Enantiomeric resolution, thermodynamic parameters, and modeling of clausenamidone and neoclausenamidone on polysaccharide-based chiral stationary phases

Article abstract of DOI:10.1002/chir.23068

The aim of the paper is to describe a new synthesis route to obtain synthetic optically active clausenamidone and neoclausenamidone and then use high-performance liquid chromatography (HPLC) to determine the optical purities of these isomers. In the process, we investigated the different chromatographic conditions so as to provide the best separation method. At the same time, a thermodynamic study and molecular simulations were also carried out to validate the experimental results; a brief probe into the separation mechanism was also performed. Two chiral stationary phases (CSPs) were compared with separate the enantiomers. Elution was conducted in the organic mode with n-hexane and iso-propanol (IPA) (80/20?v/v) as the mobile phases; the enantiomeric excess (ee) values of the synthetic R-clausenamidone and S-clausenamidone and R-neoclausenamidone and S- neoclausenamidone were higher than 99.9%, and the enantiomeric ratio (er) values of these isomers were 100:0. Enantioselectivity and resolution (α and Rs, respectively) levels with values ranging from 1.03 to 1.99 and from 1.54 to 17.51, respectively, were achieved. The limits of detection and quantitation were 3.6 to 12.0 and 12.0 to 40.0 ug/mL, respectively. In addition, the thermodynamics study showed that the result of the mechanism of chiral separation was enthalpically controlled at a temperature ranging from 288.15 to 308.15?K. Furthermore, docking modeling showed that the hydrogen bonds and π-π interactions were the major forces for chiral separation. The present chiral HPLC method will be used for the enantiomeric resolution of the clausenamidone derivatives.

Full text of DOI:10.1002/chir.23068

Received: 7 November 2018
DOI: 10.1002/chir.23068
Revised: 1 March 2019
Accepted: 5 March 2019
R E G U L A R A R T I C L E
Enantiomeric resolution, thermodynamic parameters, and
modeling of clausenamidone and neoclausenamidone on
polysaccharide‐based chiral stationary phases
Xuna Luo
| Chengqiao Fang | Junru Mi | Jingzi Xu | Hansen Lin
College of Pharmacy, Guangdong
Pharmaceutical University, Guangzhou
City, Guangdong Province, People's
Republic of China
Abstract
The aim of the paper is to describe a new synthesis route to obtain synthetic
optically active clausenamidone and neoclausenamidone and then use high‐
performance liquid chromatography (HPLC) to determine the optical purities
of these isomers. In the process, we investigated the different chromatographic
conditions so as to provide the best separation method. At the same time, a
thermodynamic study and molecular simulations were also carried out to val-
idate the experimental results; a brief probe into the separation mechanism
was also performed. Two chiral stationary phases (CSPs) were compared with
separate the enantiomers. Elution was conducted in the organic mode with
n‐hexane and iso‐propanol (IPA) (80/20 v/v) as the mobile phases; the
enantiomeric excess (ee) values of the synthetic R‐clausenamidone and
S‐clausenamidone and R‐neoclausenamidone and S‐ neoclausenamidone were
higher than 99.9%, and the enantiomeric ratio (er) values of these isomers were
100:0. Enantioselectivity and resolution (α and Rs, respectively) levels with
values ranging from 1.03 to 1.99 and from 1.54 to 17.51, respectively, were
achieved. The limits of detection and quantitation were 3.6 to 12.0 and 12.0 to
40.0 ug/mL, respectively. In addition, the thermodynamics study showed that
the result of the mechanism of chiral separation was enthalpically controlled
at a temperature ranging from 288.15 to 308.15 K. Furthermore, docking model-
ing showed that the hydrogen bonds and π‐π interactions were the major forces
for chiral separation. The present chiral HPLC method will be used for the
enantiomeric resolution of the clausenamidone derivatives.
Correspondence
Hansen Lin, College of Pharmacy,
Guangdong Pharmaceutical University,
No. 280, East Waihuan Street, Higher
Education Mega Center, Panyu District,
Guangzhou City, Guangdong Province,
People's Republic of China.
Email: linhansenyd@163.com
Funding information
Department of Science and Technology of
Guangdong Province, Grant/Award Num-
bers: 2015A020211031 and
2016A020217022; Natural Science Foun-
dation of Jiangsu Province, Grant/Award
Number: BK20160032
KEYWORDS
chiral chromatographic column, chiral recognition, docking, enantiomeric resolution, HPLC
1 | INTRODUCTION
characterized by improved learning and memory capabili-
ties in normal and memory‐impaired rodents,1,2 and
Clausenamide, a kind of pyrrolidone compound, is isolated
from an aqueous extract of the leaves of Clausena lansium.
Based on a pharmacological study, S‐clausenamide can
be used as a liver protectant or nootropic, which is
increased the [Ca2+]i concentration to a moderate level.3
Additionally, clausenamide can stimulate the central
cholinergic neurons,4,5 decrease apoptosis with high phos-
phorylation of the tau protein, and inhibit long‐term
Chirality. 2019;1–11.
wileyonlinelibrary.com/journal/chir
© 2019 Wiley Periodicals, Inc.
1

2
LUO ET AL.
potentiation (LTP) of the synaptic transmission effect,6
while other enantiomers have resulted in the opposite
effect without biological activity. To improve the pharma-
cokinetic properties and the dissolubility levels of
clausenamide and based on the principle of optimization
of the lead compound and the existing research, a series
of clausenamidone (CM) derivatives has been designed
for synthesis. To provide a preliminary resolution basis
for the quality control of CM and neoclausenamidone
(NM) derivatives, the determination of the optical purity
values of CM and NM was necessary. Optically pure CM
and NM have been synthesized before,7 but in our study,
we have explored a simpler and faster method of separa-
tion to obtain these compounds. However, how are the
optical purity values of these isomers to be determined?
According to a previous report,8 both notations of the
enantiomeric excess (ee)9 and enantiomeric ratio (er)10
can be found when the enantiomers are quantified by chro-
matography. On the basis of the IUPAC gold book, for a
mixture of (+) and (−)‐enantiomers, with the composi-
tions given as the mole or weight fractions of F(+) and
F(−) (where F(+) + F(−) = 1), ee is defined as
F(+) − F(−) and ee% as 100/F(+) − F(−); the er is defined
as the ratio of the percentage of one enantiomer in a mix-
ture to that of the other, eg, 70 (+):30 (−).11 In this regard,
the er is the normalized ratio (proportion) as it is directly
measured by the chromatographic peak area.
Moreover, the separation process of the chiral station-
ary phases (CSPs) under different chromatographic con-
ditions was also investigated, and the results of
separation were validated by modeling and thermody-
namics research.
2 | MATERIALS AND METHODS
2.1 | Chemicals and reagents
Racemic CM and NM were first synthesized according to
Rao EC,12 and then (−)‐menthyloxyacetic acid was used
as the chiral resolving agent to obtain CM and NM in
their enantiomerically pure forms (Figure 1).
Compared with an earlier report, although (−)‐
menthyloxyacetic acid was also used as the resolving
FIGURE 1 The synthesis of R‐
clausenamidone and S‐clausenamidone
and R‐neoclausenamidone and S‐
neoclausenamidone

LUO ET AL.
3
agent, the process used a highly toxic agent such as pyri-
dine, and the reaction lasted longer than 10 hours. We
used the classic esterification method with N,N′‐
J = 8.9 Hz, 1H), 4.92 (dd, J = 9.8, 2.9 Hz, 1H), 3.87 (t,
J = 9.4 Hz, 1H), 3.55 (d, J = 2.9 Hz, 1H), 2.89 (s, 3H);
HRMS (ESI): m/z [M + H]+, calculated: 296.1281, found:
296.1284.
dicyclohexylcarbodiimide
(DCC)
and
4‐
dimethylaminopyridine (DMAP) directly, and the process
could be completed in 1 hour. The chiral separation pro-
cess was as follows (CM is taken as example): To a solu-
tion of racemic CM in dichloromethane, DCC, DMAP,
and (−)‐menthyloxyacetic acid were added. After stirring
at room temperature for 1 hour, the reaction mixture was
filtered off from the insoluble substances. Then, the sol-
vent was removed, and the solid residue left behind was
separated by column chromatography and purified by
recrystallization with methanol. Two diastereomers as
white crystals were obtained, and hydrolysis of the two
diastereomers provided S‐clausenamidone (CS) and R‐
clausenamidone (CR) (Figure 2). The chiral separation
of NM was the same process as of that of CM to provide
S‐neoclausenamidone (NS), and R‐neoclausenamidone
(NR).
NM: white crystal, m.p. 192.1 to 194.7°C, 1H NMR
(300 MHz, chloroform‐d) δ 7.71 to 7.62 (m, 2H), 7.62 to
7.48 (m, 1H), 7.32 (td, J = 9.5, 8.5, 3.6 Hz, 5H), 7.28 to
7.15 (m, 2H), 5.10 (d, J = 6.6 Hz, 1H), 4.55 (d,
J = 6.6 Hz, 1H), 4.37 (s, 1H), 3.31 (t, J = 6.6 Hz, 1H),
2.94 (s, 3H); HRMS (ESI): m/z [M + H]+, calculated:
296.1281, found: 296.1279.
NS: white crystal, m.p. 182.8 to 185.7°C, [α]20 D − 14.0
(c = 1 g/cm3 in CH3OH); 1H NMR (300 MHz, chloroform‐
d) δ 7.72 to 7.15 (m, 10H), 5.10 (d, J = 6.5 Hz, 1H), 4.51 (d,
J = 6.6 Hz, 1H), 3.30 (t, J = 6.6 Hz, 1H), 4.14(d,
J = 6.6 Hz, 1H)2.95 (s, 3H); HRMS (ESI): m/z [M + H]
+, calculated: 296.1281, found: 296.1283.
NR: white crystal, m.p. 182.1 to 184.7°C, [α]20
1
D + 14.5 (c = 1 g/cm3 in CH3OH); H NMR (300 MHz,
chloroform‐d) δ 7.73 to 7.63 (m, 2H), 7.61 to 7.49 (m,
1H), 7.40 to 7.17 (m, 7H), 5.10 (d, J = 6.6 Hz, 1H), 4.52
(d, J = 6.6 Hz, 1H), 4.14(q, J = 6.6 Hz, 1H), 3.32 (t,
J = 6.6 Hz, 1H), 2.96 (s, 3H); HRMS (ESI): m/z
[M + H]+, calculated: 296.1281, found: 296.1276.
2.2 | Characterization data
CM: white crystal, melting point (m.p.) 208.1 to 209.7°C,
1H NMR (500 MHz, chloroform‐d) δ 7.56 to 7.50 (m,
2H), 7.41 (ddt, J = 8.7, 7.3, 1.3 Hz, 1H), 7.30 to 7.20 (m,
2H), 7.14 to 7.09 (m, 2H), 7.09 to 7.02 (m, 2H), 7.04 to
6.97 (m, 1H), 5.40 (d, J = 8.9 Hz, 1H), 4.95 to 4.89 (m,
1H), 3.87 (t, J = 9.4 Hz, 1H), 3.55 (d, J = 2.7 Hz, 1H),
2.89 (d, J = 0.6 Hz, 3H); HRMS (ESI): m/z [M + H]+, cal-
culated: 296.1281, found: 296.1283.
CS: white crystal, m.p. 196.6 to 198.4°C, [α]20
D − 340.2 (c = 0.5 g/cm3 in CH3OH), 1H NMR
(500 MHz, chloroform‐d) δ 7.56 to 7.50 (m, 2H), 7.41
(ddt, J = 8.7, 7.3, 1.3 Hz, 1H), 7.30 to 7.20 (m, 2H), 7.14
to 7.09 (m, 2H), 7.09 to 7.02 (m, 2H), 7.04 to 6.97 (m,
1H), 5.40 (d, J = 8.9 Hz, 1H), 4.95 to 4.89 (m, 1H), 3.87
(t, J = 9.4 Hz, 1H), 3.55 (d, J = 2.7 Hz, 1H), 2.89 (s,
3H); HRMS (ESI): m/z [M + H]+, calculated: 296.1281,
found: 296.1283.
2.3 | Chromatographic conditions
The enantiomeric separations were conducted on a high‐
performance liquid chromatography (HPLC) system
(model 1200 series, Agilent Technology) equipped with
a quaternary pump, online degasser, autosampler, col-
umn oven, and diode array detector (DAD). Two CSPs,
CHIRALPAK AD‐H based on amylose derivatized with
tris‐(3,5‐dimethylphenyl carbamate) (25 cm × 4.6 mm I.
D., 5 μm) and CHIRALCEL OJ‐H based on cellulose
derivatized with tris‐(4‐methylbenzoate) (25 cm × 4.6 mm
I.D., 5 μm), were evaluated in this study. These chiral col-
umns were purchased from Daicel Chiral Technologies
(China) Co. Ltd. The mobile phase was a mixture of n‐
hexane and iso‐propanol (IPA). By using 1.0 mL/min as
the flow rate, the injection volume was 20 μL, and the
detection wavelength was 254 nm. Chromatography vari-
ables such as the retention (k), separation (α), and resolu-
tion (Rs) factors were determined. 1,3,5‐Tri‐tert‐
butylbenzene was used to calculate the dead time. The
CR: white crystal, m.p. 198.9 to 200.5°C, [α]20
D + 312.5 (c = 0.5 g/cm3 in CH3OH), 1H NMR
(500 MHz, chloroform‐d) δ 7.56 to 7.50 (m, 2H), 7.44 to
7.37 (m, 1H), 7.27 to 7.20 (m, H), 7.15 to 7.09 (m, 2H),
7.09 to 7.03 (m, 2H), 7.03 to 6.97 (m, 1H), 5.40 (d,
FIGURE 2 The structures of R‐
clausenamidone and S‐clausenamidone
and R‐neoclausenamidone and S‐
neoclausenamidone

4
LUO ET AL.
determinations of the optical purities of CS, CR, NS, and
NR were conducted using the best separation method
which was based on the investigation of the different
chromatographic conditions to provide the ee% and er
values, where the computational methods were as fol-
lows: for example, CR (ee%) = 100/F(+) − F(−), where
F(+) represents the peak area % of CR and F(−) repre-
sents the peak area % of CS. The CM er value represents
the ratio of the peak area % of CR and the peak area % of
CS, that is F(+):F(−).
and NM; then, these two structures were converted with
Chem3D to obtain the minimal energy structures. After
that, Autodock4.2 (ADT) was used to process the struc-
tures of the receptors and ligands, which were then saved
in PDBQT files. A lattice box size of 60 Å × 60 Å × 60 Å
with a spacing of 0.371 Å was used. After docking, the
output files were opened in Discovery Studio 4.5 to obtain
the image of molecular docking interaction.
3 | RESULTS AND DISCUSSION
3.1 | Chromatographic data
2.4 | Method validation
According to a previous report, the validation of the pro-
posed method was characterized by the following param-
eters: specificity, accuracy, precision, limit of detection
(LOD) and limit of quantification (LOQ), and linearity.13
The HPLC measurements of the retention (k), separation
(α), and resolution (Rs) were ascertained for the enantio-
mers of CM and NM on amylose derivatized with
tris‐(3,5‐dimethylphenyl carbamate) and cellulose
derivatized with tris‐(4‐methylbenzoate) CSPs. The mag-
nitudes of the racemic CM and NM contents are given in
the Table 1, which indicated that the retention (k) and res-
olution (Rs) would decrease with increasing fraction of
IPA. It could be concluded that interaction between the
sample and stationary phase weakened as the IPA content
was increased. The order of separated peaks of HPLC is
shown in Figure 4 with the 80/20 (v/v) n‐hexane/IPA
mobile phase at a temperature of 298.15 K.
2.5 | Simulation studies
With the ever‐growing power of computers, software, and
computer graphics (to display structures along with their
physicochemical properties), molecular modeling has
become a practical tool for evaluating more complex
interactions such as those of the associations between
chiral selectors and selectants.14 Computer modeling
was conducted with Autodock4.2. The ligands used were
CM and NM, while the receptors were the amylose and
cellulose stationary phases with 12 units15 (Figure 3).
ChemDraw was applied to sketch the structures of CM
3.2 | Investigation of the thermodynamic
parameters
Thermodynamic experiments were carried out in the
range of 288.15 to 308.15 K, and the results are given in
Table 2 and Figure 5. In the chromatographic separation
of the enantiomers, the Gibbs free energy change ΔGθ
was measured, which represents the binding energy
between the CSPs and the samples. ΔGθ can be calculated
at a constant temperature according to the Gibbs‐
Helmholtz equation 1. In addition, the relation between
the chromatographic retention and temperature can be
described by the Van't Hoff equations 2 and 3, where
ΔHθ and ΔSθ represent the change in the standard molar
enthalpy and standard molar entropy, respectively, ΔΔHθ
and ΔΔSθ represent the difference in the standard molar
enthalpy change and in the standard molar entropy
change between the stationary phase and mobile phase,
respectively, R is the universal gas constant (8.3144
J/[mol·K]), and T is the absolute temperature (K). The
enantiomeric bonding is greatly driven by intermolecular
interactions as measured by the enthalpy change ΔHθ.
The process of complexation is usually reflected by the
entropic cost ΔSθ.16 The thermodynamic values were cal-
culated using Origin2018.
FIGURE 3 A, 2D structure of the tris‐(4‐methylbenzoate)
cellulose chiral selector. B, 2D structure of the tris‐(3,5‐
dimethylphenyl carbamate) amylose chiral selector

LUO ET AL.
5
TABLE 1 Results of separation of racemic clausenamidone and neoclausenamidone on the Chiralcel OJ‐H and Chiralpak AD‐H columns
Compound
Column (CSP)
Mobile Phase (v/v)
κ1
κ2
α
Rs
Clausenamidone
Chiralcel OJ‐H
n‐hexane/IPA 90/10
n‐hexane/IPA 85/15
n‐hexane/IPA 80/20
n‐hexane/IPA 90/10
n‐hexane/IPA 85/15
n‐hexane/IPA 80/20
2.86
1.63
1.03
11.99
6.20
4.10
5.70
3.15
1.97
18.66
10.00
6.80
1.99
1.93
1.91
1.56
1.61
1.66
13.58
11.53
9.01
16.76
16.26
16.06
Chiralpak AD‐H
Neoclausenamidone
Chiralcel OJ‐H
n‐hexane/IPA 90/10
n‐hexane/IPA 85/15
n‐hexane/IPA 80/20
n‐hexane/IPA 90/10
n‐hexane/IPA 85/15
7.77
4.30
2.93
11.74
6.44
13.04
6.83
4.04
12.27
6.63
1.68
1.59
1.38
1.05
1.03
14.57
11.56
7.08
1.54
0.95
Chiralpak AD‐H
Abbreviation: IPA, iso‐propanol.
The temperature of the column was at 298.15 K.
FIGURE 4 Chiral resolutions of clausenamidone and neoclausenamidone on the Chiralcel OJ‐H and Chiralpak AD‐H columns using n‐
hexane/IPA (80/20 v/v) as the mobile phase at 298.15 K on Chiralcel OJ‐H and n‐hexane/IPA (90:10 v/v) as the mobile phase at 298.15 K on
Chiralpak AD‐H
TABLE 2 Effect of the column temperatures of Chiralcel OJ‐H and Chiralpak AD‐H on clausenamidone
Clausenamidone
Neoclausenamidone
κ1 κ2
4.76 7.57
Column (CSP)
Temperature, K
κ1
κ2
α
α
Chiralcel OJ‐H
288.15
293.15
298.15
300.15
303.15
308.15
1.27
1.18
1.07
1.09
1.04
0.97
2.50
1.96
1.94
1.92
1.91
1.90
1.88
1.59
1.53
1.46
1.47
1.42
1.38
2.29
2.06
2.08
1.97
1.82
4.04
3.69
3.49
3.12
2.93
6.21
5.40
5.13
4.43
4.04
Chiralpak AD‐H
288.15
293.15
298.15
301.15
303.15
308.15
4.93
4.79
4.42
4.17
4.17
4.10
10.14
9.30
8.19
7.48
7.33
6.80
2.06
1.94
1.85
1.80
1.76
1.66
15.37
14.07
12.77
11.99
11.74
10.75
16.23
14.83
13.41
12.52
12.27
11.16
1.06
1.05
1.05
1.04
1.05
1.04
aThe mobile phase was 80/20 (v/v) n‐hexane/IPA.

6
LUO ET AL.
FIGURE 5 Effect of the column temperature on clausenamidone and neoclausenamidone
TABLE 3 Thermodynamic parameters of clausenamidone (at
298.15 K)
ΔGθ ¼ ΔHθ − TΔSθ
ΔHθ
(1)
(2)
(3)
Parameter
R
S
R
S
lnK ¼ −
þ ΔSθ=R
ΔΔHθ
RT
Column (CSP)
Slope
Chiralcel OJ‐H
Chiralpak AD‐H
1299.3
−4.26
1496.5
−4.27
0.97
937.1
1868.8
−4.17
0.99
lnα ¼ −ΔΔGθ=RT ¼
þ ΔΔSθ=R
RT
Intercept
R2
−1.66
0.94
0.96
As shown in the Table 2, the retention times and
retention (k) and separation (α) factors of the two enan-
tiomers of these two compounds exhibited downward
trends from 288.15 to 308.15 K, indicating that the inter-
action between the sample and stationary phase weak-
ened as the temperature was increased, which was
related to the Van't Hoff equation 2. The temperature
was raised to facilitate the processes of adsorption and
desorption between the solute and the stationary phase.
The separation of the chromatogram parameters was a
reflection of the reduction in the capacity factor k and
the separation factor α. Thus, the relation between the
chromatographic retention and column temperature
could be described by Van't Hoff plots to obtain the linear
plots of lnk versus 1/T and lnα versus 1/T; the date of the
latter linear plots is given in Tables 3 and 4. It was
ΔΗθ
ΔSθ
ΔGθ
ΔΔHθ(kJ/mol)
ΔΔSθ
−10.80
−0.04
−12.44
−0.04
−1.87
−7.79
−0.01
−3.69
−7.75
−0.0209
−1.52
370.66
−15.54
−0.03
−6.60
−0.25
−1.64
−0.0001
−1.62
ΔΔGθ(kJ/mol)
Tiso (K)
22930.23
observed that the linear plots of lnk versus 1/T and lnα
versus 1/T had regression coefficients R2 ≥ 0.90 for both
compounds. From the Van't Hoff plots, the values of
ΔHθ,ΔSθ,ΔΔHθ, ΔΔSθ, and ΔΔGθ were thus calculated,
and the thermodynamic parameters for the chiral separa-
tions are provided in Tables 3 and 4.

LUO ET AL.
7
TABLE 4 Thermodynamic parameters of neoclausenamidone (at
298.15 K)
Chiralcel OJ‐H or below 365 K on Chiralpak AD‐H, the
value of separation (α) would decrease as the temperature
was increased. Given that a high temperature would
damage the column, the verification experiments were
not carried out.
Parameter
R
S
R
S
Column (CSP)
Slope
Chiralcel OJ‐H
Chiralpak AD‐H
In addition, as reported19 before, when the absolute
value of ΔΔHθ was lower than 0.1 KJ/mol, the chiral rec-
ognition was only attributed to the steric hindrance and
did not include other types of forces. In the range of 0.5
to 1.0 kJ/mol, the contribution of the steric hindrance
was amplified by weak interactions such as the weak
π‐π action, weak hydrogen bonding action, and the like.
When the absolute value of ΔΔHθ was higher than 1.0
kJ/mol, it was considered that the postpeak isomer was
subjected to an additional strong π‐π action or strong
hydrogen bonding, which showed somewhat strong chi-
ral recognition levels. For CM, the value obtained by
Chiralcel OJ‐H was ΔΔHθ= −1.64 kJ/mol and that by
Chiralpak AD‐H was ΔΔHθ = −7.75 kJ/mol, which
reflected strong π‐π interactions or hydrogen bonding
with the CSPs resulting in chiral separation. For NM,
the value obtained by Chiralcel OJ‐H was ΔΔHθ= −5.34
kJ/mol, which indicated strong π‐π interactions or hydro-
gen bonding with the CSPs resulting in chiral separation.
However, on Chiralpak AD‐H, the value was ΔΔHθ
= −0.63 kJ/mol, which was in the range of 0.5 to 1.0
kJ/mol, thereby indicating that the contribution of steric
hindrance was amplified by weak interactions such as
the weak π‐π action, weak hydrogen bonding action,
and the like.
2821.1
−7.79
0.99
2179.4
−6.02
0.98
1675.2
−3.03
0.99
1599.6
Intercept
R2
−2.82
0.99
ΔΗθ
ΔSθ
ΔGθ
ΔΔHθ(kJ/mol)
ΔΔSθ
−23.46
−0.07
−4.16
−5.34
−0.015
−0.96
363.94
−18.12
−0.05
−3.19
−13.93
−0.03
−6.42
−0.63
−0.002
−0.12
365.75
−13.30
−0.02
−6.31
ΔΔGθ(kJ/mol)
Tiso (K)
using n‐hexane/IPA (80/20 v/v) as the optimum
mobile‐phase composition
The ΔΔHθ and ΔΔSθ data in Tables 3 and 4 both
exhibited negative values, which indicated that the chi-
ral separation of CM and NM was somewhat
enthalpically controlled under the chromatographic con-
ditions. When ΔΔHθ and ΔΔSθ were both negative, the
second‐eluted enantiomer fit much better in the cavity
of the selector, thereby forming a more stable complex
than in the case of the first‐eluted enantiomer; the neg-
ative entropy was less favorable for enantioseparation.
The enantioseparation was enthalpically driven, as was
common, and the selectivity decreased with increasing
temperature.17
Thus, it was concluded that when the temperature
was below 363 K, the enantioseparation of these samples
was enthalpically driven, and according to the thermody-
namic parameters, the π‐π and hydrogen bonding
actions were the major interactions between the samples
and the CSPs.
Moreover, according to Equation 2, when the value of
separation (α) was 1, the corresponding temperature Tiso
was obtained, which meant that Tiso=ΔΔHθ/ΔΔSθ; the
contributions of the enthalpy and entropy in the chiral
separation process were offset. When the column temper-
ature was below Tiso, the separation would decrease as
the temperature was increased. When the column tem-
perature was Tiso, the isomers would be eluted simulta-
neously. However, if the column temperature was
higher than Tiso, the elution order of the enantiomers
would be reversed, and the mechanism of chiral separa-
tion would be entropically controlled.18
3.3 | Method validation
3.3.1 | System suitability
HPLC suitability testing was performed using six repli-
cate injections (n = 6) of the standard solution
(1000 μg/mL) and the tailing factor, % relative standard
deviation (RSD) of the peak area, and % RSD of the reten-
tion time were deemed suitable as shown in Table 5.
For CM, the values of Tiso = ΔΔHθ/ΔΔSθ = 22 930 K
on Chiralcel OJ‐H and Tiso = ΔΔHθ/ΔΔSθ = 371 K on
Chiralpak AD‐H showed that if the temperatures were
below 22 930 K on Chiralcel OJ‐H or below 371 K on
Chiralpak AD‐H, the value of separation (α) would
decrease as the temperature increased. For NM, the
values of Tiso = ΔΔHθ/ΔΔSθ = 363 K on Chiralcel OJ‐
H and Tiso = ΔΔHθ/ΔΔSθ = 365 K on Chiralpak AD‐H
showed that if the temperatures were below 363 K on
3.3.2 | Specificity
Specificity of the method was evaluated for any interfer-
ence by other peaks along with the main peaks. The elu-
ent was loaded onto the machine as a blank. No peak was

8
LUO ET AL.
TABLE 5 The results of the method validation
Method Validation
CR
0.59
CS
0.55
5.01
0.07
0.27
NS
0.55
5.64
0.05
0.33
NR
System suitability
Tailing factor
Resolution
% RSD peak area
% RSD retention time
0.59
3.79
0.14
0.53
‐
0.90
0.15
Linearity
Slope
40402
698.13
0.9996
38502
465.05
0.9992
33167
394.97
0.9992
31621
326.26
0.9993
Y‐intercept
Regression coeff. (R2)
LOD and LOQ
LOD (μg/mL)
LOQ (μg/mL)
3.6
12.0
4.8
16.0
6.0
20.0
12.0
40.0
Accuracy (percentage
recovery data)
CR (300 μg/mL) CS,
NS, NR (350 μg/mL)
1.04
1.03
1.02
1.04
1.03
1.03
1.00
0.99
1.00
1.01
1.01
1.01
1.01
1.02
0.99
0.99
0.99
0.99
0.99
1.00
1.00
1.00
0.99
0.99
0.99
1.00
1.00
0.99
0.98
0.98
1.00
1.00
1.00
1.01
1.01
0.99
CR (330 μg/mL) CS,
NS, NR (440 μg/mL)
CR (450 μg/mL) CS,
NS, NR (570 μg/mL)
Precision (% RSD)
120 μg/mL
240 μg/mL
400 μg/mL
Aa
Ta
A
T
A
T
0.07
0.06
0.93
0.10
0.62
0.07
0.37
0.54
0.60
0.07
0.31
0.08
0.62
0.10
0.19
0.11
0.99
0.06
0.38
0.71
0.40
0.10
0.25
0.16
aA represents the peak area; T represents the retention time.
observed from the blank solution. The latter indicated the
specificity of the system.
3.3.5 | Accuracy
The accuracy was determined by calculating the recovery
from the spike amounts, and the percent recoveries were
found to be in the range of 97.88% to 103.94% for all four
enantiomers at three different spiked concentrations
(300‐570 μg/mL) as shown in Table 5. This array indi-
cated a high accuracy of the method.
3.3.3 | Linearity
Linearity was evaluated by regression analysis of the
curve. The linearity values of the standardized peak areas
vs concentrations for all four enantiomers were tested in
the 80 to 400 μg/mL concentration range. The slopes, y‐
intercepts, and regression coefficients (R2) were calcu-
lated These values are given in Table 5. The regression
coefficients of the establishment were 0.9992 to 0.9996
for the four enantiomers (n = 6).
3.3.6 | Precision
The precision was determined by three concentrations of
all of the enantiomers (120, 240, and 400 μg/mL) (n = 3).
The % RSDs of the retention times and peak area were
0.06 to 0.71 and 0.07 to 0.62, respectively, which indicated
a suitably precise HPLC method.
3.3.4 | LOD and LOQ
The LOD and LOQ values were calculated as signal‐to‐
noise ratios of 3 and 10, respectively. The LODs of CR,
CS, NS, and NR were 3.6, 4.8, 6.0, and 12 μg/mL, respec-
tively. The LOQs of CR, CS, NS, and NR were 12, 16, 20,
and 40 μg/mL, respectively. The % RSD values of the peak
areas for the LOD were 0.66 to 0.99. Similarly, the % RSD
of the peak areas for the LOQ were 0.62 to 0.84.
3.4 | Determination of the optical purity
It could be concluded that the best separation condition
of racemic CM and NM was n‐hexane/IPA (80/20 v/v)
at 298.15 K using Chiralcel OJ‐H according to Sections
3.1 and 3.2. Based on this condition, racemic CM and
NM were used as the reference objects to determine the

LUO ET AL.
9
optical purities of CS, CR, NS, and NR. The results are
shown in Figure 6, which indicated that the ee and the
er values for the synthetic CM and NM were higher than
99.9% and 100:0, respectively.
3.5 | Simulation studies
The earlier reports20-26 clearly were in support of chiral
recognition being controlled by hydrogen bonding, and
hydrophobic, π‐π, steric interactions, etc. For example,
the interactions between the π‐π stacked and hydrogen
bonds are clearly shown between S‐clausenamidone and
CSP with tris‐(4‐methylbenzoate) cellulose in Figure 7.
In Section 3.1 it was concluded that the interaction
between the sample and stationary phase was strength-
ened as the ratio of IPA decreased, which corresponded
to the docking result that the separation was controlled
by the hydrogen bonds. Iso‐propanol functioned as the
proton donor providing hydrogen atoms to form hydro-
gen bonds with the hydrogen atom of sample. If the
IPA in the mobile phase was reduced, the hydrogen bond
interaction between IPA and enantiomer would be rela-
tively weakened, and the hydrogen bond force between
the CSP and the enantiomer would be strengthened given
the long retention time.
From the comparison between the different columns,
CM and NM had longer retention times in Chiralpak AD‐
H with tris‐(3,5‐dimethylphenyl carbamate) amylose than
those in Chiralcel OJ‐H with tris‐(4‐methylbenzoate) cellu-
lose. These results corresponded to the report27 in which
the phenylcarbamate moiety attached to the amylose
backbone exhibited a rather high enantiomer resolving
ability. It could be explained by the structure of Chiralpak
FIGURE 6 Chiral resolutions of racemic clausenamidone and
neoclausenamidone and enantiomers on Chiralcel OJ‐H using n‐
hexane/IPA (80/20 v/v) as the mobile‐phase composition at
298.15 K. a. % peak area
FIGURE 7 Docking model of S‐
clausenamidone with the tris‐(4‐
methylbenzoate) cellulose chiral selector

10
LUO ET AL.
AD‐H, which has been reported to be a left‐handed four-
fold (4/l) helix on the basis of X‐ray analysis.28 The carba-
mate residue formed a large number of grooves around
the main chain, and the interaction between the sample
and the stationary phase included ammonia hydrogen
bond interactions (or dipole‐dipole interactions),
clathrations, and π‐π interactions. The strongest chiral
adsorption positions of Chiralcel OJ‐H were only the car-
bonyl group, which had formed hydrogen bond interac-
tions, and the phenyl group through π‐π interactions
with aromatic compounds to complete the chiral recogni-
tion. Hence, the interaction between the compound and
Chiralpak AD‐H was stronger than that between the com-
pound and Chiralcel OJ‐H. Moreover, the elution time in
Chiralpak AD‐H was longer than that in Chiralcel OJ‐H.
ORCID
Xuna Luo https://orcid.org/0000-0003-3115-0258
REFERENCES AND NOTES
1. Zhang J, Cheng Y, Zhang JT. Protective effect of (−)
clausenamide against neurotoxicity induced by okadaic acid
and beta‐amyloid peptide25–35[J]. Yao xue xue bao Acta
pharmaceut Sinica. 2007;42(9):935‐942.
2. Tang K, Zhang JT. (−) clausenamide improves long‐term poten-
tiation impairment and attenuates apoptosis after transient
middle cerebral artery occlusion in rats[J]. Neurol Res.
2003;25(7):713‐717.
3. Duan W, Zhang J. Effects of (−),(+) clausenamide on anisodine‐
induced acetylcholine decrease and associated memory deficits
in the mouse brain[J]. Yao xue xue bao Acta pharmaceut Sinica.
1998;33(4):259‐263.
4. Bayer AJ, Bullock R, Jones RW, et al. Evaluation of the safety
and immunogenicity of synthetic Aβ42 (AN1792) in patients
with AD[J]. Neurology. 2005;64(1):94‐101.
4 | CONCLUSION
5. Selkoe DJ. Alzheimer's disease is a synaptic failure[J]. Science.
2002;298(5594):789‐791.
This study describes the synthesis of optically active CM
and NM and discusses their enantioseparations using two
different polysaccharide‐based CSPs, namely, Chiralcel
OJ‐H and Chiralpak AD‐H. The best separation condition
was n‐hexane/IPA (80/20 v/v) at 298.15 K using Chiralcel
OJ‐H; the enantioselectivity and resolution (α and Rs,
respectively) values were 1.90 and 1.38, and 9.06 and
7.08, respectively. Based on this condition, the ee values
were higher than 99.9% and the er values were 100:0 for
all synthetic optically active CM and NM. Meanwhile, all
of the isomers were separated successfully within
25 minutes, and all validation parameters showed accept-
able results; hence, the developed method was efficiently
used for the quantitative determination of the enantio-
meric purities of the optically active CM and NM. In addi-
tion, according to the results of the thermodynamic
parameters, it could be concluded that the separation of
racemic CM and NM was somewhat enthalpically con-
trolled. At the same time, the results showed that the π‐π
interactions and hydrogen bonding with the CSPs resulted
in chiral separation, which corresponded to the results of
docking modeling. In all, the described method was suc-
cessful for evaluating the enantiomers of CM and NM
and is of great reference significance for future research
on CM and NM and their derivatives.
6. Li XZ, Lai K, Wu KM, Huang DF, Huang L. Studies on the syn-
thesis and biological evaluation of the metabolite of
clausenamide CM2 and its stereoisomers[J]. Eur J Med Chem.
2014;74:736‐741.
7. Hansen L. Resolution of neoclausenamidone[J]. Acad J Guang-
dong Coll Pharma. 1999;2.
8. Tiritan ME, Fernandes C, Maia AS, Pinto M, Cass QB. Enantio-
meric ratios: why so many notations?[J]. J Chromatogr A.
2018;1569:1‐7.
9. Han J, Kitagawa O, Wzorek A, Klika KD, Soloshonok VA. The
self‐disproportionation of enantiomers (SDE): a menace or an
opportunity?[J]. Chem Sci. 2018;9(7):1718‐1739.
10. Sriyatep T, Tantapakul C, Andersen RJ, et al. Resolution and
identification of scalemic caged xanthones from the leaf extract
of Garcinia propinqua having potent cytotoxicities against colon
cancer cells[J]. Fitoterapia. 2018;124:34‐41.
11. Book G. Compendium of chemical terminology[J]. Int Union
Pure Appl Chem. 2014;528.
12. Rao EC, Hong H, Cheng JC, Yang GZ. Total synthesis of
clausenamide[J]. Chin Chem Lett. 1994;5:267‐267.
13. Ali I, Khattab RA, Alothman ZA, Badjah AY, Alwarthan A.
Enantiomeric resolution and modeling of DL‐alanine‐DL‐
tryptophan dipeptide on amylose stationary phase[J]. Chirality.
2018;30(4):491‐497.
14. Lämmerhofer M. Chiral recognition by enantioselective liquid
chromatography: mechanisms and modern chiral stationary
phases[J]. J Chromatogr A. 2010;1217(6):814‐856.
15. Ye YK, Bai S, Vyas S, Wirth MJ. NMR and computational
studies of chiral discrimination by amylose tris (3, 5‐
dimethylphenylcarbamate)[J]. J Phys Chem B. 2007;111(5):
1189‐1198.
ACKNOWLEDGMENTS
This study was supported by the Department of
Science and Technology of Guangdong Province
(2015A020211031 and 2016A020217022) and the Natural
Science Foundation of Jiangsu Province (BK20160032).
16. Wani DV, Rane VP, Mokale SN. Liquid chromatographic sepa-
ration and thermodynamic investigation of lorcaserin

LUO ET AL.
11
hydrochloride enantiomers on immobilized amylose–based chi-
macrocyclic glycopeptide antibiotic chiral stationary phases[J].
ral stationary phase[J]. Chirality. 2018;30(3):284‐292.
Il Farmaco. 2002;57(7):513‐529.
17. Lajkó G, Orosz T, Kiss L, et al. High‐performance liquid chro-
matographic enantioseparation of fluorinated cyclic β3‐amino
acid derivatives on polysaccharide‐based chiral stationary
phases. Comparison with nonfluorinated counterparts[J].
Biomed Chromatogr. 2016;30(9):1441‐1448.
25. Aboul‐Enein HY, Ali I. Studies on the effect of alcohols on the
chiral discrimination mechanisms of amylose stationary phase
on the enantioseparation of nebivolol by HPLC[J]. J Biochem
Biophys Methods. 2001;48(2):175‐188.
26. Ali I, Sanagi MM, Aboul‐Enein HY. Advances in chiral
separations by nonaqueous capillary electrophoresis in
pharmaceutical and biomedical analysis[J]. Electrophoresis.
2014;35(7):926‐936.
18. Stringham RW, Blackwell JA. Factors that control successful
entropically driven chiral separations in SFC and HPLC[J]. Anal
Chem. 1997;69(7):1414‐1420.
19. Küsters E, Spöndlin C. Influence of temperature on the
enantioseparation of rolipram and structurally related racemates
on Chiracel‐OD[J]. J Chromatogr A. 1996;737(2):333‐337.
27. Chankvetadze B, Kartozia I, Yamamoto C, Okamoto Y. Compar-
ative enantioseparation of selected chiral drugs on four different
polysaccharide‐type chiral stationary phases using polar organic
mobile phases[J]. J Pharm Biomed Anal. 2002;27(3–4):467‐478.
20. Morris GM, Goodsell DS, Halliday RS, et al. Automated docking
using a Lamarckian genetic algorithm and an empirical binding
free energy function[J]. J Comput Chem. 1998;19(14):1639‐1662.
28. Okamoto Y, Kaida Y. Resolution by high‐performance
liquid chromatography using polysaccharide carbamates and
benzoates as chiral stationary phases[J]. J Chromatogr A.
1994;666(1–2):403‐419.
21. Ali I, Alam SD, Al‐Othman ZA, Farooqi JA. Recent advances
in SPE–chiral‐HPLC methods for enantiomeric separation of
chiral drugs in biological samples[J].
J Chromatogr Sci.
2013;51(7):645‐654.
22. Ali I, Saleem K, Hussain I, Gaitonde VD, Aboul‐Enein HY. Poly-
saccharides chiral stationary phases in liquid
chromatography[J]. Sep Purif Rev. 2009;38(2):97‐147.
How to cite this article: Luo X, Fang C, Mi J, Xu
J, Lin H. Enantiomeric resolution, thermodynamic
parameters, and modeling of clausenamidone and
neoclausenamidone on polysaccharide‐based chiral
stationary phases. Chirality. 2019;1–11. https://doi.
org/10.1002/chir.23068
23. Aboul‐Enein HY, Ali I. Applications of Polysaccharide‐Based
Chiral Stationary Phases for Resolution of Different Compound
Classes[M]//chiral separations. Humana Press; 2004:183‐196.
24. Aboul‐Enein HY, Ali I. Optimization strategies for HPLC
enantioseparation of racemic drugs using polysaccharides and