23.2% ee. The enantioselectivities via ultrafiltra-
tion were still in agreement with the above dialysis results.
From the above comprehensive studies, all of the experi-
mental results showed that one enantiomer was adsorbed first
on the surface of the optical resolution membrane, but the
enantioselective permeation was always a different enantio-
mer, regardless of whether the separation was membrane
chromatography, ultrafiltration, or dialysis. What is the chiral
separation mechanism of the optical resolution membranes
for racemates?
3.14 | Optical resolution mechanism
Cellulose, sodium alginate, and hydroxypropyl‐β‐cyclodex-
trin are composed of saccharides with high concentrations
of chiral active carbons on their backbones. Cellulose and
sodium alginate can form helical, chirally active small spaces
in their main chain backbone structures. Hydroxypropyl‐β‐
cyclodextrin is a macrocyclic ring that can form inclusion
complexes with various organic compounds. The chiral rec-
ognition on the surfaces of the optical resolution membranes
is a result of the steric fit of the conformations of the enantio-
mers in the chiral cavities of the membranes, hydrogen bond-
ing, and hydrophobic interactions between the racemates and
the membranes.
Due to the association characteristic of the D,L‐mandelic
acid and D,L‐p‐hydroxy phenylglycine racemates, it is possi-
ble that one kind of conformation enantiomer was adsorbed
first on the surface of the enantioselective membrane, then
the other enantiomer was associated with the first configura-
tion layer. Then the enantioselective permeation of the oppo-
site enantiomer was carried out via the concentration driving
force due to the weak association force between the two enan-
tiomers layers (Figure 11). The enantioseparation mechanism
of the membranes includes three steps of “adsorption –
association – diffusion.”
3.13 | Crystal structure of mandelic acid and
p‐hydroxy phenylglycine racemates
There are three kinds of crystal structures for racemates, mix-
tures, compounds, and solid solutions. Most racemates are
composed of compounds. We recrystallized D,L‐mandelic
acid in water to obtain a single crystal, which was identified
using single‐crystal X‐ray diffraction analysis (Figure 10).
In the single crystal of D,L‐mandelic acid, there is a 1:1
ratio between D‐mandelic acid and L‐mandelic acid and a
regular arrangement of D‐mandelic acid alternating with
L‐mandelic acid.
Unfortunately, we have not obtained a large enough
single crystal of p‐hydroxy phenylglycine to perform
FIGURE 11 Chiral separation mechanism via “adsorption
association – diffusion” in the optical resolution membranes
–
FIGURE 10 Single‐crystal structure of D,L‐ mandelic acid

YUAN ET AL.
9
10. Higuchi A, Tamai M, Ko YA, et al. Polymeric membranes for chiral
separation of pharmaceuticals and chemicals. Polymer Rev.
2010;50:113‐143.
4 | CONCLUSION
Chiral separations of D,L‐mandelic acid or D,L‐p‐hydroxy
phenylglycine may be realized via dialysis through
solid membranes of cellulose, sodium alginate, or
hydroxypropyl‐β‐cyclodextrin. The enantiomeric excess
values can be affected by the membrane material concentra-
tions, membrane preparation conditions, and feed concentra-
tions. The chiral separation by the membranes is possibly a
result of adsorption of one kind of enantiomer followed by
association of the opposite enantiomer and then diffusion of
the opposite configuration enantiomer. The chiral adsorption
depends on the enantioselective properties of the membrane.
The association of the opposite configuration enantiomer is
related to the characteristics of the racemate. The diffusion
of the opposite configuration enantiomer is a result of the
concentration difference. This study will be very useful for
the development of optical resolution membranes.
11. Dai RJ, Su CL, Wu HY, Deng YL. Separation of salsolinol using β‐
cyclodexytrin‐modified chiral membrane. Trans Beijing Inst
Technol. 2008;28:557‐557.
12. Yoshikawa M, Tharpa K, Dima ŞO. Molecularly imprinted mem-
branes: Past, present, and future. Chem Rev. 2016;116:11500‐11528.
13. Ingole PG, Bajaj HC, Singh K. Membrane separation processes:
Optical resolution of lysine and asparagine amino acids. Desalina-
tion. 2014;343:75‐81.
14. Ingole PG, Singh K, Bajaj HC. Enantioselective permeation of α‐
amino acid isomers through polymer membrane containing chiral
metal–Schiff base complexes. Desalination. 2011a;281:413‐421.
15. Ingole PG, Bajaj HC, Singh K. Optical resolution of alpha‐amino
acid derivative through membrane process. Ind J Chem Technol.
2011b;18:197‐206.
16. Kong D, Zhou ZY, Zhu HY, et al. Selective separation of salbutamol
enantiomers with simultaneously synergistic extraction and
stripping method. J Membr Sci. 2016;499:343‐351.
ACKNOWLEDGEMENTS
17. Bushell AF, Budd PM, Attfield MP, et al. Nanoporous organic
polymer/cage composite membranes. Angew Chem Int Ed.
2013;52:1253‐1256.
This work was supported by the National Nature Science
Foundation of China (21365025, 21265026 and 21165022).
18. Weng X, Baez JE, Khiterer M, Hoe MY, Bao Z, Shea KJ. Chiral
polymers of intrinsic microporosity: Selective membrane perme-
ation of enantiomers. Angew Chem Int Ed. 2015;54:11214‐11218.
REFERENCES
19. Liu Y, Li P, Xie L, Fan DY, Huang SS. β‐cyclodextrin modified
silica nanochannel membrane for chiral separation. J Membr Sci.
2014;453:12‐17.
1. Shen J, Okamoto Y. Efficient separation of enantiomers using
stereoregular chiral polymers. Chem Rev. 2016;116:1094‐1138.
2. Yu J, Ryoo DH, Lee JM, Ryoo JJ. Synthesis and application of C2
and C3 symmetric (R)‐phenylglycinol‐derived chiral stationary
phases. Chirality. 2016;28:186‐191.
20. Singh K, Devi S, Bajaj HC, Ingole P, Choudhari J, Bhrambhatt H.
Optical resolution of racemic mixtures of amino acids through
nanofiltration membrane process. Sep Sci Technol. 2014;49:
2630‐2641.
3. Liu Y, Xuan W, Cui Y. Engineering homochiral metal‐organic
frameworks for heterogeneous asymmetric catalysis and
enantioselective separation. Adv Mater. 2010;22:4112‐4135.
21. Sueyoshi Y, Utsunomiya A, Yoshikawa M, Robertson GP, Guiver
MD. Chiral separation with molecularly imprinted polysulfone‐
aldehyde derivatized nanofiber membranes.
2012;401‐402:89‐96.
J Membr Sci.
4. Xie SM, Zhang ZJ, Wang ZY, Yuan LM. Chiral metal‐organic
frameworks for high‐resolution gas chromatographic separations.
J Am Chem Soc. 2011;133:11892‐11895.
22. Kang Z, Xue M, Fan L, et al. “single nickel source” in situ fabrica-
tion of a stable homochiral MOF membrane with chiral resolution
properties. Chem Commun. 2013;49:10569‐10571.
5. Ge L, Zhao QQ, Yang KD, Liu SS, Xia F, et al. Chirality.
2015;27:131‐136.
23. Ingole PG, Bajaj HC, Srivastava DN, Rebary B, Singh K. Prepara-
tion of thin film polymer composite membranes for optical
resolution of racemic mixture of α‐amino acids. Sep Sci Technol.
2013;48:1777‐1787.
6. Schuur B, Blahušiak M, Vitasar CR, Grambličkam M, Haan ABD,
Visser TJ. Selector screening for enantioseparation of DL‐α‐methyl
phenylglycine amide by liquid–liquid extraction. Chirality.
2015;27:123‐130.
24. Singh K, Ingole PG, Chaudhari J, Bhrambhatt H, Bhattacharya A,
Bajaj HC. Resolution of racemic mixture of ‐amino acid derivative
through composite membrane. J Membr Sci. 2011;378:531‐540.
7. Osad Y, Ohta F, Mizumoto A, Takase M, Kurimura Y. Specific per-
meation and adsorption of aqueous amino acids by plasma‐
polymerized membranes of d‐camphor and l‐menthol. J Chem Soc
Japan. 1986;7:866‐872.
25. Tian FY, Zhang JH, Duan AH, Wang BJ, Yuan LM. Chiral separa-
tion of D,L‐mandelic acid using an enantioselective membrane
formed by polycondensation of β‐cyclodextrin with 1,6‐
diisocyanatohexane on a polysulfone membrane. J Membr Sep Tech.
2012a;72‐78.
8. Lakshmi BB, Martin CR. Enantioseparation using apoenzymes
immobilized in
a
porous polymeric membrane. Nature.
1997;388:758‐760.
9. Xie R, Chu LY, Deng JG. Membranes and membrane processes for
26. Feng X, Zhang Q, Liang X, Li J, Zhao Y, Chen L. Preparation, char-
chiral resolution. Chem Soc Rev. 2008a;37:1243‐1263.
acterization and application of a chiral thermo‐sensitive membrane

10
YUAN ET AL.
for phenylalanine separation of the racemic mixture. J Polym Res.
2014;21:475‐484.
33. Xiong WW, Wang WF, Zhao L, Song Q, Yuan LM. Chiral separa-
tion of (R,S)‐2‐phenyl‐1‐propanol through glutaraldehyde‐
crosslinked chitosan membranes. J Membr Sci. 2009;328:268‐272.
27. Duerinck T, Denayer JFM. Metal‐organic frameworks as stationary
phases for chiral chromatographic and membrane separations. Chem
Eng Sci. 2015;124:179‐187.
34. Kim JH, Kim JH, Jegal J, Lee KH. Optical resolution of α‐amino
acids through enantioselective polymeric membranes based on
polysaccharides. J Membr Sci. 2003;213:273‐283.
28. Senthilkumar T, Asha SK. An easy ‘filter‐and‐separate’ method
35. Tian FY, Zhang JH, Duan AH, Wang BJ, Yuan LM. Chiral separa-
tion of D,L‐mandelic acid using an enantioselective membrane
formed by polycondensation of β‐cyclodextrin with 1,6‐
diisocyanatohexane on a polysulfone membrane. J Memb Sep Tech.
2012b;72‐78.
for enantioselective separation and chiral sensing of substrates
using
a biomimetic homochiral polymer. Chem Commun.
2015;51:8931‐8934.
29. Song Q, Jiang S, Hasell T, et al. Porous organic cage thin films and
36. 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. 2008b;321:293‐298.
molecular‐sieving membranes. Adv Mater. 2016;28(13):2629‐2637.
30. Ent EMVD, Riet KV, Keurentjes JTF, Padt AVD. Design criteria for
dense permeation‐selective membranes for enantiomer separations.
J Membr Sci. 2001;185:207‐221.
How to cite this article: Yuan L‐M, Ma W, Xu M,
et al. Optical resolution and mechanism using
enantioselective cellulose, sodium alginate and
hydroxypropyl‐β‐cyclodextrin membranes. Chirality.
2017. https://doi.org/10.1002/chir.22693
31. Paul DR, Koros WJ. Effect of partially immobilizing sorption on
permeability and the diffusion time lag. J Polym Sci Pol Phys.
1976;14:675‐683.
32. Ma C, Xu XL, Ai P, et al. Chiral separation of D,L‐mandelic acid
through cellulose membranes. Chirality. 2011;23:379‐382.