HPLC and SFC enantioseparation of (±)-Corey lactone diol: Impact of the amylose tris-(3,5-dimethylphenylcarbamate) coating amount on chiral preparation

Article abstract of DOI:10.1002/chir.23118

As an important intermediate of prostaglandins and entecavir, optically pure Corey lactone diol (CLD) has great value in the pharmaceutical industry. In this work, the enantioseparation of (±)-CLD was evaluated using high-performance liquid (HPLC) and supercritical fluid chromatography (SFC). In HPLC, the separations of CLD enantiomers on polysaccharide-based chiral stationary phases with both normal phase and polar organic phase were screened. And the conditions for the enantioseparation were optimized in HPLC and SFC, including the selection of mobile phase, temperature, back-pressure, and other conditions. More important, it was found that the chiral resolutions were greatly enhanced by the increase of the coating amount of ADMPC (amylose tris-(3,5-dimethylphenylcarbamate)) under both HPLC and SFC conditions, which can lead to the increase of the productivity and the decrease of the solvent consumption. The preparations of optically pure CLD were evaluated on a semi-preparative (2?×?25?cm) column packed with 30% ADMPC-coated CSP under HPLC and SFC conditions. Preparative performances in terms of kkd are 1.536?kg racemate/kg CSP/day and 1.248?kg racemate/kg CSP/day in HPLC and SFC, respectively.

Full text of DOI:10.1002/chir.23118

Received: 9 April 2019
DOI: 10.1002/chir.23118
Revised: 17 July 2019
Accepted: 18 July 2019
R E G U L A R A R T I C L E
HPLC and SFC enantioseparation of ( )‐Corey lactone diol:
Impact of the amylose tris‐(3,5‐dimethylphenylcarbamate)
coating amount on chiral preparation
Huiying Wang | Qiuhua Wang | Yaling Wu | Lingping Cheng | Lunan Zhu |
Junchen Zhu | Yanxiong Ke
Engineering Research Center of
Abstract
Pharmaceutical Process Chemistry, School
As an important intermediate of prostaglandins and entecavir, optically pure
Corey lactone diol (CLD) has great value in the pharmaceutical industry. In
this work, the enantioseparation of ( )‐CLD was evaluated using high‐
performance liquid (HPLC) and supercritical fluid chromatography (SFC). In
HPLC, the separations of CLD enantiomers on polysaccharide‐based chiral sta-
tionary phases with both normal phase and polar organic phase were screened.
And the conditions for the enantioseparation were optimized in HPLC and
SFC, including the selection of mobile phase, temperature, back‐pressure,
and other conditions. More important, it was found that the chiral resolutions
were greatly enhanced by the increase of the coating amount of ADMPC (amy-
lose tris‐(3,5‐dimethylphenylcarbamate)) under both HPLC and SFC condi-
tions, which can lead to the increase of the productivity and the decrease of
the solvent consumption. The preparations of optically pure CLD were evalu-
ated on a semi‐preparative (2 × 25 cm) column packed with 30% ADMPC‐
coated CSP under HPLC and SFC conditions. Preparative performances in
terms of kkd are 1.536 kg racemate/kg CSP/day and 1.248 kg racemate/kg
CSP/day in HPLC and SFC, respectively.
of Pharmacy, Ministry of Education, East
China University of Science and
Technology, Shanghai, China
Correspondence
Yanxiong Ke, Engineering Research
Center of Pharmaceutical Process
Chemistry, School of Pharmacy, Ministry
of Education, East China University of
Science and Technology, Meilong Road
130, Shanghai, China.
Email: key@ecust.edu.cn
Funding information
National Natural Science Foundation of
China, Grant/Award Number: 21375038
KEYWORDS
chiral separation, chiral stationary phase, mobile phase, polar organic mobile phase, semi‐
preparation
1 | INTRODUCTION
pregnancy,2 immune and inflammatory response,3 human
colon cancer cells,4,5 and gastrointestinal tract illnesses.6
Prostaglandins (PGs) are a class of endogenous natural
products with biological activity. PGs have the advantages
of mild toxicity to the human body and have various
physiological regulatory functions on the human body.
Therefore, PGs are widely used in the field of medicine.
PGs derivatives have succeeded in becoming an important
class of drugs, offer advantages in the treatment of idio-
pathic pulmonary hypertension,1 termination of early
In 1969, Corey first proposed that the use of Corey lac-
tone diol (CLD) in the synthesis of prostaglandin could
control its stereochemistry at an early stage, and all the
primary prostaglandins and a variety of analogs could
be synthesized from this precursor.7 The synthesis of
PGs has been greatly simplified by using (−)‐CLD. Now-
adays (−)‐CLD has become the key intermediate in the
syntheses of various types of prostanoidins industry. In
Chirality. 2019;1–10.
wileyonlinelibrary.com/journal/chir
© 2019 Wiley Periodicals, Inc.
1

2
WANG ET AL.
addition, (+)‐CLD is also an intermediate in the synthesis
of entecavir, a medicine for anti‐virus therapy of chronic
hepatitis B. According to the statistics by World Health
Organization, more than 2 billion of the world's residents
are infected with hepatitis B virus (HBV).8 Entecavir is the
most selective against HBV. Long‐term renal tolerance of
entecavir was good and side effects were minimal. An
extensive review on this topic including patent data has
been published by Slova.9 Recent reports showed that the
synthetic steps of entecavir could be largely shortened from
the CLD.10,11 The structure of CLD is shown in Figure 1.
Up to the date, many efforts have been done to prepare
optically pure CLD because of its great value in the phar-
maceutical industry. Records included asymmetric synthe-
sis,12 chiral pool synthesis,13 and chiral separation.14,15 The
asymmetric synthesis and chiral pool synthesis have been
developed to synthesize optically pure CLD.14,16 However,
these methods often suffered low overall yields because
of their requirements of multiple reaction steps in the syn-
theses.17 Crystallization of CLD derivatives and enzymatic
kinetic resolution methods were reported to separate CLD
enantiomers.18 In recent decades, direct chiral separation
using chiral stationary phases (CSPs) in high‐performance
liquid chromatography (HPLC) has become a popular and
reliable tool for enantioseparation. It has been reported
that CLD enantiomers could be separated on commercial
20 μm Chiralpak AD and Chiral AS chiral stationary
phases (CSPs).19
Polysaccharide derivatives are the most widely used
CSPs in the analysis of enantiomer compositions and
the preparation of pure enantiomers. It has been reported
that the highly ordered helical structure enhanced the
enantiorecognition abilities of the polysaccharide deriva-
tives.20 The enantioselectivities of polysaccharide‐based
CSPs not only depend on the structures of polysaccharide
derivatives but also on the coating method and coating
amount of the chiral selectors. For example, there are
reports that investigated the separation of chrysanthemate
isomers on the CSP of cellulose tris‐(4‐methylbenzoate)
(CTMB), and the separation factor for chrysanthemic acid
ethyl ester significantly depended on the preparation con-
ditions of the CSP, such as the coating amount of CTMB
and the type and amount of coating solvents.21 Another lit-
erature has shown that the amount of cellulose tris‐(3, 5‐
dimethylphenylcarbamate) (CDMPC) adsorbed on the sil-
ica gel greatly influenced the chiral recognition of some
racemates. Loading capacity of racemates increased with
an increase of the amount of CDMPC supported on the sil-
ica gel.22 These results indicate that the optimization of
coating method and coating amount of polysaccharide‐
based CSP are important for chiral separation in HPLC,
especially for the method development with the purpose
of chiral preparation, because small increases in selectiv-
ity, resolution and loading capacity will yield major
increases in productivity in large‐scale separations.23
Besides the chiral selectivity and resolution, the sample
loading capacity is another critical factor that influences
the separation efficiency in chiral preparative chromatog-
raphy, which is closely related to solubility of solute in
the mobile phase. ( )‐CLD is highly soluble in water, well
soluble in alcohol and sparely soluble in alkane. Compared
with traditional normal phase liquid chromatography
(NPLC), polar organic solvent chromatography (POSC) is
more preferred for the preparative purpose because ( )‐
CLD has a much better solubility in the solution. In recent
years, supercritical fluid chromatography (SFC) is more
and more popular in enantioseparation because of its
advantages, such as high efficiency, less organic solvent
consumption, and easy to workup. Moreover, the charac-
teristics of the low viscosity of supercritical CO2 and the
high mass transfer allow SFC to use of a more efficient
media, such as a small particle chiral stationary phase
(5 μm), with a much faster flow rate in the separation,
which can lead to the significant improvement of the sepa-
ration efficiency.
The aim of this work was to develop a more efficient
chromatographic separation method for the preparation
of optically pure CLD. Both HPLC and SFC were explored
when undertaking the method development for ( )‐CLD
enantioseparation. The type of polysaccharide‐based
CSP, mobile phase, mobile phase additive, and tempera-
ture were optimized. More importantly, the impact
of the coating amount of amylose tris‐(3,5‐
Dimethylphenylcarbamate) (ADMPC) on the separation
efficiency was also investigated. The CSP with a high coat-
ing amount of ADMPC gave a much higher selectivity and
resolution for CLD enantiomers. The chiral preparations of
( )‐CLD were performed on a semi‐preparative column
(250 × 20 mm i.d. 5 μm) under HPLC and SFC conditions.
2 | MATERIALS AND METHODS
2.1 | Chemicals
( )‐CLD and optically pure CLD were supplied by Bide
FIGURE 1 The structure of Corey lactone diol
Pharmatech Ltd (Shanghai, China). Four packed columns

WANG ET AL.
3
Chiralpak AD‐H, Chiralcel OD‐H, Chiralcel OJ‐H, and
Chiralpak AS‐H (250 × 4.6 mm i.d.) were obtained from
Daicel Chiral Technologies (Shanghai, China) Co., Ltd.
Spherical silica gel (5 μm, 1000 Å, 30 m2/g) was provided
by Acchrom Ltd (Beijing, China). Other important
chemicals of this research are listed as below: DMF and
THF were purchased from Titan Technologies Inc (Shang-
hai, China). All solvents and reagents used were of HPLC
grade: methanol, ethanol, ACN (acetonitrile), and n‐
hexane obtained from J&K Chemicals (Shanghai, China).
2.4 | Instrumentation and
chromatographic conditions
The morphologies and microstructure of ADMPC were
observed by a Hitachi S‐3400 N scanning electron micro-
scope (SEM) with 15 kV accelerating voltage. And the FT‐
IR spectra of ADMPC were recorded on a NICOLET 6700
FT‐IR spectrophotometer.
Analyses in NPLC mode were carried out by HPLC
system equipped with a 515 pump (Waters, Milford,
Massachusetts), a 7725i manual injector and a 2489 UV/
visible detector (Waters, Milford, Massachusetts). Chro-
matographic analyses in POSC mode were performed on
an Alliance HPLC system (Waters, Milford, Massachu-
setts) consisted of a 2695 pump and a 2998 UV/visible
detector. SFC separations were performed on the system
of Waters ACQUITY Ultra Performance Convergence
Chromatography (ACQUITY UPC2), which consists of a
binary solvent delivery pump, as sampler manager‐FL, a
column manager Aux, a photodiode array (PDA) detector,
and an automatic backpressure regulator (ABPR). All data
acquisition and processing were conducted by the Waters
Empower Pro 3 software.
2.2 | Packing materials
ADMPC was synthesized according to the reported
method.24 Amylose of 15 g was reacted with 52.5 g of
3,5‐dimethylphenyl isocyanate in 105 mL anhydrous
pyridine at 100°C for 20 hours. The reaction solution
was poured into 1 L methanol, and the precipitate was
collected by filtration and washed with large amount of
methanol and dried at 50°C under vacuum for 12 hours
to yield 54 g of ADMPC.
The elemental analysis data found w (N) %, w (C) %,
and w (H) % were very closed to the calculated data,
which showed that all of the hydroxyl groups in amylose
were almost converted to carbamate groups (Anal. Calcd.
for ADMPC: C, 65.66; H, 6.18; N, 6.96; found: C, 64.65; H,
6.09; N, 7.01).
The preparative POSC was carried out on a Waters Auto
purification System, which includes a 2545 binary gradient
module, a 2767 sample manager, and a 2489 UV/
visible detector. Data were collected using a Masslynx4.1
workstation.
FTIR (cm−1) 3382, 3320 (νNH), 1722 (νC=O), which also
indicates that the hydroxyl groups of the amylose have
been converted into the carbamate moieties.
The prepared ADMPC was coated on 5 μm silica gel.
The ratio of the derivative to the silica gel (w/w) was 1:10;
2:10; 3:10, and 4:10, respectively, denoted as ADMPC‐10,
ADMPC‐20, ADMPC‐30, and ADMPC‐40, respectively.
The details of preparation and characterization of
ADMPC and the packing materials were described in
the Supporting Information.
The preparative SFC was performed using an SFC
prep‐80 system, which includes a high‐pressure CO2
pump, a high‐pressure co solvent pump, a mass flow
meter, a PDA detector, an ABPR, a manual backpressure
regular, and six high‐pressure fraction collection cyclones
(Waters, Milford, Massachusetts). Data acquisition and
processing were conducted by SuperChrom software.
( )‐CLD sample was dissolved at the concentration of
10 mg/mL in mobile phase for HPLC analysis and
10 mg/mL in methanol for SFC analysis, respectively.
The injection volume was 5 μL for analyses. Column tem-
perature was controlled at 20°C in HPLC and 35°C in
SFC if there were no other specifications. Detection wave-
length was set at 220 nm.
2.3 | Column packing
1,3,5‐tri‐tert‐butylbenzene (TTB) is generally used as the
non‐retained compound for estimating the dead time (t0)
under the same condition as column efficiency determina-
tion.25 The dead times of the columns were measured by
using 1, 3, 5‐tri‐tert‐butylbenzene (TTB) with the mobile
phase of hexane/isopropanol (90/10, v/v). The measured
dead times on ADMPC‐10, ADMPC‐20, ADMPC‐30, and
ADMPC‐40 were 3.07, 2.75, 2.37, and 2.03, respectively.
These values were used for the calculation of the selectivity
of CLD enantiomers in HPLC. The dead time of the SFC
was determined by the retention time of the solvent peak.
A slurry packing procedure was employed to pack the
chromatography columns (250 × 4.6 mm i.d. for HPLC,
150 × 4.6 mm i.d. for SFC, and 250 × 20 mm i.d. for
semi‐preparation). Briefly, the stationary phase was
suspended in methanol and ultrasonicated for 15 minutes.
Methanol was employed as packing solvent. The column
was packed at 400 to 500 kg/m2 on an SP‐100 liquid
chromatography column packing machine supplied by
Dalian Sipore Co (Dalian, China) with a DSF‐100 air‐
driven hydraulic pump provided by Mpapower (Shang-
hai, China).

4
WANG ET AL.
very low in this solvent, which was impractical for prepar-
ative purposes. A high selectivity under high‐solubility
conditions would be more favorable for preparative separa-
tion. Thus, the enantioseparation of ( )‐CLD on the
Chiralpak AD‐H with polar organic mobile phase was car-
ried out in the following optimization.
3 | RESULTS AND DISCUSSION
3.1 | HPLC separation of ( )‐CLD
3.1.1 | Choice of chiral column
The various of polysaccharide‐based chiral analytical
HPLC columns (Chiralpak AD‐H, Chiralcel OD‐H,
Chiralcel OJ‐H, and Chiralpak AS‐H) and mobile phases
(a normal mobile phase consisted of n‐hexane/ethanol
(80/20, v/v) and a polar organic mobile phase consisted
of 100% ACN) were screened. The separation results are
shown in Table 1. The elution order was determined by
the injection of optically pure CLD standard sample.
No chiral resolution was observed on the Chiralcel
OD‐H and Chiralcel OJ‐H columns with either normal
mobile phase or polar organic mobile phase. The best res-
olution was achieved on the Chiralpak AD‐H column
with separation factor (α value) of 2.03 and resolution fac-
tor (RS value) of 5.17 in ACN. Normal phase also gave sat-
isfied separation result on this column with α value of
1.53 and RS value of 4.75. Under normal phase condition,
CLD enantiomers were also separated on the Chiralpak
AS‐H column with a lower separation factor and a lower
resolution factor. Only a limited resolution was achieved
on this column with a mobile phase of 100% ACN. It
should be noted that the elution order of the two enantio-
mers on the Chiralpak AD‐H was different from that
on the Chiralpak AS‐H column. (+)‐CLD was eluted first
on the Chiralpak AD‐H column, while (−)‐CLD was
eluted first on the Chiralpak AS‐H column.
3.1.2 | Influence of polar modifiers in
polar organic mobile phase
ACN is an attractive mobile phase for preparative chro-
matography because of its low viscosity. The polar modi-
fiers of methanol, ethanol, and H2O were evaluated using
ACN/modifier (97:3, v/v) mobile phases. The solubility of
( )‐CLD of in these mobile phases were measured. The
separation results and the solubility values were summa-
rized and listed in Table 1.
The sample solubility was estimated by weighting
100 mg of sample into the vials. A certain amount of
selected solvents were added into the vials and stirred at
a temperature of 25°C. The solubility of ( )‐CLD was
only about 55 mg in ACN. It was dramatically increased
to about 200 mg/mL in ACN/water (97:3, v/v). The solu-
bility in ACN/methanol (97:3, v/v) and ACN/ethanol
(97:3, v/v) solution was comparable, which was about
the half of that in ACN/H2O (97:3, v/v) solution.
The retentions of both enantiomers decreased with the
increase of modifier's polarity. The selectivity and the res-
olution were 1.96 and 3.76, respectively, with water as co‐
solvent, which were lower than those with the co‐solvent
of methanol or ethanol. However, the column efficiency
was much better when using water as a modifier. Also,
considering the good resolution, the high solubility in this
mobile phase and the convenience of the solvent recov-
ery, water was used in the preparation application.
The influence of water concentration on the separation
was examined. As shown in Figure 2, the influence of
water concentration on the selectivity was small. Both
of the retention and resolution kept increasing with the
decrease of water concentration. Thus, concentration of
water in mobile phase should be as small as possible
when the solubility of ( )‐CLD was acceptable for prepar-
ative purpose.
For preparative chromatography, the sample solubility
is a crucial issue. In this case, the satisfied separation was
obtained with the normal mobile phase on the Chiralpak
AD‐H column. However, the solubility of ( )‐CLD was
TABLE 1 The separations of Corey lactone diol (CLD) on the
chiral stationary phases (CSPs)
Solubility
Column k1
α
Rs
EO N1
MP (mg/mL)
AD‐H
7.75 1.53 4.75 (+) 2190
0.65 2.03 5.17 (+) 6558
0.62 2.35 4.95 (+) 1500
0.67 2.17 4.94 (+) 1690
0.38 1.96 3.76 (+) 5535
1.74 1.47 2.71 (−) 2422
0.36 1.37 1.03 (−) 2729
A
B
C
D
E
A
B
~55
~110
~100
~200
3.1.3 | Influence of temperature
AS‐H
The column temperature is expected to influence the
enantioselectivity in chiral separation procedures.26 The
retention factor and separation factor will associate with
the column temperature.27 In order to elucidate the ther-
modynamic characteristics of the chiral recognition pro-
cess as well as to optimize the separation selectivity, the
~55
Note. EO, elution order, absolute configuration of the first‐eluted enantio-
mer. N1, column efficiency (no. of theoretical plates, /m) of first‐eluted enan-
tiomer. MP, mobile phase, A, hexane/ethanol (80/20, v/v); B, ACN; C, ACN/
methanol (97:3, v/v); D, ACN/ethanol (97:3, v/v), E, ACN/water (97:3, v/v);
flow‐rate: 1.0 mL/min.

WANG ET AL.
5
FIGURE 2 Plots of the retention (A), separation factor and resolution factor (B), and column efficiency (C) of ( )‐CLD in the different
composition of mobile phases (acetonitrile‐water). Chromatographic conditions: Column, Chiralpak AD‐H; mobile phase, acetonitrile‐water
mixtures; flow‐rate, 1.0 mL/min
enantioseparation data were recorded on the Chiralpak
AD‐H column with stepwise raise over the range of
20°C to 55°C in 5°C increments. The obtained chromato-
graphic data were correlated to the adsorption thermody-
namic parameters by the following Van't Hoff equation:
these thermodynamic parameters were constant, which
meant that the enantioselective mechanism remained
unchanged. ΔH0 was less than 0, indicating that the
adsorption of CLD enantiomer on the stationary phase
was an exothermic process. The negative value of ΔΔH
(−3.6 0.1 kJ mol−1) indicated that the enantioseparation
of CLD was predominantly enthalpically driven. A nega-
ΔH ΔS
lnk ¼ −
þ
þ lnΦ
R
(1)
tive ΔΔS value (−8.1
0.2 J mol−1 −1) indicates that
K
RT
(−)‐CLD had less degree of freedom in binding to the chiral
selector compared with (+)‐CLD. Thus, a lower tempera-
ture would be preferred for the separation of CLD enantio-
mers since the term of –TΔΔS will cause the increase of
ΔΔG in Equation (3) at an elevated temperature. As
expected, the separation factor increased from 1.44 to
1.70 by lowering the temperature from 55°C to 20°C. The
resolution value also increased by lowering the tempera-
ture from Rs = 1.46 at 55°C to Rs = 3.76 at 20°C.
ΔΔH ΔΔS
lnα ¼ −
þ
(2)
(3)
RT
R
G ¼ H − TS;
where k represents the retention factor; R is the universal
gas constant (8.3144 J mol−1 −1); T is the absolute tem-
K
perature; ΔH and ΔS are the molar enthalpy and molar
entropy of the adsorption; Φ is the column phase ratio,
ΔΔH, ΔΔS, and ΔΔG are the differences of the molar dif-
ferential enthalpy, entropy, and Gibbs‐free energy of the
enantioselective adsorption, respectively.
As shown in Figure 3, the fitted curve of lnk versus
1/T × 10−3 and lnα versus 1/T × 10−3 possessed
high values of regression coefficients (linear correlation
coefficient R2 > .99). The linear plots indicated that
3.1.4 | Impact of ADMPC coating amount
on HPLC enantioseparation
The improvement of selectivity and resolution can lead to a
higher sample loadability, and thus increase the productiv-
ity and decrease the consumption of solvent. Up to the
day, most of the investigations mainly focus on the CSP
selection and the mobile phase optimization. It has been
FIGURE 3 Retention (A) and separation (B) factors versus 1/T of chiral separation of Corey lactone diol (CLD). Chromatographic
conditions: Column, Chiralpak AD‐H; mobile phase, ACN/water (97:3, v/v); flow‐rate, 1.0 mL/min

6
WANG ET AL.
reported that the selectivity of polysaccharide‐based
CSP is also affected by the coating amount of the polysac-
charide derivative. The enantiorecognition ability of
polysaccharide‐based CSPs not only depends on the type
of chiral selector, but also on its spatial configuration. A
higher amount coating of polysaccharide derivative on
the stationary phase may result in a more regular confor-
mation and orientation, which can lead to more chiral
recognition sites and enhanced the chiral selectivity and
resolution for some chiral compounds.
Here, four chromatographic columns (ADMPC‐10,
ADMPC‐20, ADMPC‐30, and ADMPC‐40) with the
increasing of ADMPC coating amount were prepared,
and their performances were evaluated. The column was
first tested using trans‐stilbene oxide (TSO) as a standard
with the mobile phase of hexane/isopropanol (90/10, v/
v). As shown in Figure 4, the retention and selectivity of
TSO enantiomers kept increasing with the increase of the
ADMPC coating amount. It was interesting to find that
the separation factors were even better than those of the
FIGURE 4 Effect of the coating amount of ADMPC on the separation of (A) TSO in n‐hexane/isopropanol (90/10, v/v), (B) ( )‐CLD in n‐
hexane/ethanol (80/20, v/v) and (C) ( )‐CLD in ACN/water (97/3, v/v). Flow‐rate: 1.0 mL/min; injection: 2 μL, 4 mg/mL. (D) The effect of
ADMPC coating amount on maximum loading capacity. Chromatographic conditions: column: ADMPC‐20, ADMPC‐30, ADMPC‐40; mobile
phase: ACN/water (97/3, v/v); flow‐rate: 1.0 mL/min; injection: 40 μL, 200 mg/mL. (E) The SEM images of the ADMPC‐30 and ADMPC‐40

WANG ET AL.
7
commercial AD‐H column when the coating amount was
higher than 30%. Although the selectivity on ADMPC‐40
was the highest, the resolution factor was lower than that
on ADMPC‐30 because of the lower column efficiency.
From the SEM images of the ADMPC‐30 and ADMPC‐
40, no obvious particle aggregation was found in Figure 4
E. Both of them were well‐coated. The attempt to improve
column efficiency of ADMPC‐40 by optimizing the column
packing condition in our lab failed. We suspected that the
high solid‐liquid mass transfer resistance caused by the
overcoating of ADMPC‐40 on the packing material should
be responsible for the low column efficiency.
For the separation of CLD enantiomers, the retention
and resolution of CLD enantiomers also kept increasing
with the increase of ADMPC coating amount. The best
separation factor (1.95) was achieved on ADMPC‐30,
and the best resolution factor (5.49) was obtained on
ADMPC‐40. It was interesting to find that the column
efficiency of ADMPC‐40 was comparable with the other
three columns. This finding may indicate that the mass
transfer mechanism in polar organic phase was different
from that in normal phase.
The maximum of sample loading capacities on the col-
umns were evaluated on analytical columns. The loading
quantity was constantly increased until enantiomers of
CLD on the column could not reach the baseline separa-
tion level. As shown in Figure 4D, a higher ADMPC coat-
ing amount led to a higher sample loading capacity. The
maximum loading capacity dramatically increased from
3.6 mg on ADMPC‐20 to 8 mg on ADMPC‐30. Further
increasing the ADMPC coating amount to 40% increased
the sample loading capacity to 10 mg.
We performed stable enantioseparation of ( )‐CLD
using stacking injection on ADMPC‐30 and ADMPC‐40,
respectively. Each injection interval (the time span from
the start of the first peak to the end of second peak) on
ADMPC‐30 and ADMPC‐40 were set at 3 and 4 minutes,
respectively. Thus, the estimated injection amount of ( )‐
CLD sample on ADMPC‐30 and ADMPC‐40 per hour
were 160 and 150 mg, respectively. From this point of
view, the throughput on ADMPC‐30 was slightly higher
than that on ADMPC‐40 when taking the separation time
into account.
and back pressure on the enantioseparation were also
investigated.
As shown in Table 2, the CLD enantiomers were
not baseline separated when using ethanol as modifier.
The using of 20% methanol co‐solvent offered much bet-
ter selectivity (α = 1.44) and resolution (RS = 3.95) in
SFC. The peak tailing was minimized and the separation
was finished in 3 minutes.
With the decrease of methanol concentration, the elu-
tion strength of the mobile phase decreased, resulting in
an increase of the retention time. The retention factor of
the first elution peak dramatically increased from 3.19
to 7.17 when the concentration of methanol decreased
from 20% to 10%. And the resolution increased when
the percentage of methanol decreased.
In SFC, the increase of the back pressure will increase
the density of CO2 supercritical fluid, which will lead to
the change of retention. In this experiment, the back pres-
sure had little effect on separation. Only minor changes on
retention, selectivity, and resolution were observed when
the back pressure was varied from 1.6 kPa to 2.2 kPa.
As shown in Table 2, a slight increase of k1 and a slight
decrease of α value were observed with the increase of
temperature from 293 to 328 K. The fluctuation of the
Rs value was kept in a small range. Normally, the density
of CO2 supercritical fluid at a high temperature is lower
than that at a lower temperature, which can cause the
decrease of elution strength. However, the mass transfer
will increase with the increase of temperature. Here, the
impact of column temperature on the SFC separation
retention was small.
In general, the SFC separation of CLD enantiomers is
largely influenced by the type and concentration of mod-
ifier. The influences of back pressure and temperature
on the enantioseparation are small. These characteristics
TABLE 2 The effect of different modifiers, temperature, and
back‐pressure on SFC chiral separation on ADMPC‐30
MP
BP/psi
T/K
k1
α
RS
10% methanol
15% methanol
20% methanol
20% ethanol
2000
2000
2000
2000
2000
1600
1800
2200
2000
2000
2000
308
308
308
308
308
308
308
308
298
318
328
7.17
5.57
3.19
3.87
2.76
3.40
3.29
3.09
3.22
3.13
3.08
1.45
1.44
1.44
1.10
1.22
1.44
1.44
1.44
1.44
1.43
1.42
4.82
4.38
3.95
1.11
2.08
3.99
4.02
3.91
3.88
4.06
3.92
20% isopropanol
20% methanol
20% methanol
20% methanol
20% methanol
20% methanol
20% methanol
3.2 | SFC enantioseparation of ( )‐CLD
3.2.1 | Optimization of SFC conditions
Under the SFC mode, three common organic modifiers,
including methanol, ethanol, and isopropanol, were evalu-
ated on an ADMPC‐30 column based on the above results.
And the effects of modifier concentration, temperature,

8
WANG ET AL.
FIGURE 5 The effect of ADMPC
coating amount on the SFC
enantioseparation (A) and maximum
loading capacity (B). Chromatographic
conditions: Column: ADMPC‐20,
ADMPC‐30, ADMPC‐40; mobile phase:
10% methanol in CO2; flow‐rate: 3.0 mL/
min; column temperatures: 308 K; back‐
pressure: 2000 psi
will make the enantioseparation of ( )‐CLD in SFC easy
to control.
6.0 mg of the sample was injected in ADMPC‐40. From
the chromatogram, there was still room for improvement
in the sample load. Again, these results showed that the
sample loading capacity was improved by the ADMPC
coating amount.
3.2.2 | Impact of ADMPC coating amount
on SFC separation
3.3 | Semi‐preparation
The impact of ADMPC coating amount on the chiral sep-
aration was examined in SFC. As expected, the retention
and selectivity of the CLD enantiomers increased with
increasing coating amount of ADMPC. ADMPC‐30 gave
the best resolution for ( )‐CLD.
According to the above experimental results, the semi‐
preparations of CLD enantiomers were performed on a
250 × 20 mm column packed with ADMPC‐30 packing
material in both HPLC and SFC (Figure 6).
The maximum of sample loading capacities on analyt-
ical column under SFC condition were tested on
ACQUITY UPC2. A CLD solution in methanol with con-
centration of 600 mg/mL was used for the testing. As
shown in Figure 5, the maximum loading of ADMCP‐20
and ADMCP‐30 were 2.4 mg and 4.0 mg, respectively.
Because of the limitation of sample loop equipped on this
instrument (the volume of sample loop is only 10 μL),
A “stack injection” was used to increase the productiv-
ity. The injection cycle time was set at 3 minutes in HPLC
and SFC. To separate 1 g CLD, the HPLC method took
about 20 minutes with 390 mL ACN consumption, and
the SFC method took about 23 minutes with 460 mL meth-
anol and 920 g CO2 consumption. The enantiomeric
purities of the two peaks are 98.33% and 99.41% in HPLC,
and 99.02% and 94.26% in SFC. Preparative performances
in terms of kkd (kilogram of compound purified per
kilogram of stationary phase per day) are 1.536 kg
racemate/kg CSP/day and 1.248 kg racemate/kg CSP/day
in HPLC and SFC, respectively. On the basis of the above
results, the HPLC method can provide higher productivity
and less solvent consumption. Thus, the HPLC method is a
better choice for CLD enantio‐preparation.
4 | CONCLUSION
In this paper, four coated polysaccharide‐based CSPs
were screened in HPLC and SFC modes for the
enantioseparation of CLD. The ADMPC‐coated chiral sta-
tionary phase (CSP) showed the best performance for the
enantioseparation. Polar organic mobile phase (ACN/
water = 97:3, v/v) is practical for preparative chromatog-
raphy because of the advantages of high resolution of the
enantiomers and the high solubility of ( )‐CLD. In
SFC, the modifier of methanol exhibits the best chiral
FIGURE 6 (A) The semi‐preparation on ADMPC‐30 and the
analytical purity of the prepared enantiomers; mobile phase, ACN
and water (97:3, v/v); flow‐rate, 20.0 mL/min; injection, 160 mg. (B)
the semi‐preparation on SFC and the analytical purity of the
prepared enantiomers; mobile phase, 20% MeOH; flow‐rate, 60 g/
min; injection, 130 mg

WANG ET AL.
9
10. Vostrikov NS, Lobko IF, Valiullina ZR, Miftakhov MS. Synthesis
resolution. The chiral resolution is greatly enhanced by
the increase of the coating amount of ADMPC under both
HPLC and SFC conditions. Semi‐preparations were car-
ried out under the conditions of HPLC and SFC with a
semi‐preparative column packed with 30% ADMPC
coated CSP, respectively. The CSP with high ADMPC
chiral selector would be more favorited for the enantio‐
preparation purpose, which can lead to high sample load-
ing capacity, high separation efficiency and low solvent
consuming.
of racemic entecavir. Mendeleev Commun. 2017;27(1):12‐13.
11. Valiullina ZR, Akhmet'yanova VA, Vostrikov NS, Miftakhov
MS. A short synthesis of the carbocyclic core of entecavir from
Corey lactone. Mendeleev Commun. 2016;26(1):9‐10.
12. Doyle MP, Catino AJ. A short stereoselective synthesis of (+)‐
and (−)‐2‐oxabicyclo[3.3.0]oct‐6‐en‐3‐one by intramolecular
carbon‐hydrogen insertion catalyzed by chiral dirhodium (II)
carboxamidates. Tetrahedron: Asymmetry. 2003;14(7):925‐928.
13. Kasai S, Nagasawa H, Shimamura M, Uto Y, Hori H. Design and
synthesis of antiangiogenic/heparin‐binding arginine dendrimer
mimicking the surface of endostatin. Bioorg Med Chem Lett.
2002;12(6):951‐954.
ACKNOWLEDGMENT
14. Nakashima H, Sato M, Taniguchi T, Ogasawara K. Chiral prep-
aration of polyoxygenated cyclopentanoids. Synthesis. 2000;
2000(06):817‐823.
This work was supported by the National Natural Science
Foundation of China (grant no. 21375038).
15. Nakashima H, Sato M, Taniguchi T, Ogasawara K. Lipase‐medi-
ated resolution of cis‐4‐cumyloxy‐2‐cyclopenten‐1‐ol and its
utilization for enantioconvergent preparation of (−)‐oxabicyclo
3.3.0 ‐oct‐6‐en‐3‐one. Synlett. 1999;1999(11):1754‐1756.
Declarations of interest
None.
16. Seiichi T, Tsutomu S. Method for optical resolution of corey lac-
tone diols patent US5219743 (A). 1993 July 15.
17. Yakura T. A short synthesis of optically active Corey lactone by
means of the Dirhodium (II)‐catalyzed intramolecular C‐H
insertion reaction of an α‐diazo ketone. Synthesis. 1998;
1998(07):973‐974.
ORCID
Yanxiong Ke https://orcid.org/0000-0002-5412-2869
18. Romano A, Romano D, Molinari F, Gandolfi R, Costantino F. A
new chemoenzymatic synthesis of D‐cloprostenol. Tetrahedron
Asymmetry. 2005;16(19):3279‐3282.
REFERENCES
1. Li JH, Zhang HD, Wang ZZ, et al. Acute iloprost inhalation
improves right ventricle function in pulmonary artery hyperten-
sion: a cardiac magnetic resonance study. Front Pharmacol.
2019;9:7.
19. Szanya T, H‐Veszprem, Hanak L. Process for the separation of
optical isomers of the corey lactone patent WO2007066160/A1.
2007 June 14.
2. Lievre M, Sitruk‐Ware R. Meta‐analysis of 200 or 600 mg mifep-
ristone in association with two prostaglandins for termination of
early pregnancy. Contraception. 2010;81(2):173‐174.
20. Ikai T, Okamoto Y. Structure control of polysaccharide deriva-
tives for efficient separation of enantiomers by chromatography.
Chem Rev. 2009;109(11):6077‐6101.
3. Kelly RW. Pregnancy maintenance and parturition: the role of
prostaglandin in manipulating the immune and inflammatory
response. Endocr Rev. 1994;15(5):684‐706.
21. Yamamoto C, Yamada K, Motoya K, et al. Preparation of HPLC
chiral packing materials using cellulose tris(4‐methylbenzoate)
for the separation of chrysanthemate isomers. J Polym Sci, Part
A: Polym Chem. 2010;44:5087‐5097.
4. Pai R, Nakamura T, Moon WS, Tarnawski AS. Prostaglandins
promote colon cancer cell invasion; signaling by cross‐talk
between two distinct growth factor receptors. FASEB J. 2003;
17(12):1640‐1647.
22. Yashima E, Sahavattanapong P, Okamoto Y. HPLC
enantioseparation on cellulose tris(3,5‐dimethylphenylcarbamate)
as a chiral stationary phase: influences of pore size of silica gel,
coating amount, coating solvent, and column temperature on chi-
ral discrimination. Chirality. 2015;8:446‐451.
5. Hata AN, Breyer RM. Pharmacology and signaling of prosta-
glandin receptors: multiple roles in inflammation and immune
modulation. Pharmacol Ther. 2004;103(2):147‐166.
23. Subramanian G. Chiral Separation Techniques: A Practical
Approach. Weinheim: Wiley‐VCH; 2001:199.
6. Wang D, Mann JR, Dubois RN. The role of prostaglandins and
other eicosanoids in the gastrointestinal tract. Gastroenterology.
2005;128(5):1445‐1461.
24. Chiyo Y, Eiji Y, Yoshio O. Structural analysis of amylose tris(3,5‐
dimethylphenylcarbamate) by NMR relevant to its chiral recogni-
tion mechanism in HPLC. J Am Chem Soc. 2002;124:12583‐12589.
7. Corey EJ, Weinshenker NM, Schaaf TK, Huber W. Stereo‐con-
trolled synthesis of dl‐prostaglandins F2.Alpha. And E2. J Am
Chem Soc. 1969;91(20):5675‐5677.
25. Li WZ, Wang X, Zhang WQ, et al. Enantioseparation of
novel chiral tetrahedral clusters on an amylose tris‐(3,5‐
dimethylphenylcarbamate) chiral stationary phase by normal
phase HPLC. Chin J Chem. 2004;22:1070‐1074.
8. Faseeha N. Hepatitis B virus infection: an insight into infection
outcomes and recent treatment options. Virus. 2015;26:1‐8.
9. Campian M, Putala M, Sebesta R. Bioactive carbocyclic nucleo-
side analogues–syntheses and properties of entecavir. Curr Org
Chem. 2014;18(22):2808‐2832.
26. Gao J, Qu H, Zhang C, Li W, Wang P, Zhou Z. Direct chiral
separations of the enantiomers of phenylpyrazole pesticides
and the metabolites by HPLC. Chirality. 2017;29(1):19‐25.

10
WANG ET AL.
27. Shamsipur M, Heydari R, Jamasbi ES, Abdollahpour A. Chiral
separation and quantitation of dorzolamide hydrochloride enan-
tiomers by high‐performance liquid chromatography. J Sep Sci.
2010;33(15):2328‐2333.
How to cite this article: Wang H, Wang Q, Wu Y,
et al. HPLC and SFC enantioseparation of ( )‐Corey
lactone diol: Impact of the amylose tris‐(3,5‐
dimethylphenylcarbamate) coating amount on
chiral preparation. Chirality. 2019;1–10. https://doi.
org/10.1002/chir.23118
SUPPORTING INFORMATION
Additional supporting information may be found online
in the Supporting Information section at the end of the
article.