Synthesis of new C3 symmetric amino acid- and aminoalcohol-containing chiral stationary phases and application to HPLC enantioseparations

Article abstract of DOI:10.1002/chir.22766

We recently reported a new C3-symmetric (R)-phenylglycinol N-1,3,5-benzenetricarboxylic acid-derived chiral high-performance liquid chromatography (HPLC) stationary phase (CSP 1) that demonstrated better results as compared to a previously described N-3,5-dintrobenzoyl (DNB) (R)-phenylglycinol-derived CSP. Over a decade ago, (S)-leucinol, (R)-phenylglycine, and (S)-leucine derivatives were used as the starting materials of 3,5-DNB-based Pirkle-type CSPs for chiral separation. In this study, three new C3-symmetric CSPs (CSP 2, 3, and 4) were prepared by combining the ideas and results mentioned above. Here we describe the synthetic procedures and applications of the new C3-symmetric CSPs (CSP 2–CSP 4).

Full text of DOI:10.1002/chir.22766

Received: 13 June 2017
DOI: 10.1002/chir.22766
Revised: 25 July 2017
Accepted: 30 July 2017
R E G U L A R A R T I C L E
Synthesis of new C3 symmetric amino acid‐ and
aminoalcohol‐containing chiral stationary phases and
application to HPLC enantioseparations
Jeongjae Yu1 | Daniel W. Armstrong2 | Jae Jeong Ryoo3
1 Department of Chemistry, Kyungpook
Abstract
Nat'l University, Daegu, South Korea
We recently reported a new C3‐symmetric (R)‐phenylglycinol N‐1,3,5‐
benzenetricarboxylic acid‐derived chiral high‐performance liquid chromatogra-
phy (HPLC) stationary phase (CSP 1) that demonstrated better results
2 Department of Chemistry, University
of Texas, Arlington, Texas, USA
3 Department of Chemistry Education,
Kyungpook Nat'l University, Daegu,
South Korea
as compared to
a
previously described N‐3,5‐dintrobenzoyl (DNB)
(R)‐phenylglycinol‐derived CSP. Over a decade ago, (S)‐leucinol, (R)‐
phenylglycine, and (S)‐leucine derivatives were used as the starting materials
of 3,5‐DNB‐based Pirkle‐type CSPs for chiral separation. In this study, three
new C3‐symmetric CSPs (CSP 2, 3, and 4) were prepared by combining the
ideas and results mentioned above. Here we describe the synthetic procedures
and applications of the new C3‐symmetric CSPs (CSP 2–CSP 4).
Correspondence
J. J. Ryoo, Department of Chemistry
Education, Kyungpook Nat'l University,
Daegu, 702‐701, South Korea.
Email: jjryoo@knu.ac.kr
Funding information
Korean Research Foundation,
Grant/Award Number: KRF‐
2016R1D1A1B03935000
KEYWORDS
(R)‐phenylglycine, (S)‐leucine, (S)‐leucinol, C3 symmetry, chiral stationary phases, HPLC, N‐phenyl
amide
1 | INTRODUCTION
chiral selectors of C3‐symmetric CSPs. An easily
synthesizable N‐3,5‐DNB‐leucinol‐derived CSP (DNB‐
Various types of high‐performance liquid chromatogra-
phy (HPLC) chiral stationary phases (CSPs) such as
brush/Pirkle‐type stationary phases,1 ligand exchange,2
crown ethers,3 proteins,4 polysaccharides,5 cyclodextrin,6
and cyclofructan7 have been developed.8,9 In a previous
study, a new (R)‐phenylglycinol‐derived C3‐symmetric
CSP and three C2‐symmetric CSPs were prepared and
leu‐ol), reported in 2003, showed better results compared
to other aminoalcohol‐derived CSPs.11 However, despite
its similar structure to N‐3,5‐DNB‐phenylglycinol, the
latter did not show a similar chiral separation pattern as
that of N‐3,5‐DNB‐leucinol. Instead, the N‐phenyl‐DNB‐
leucine‐derived CSP (N‐ph‐DNB‐leu) reported in 1998
showed a much better resolution than the previously
reported original N‐3,5‐DNB‐leucine‐derived CSP
(DNB‐leu) did.12 The structures of these three CSPs,
DNB‐Leu‐ol, DNB‐Leu, and N‐ph‐DNB‐leu CSP, are
shown in Figure 1.
The N‐phenyl moiety in the amide group plays an
important role in chiral separation. In this study, we
introduced a phenyl group at the same amide nitrogen
position and designed two amino acid‐derived C3‐sym-
metric CSPs, (R)‐phenylglycine and (S)‐leucine, and
one aminoalcohol‐derived CSP, (S)‐leucinol. We
their
chiral
separation
performances
for
25
chiral samples were evaluated and compared with that
of N‐3,5‐dintrobenzoyl (DNB)‐(R)‐phenylglycinol‐derived
CSP, which was previously reported as a brush/Pirkle‐
type CSP, under the same separation conditions.10 The
results revealed that the (R)‐phenylglycinol‐derived C3‐
symmetric CSP had a better resolution and separated
more chiral samples compared to the other CSPs.10
(S)‐Leucinol, (S)‐leucine, and (R)‐phenylglycine deriv-
atives were considered good candidates for improved
Chirality. 2017;1–11.
wileyonlinelibrary.com/journal/chir
© 2017 Wiley Periodicals, Inc.
1

2
YU ET AL.
O
O
O
O
O
O
O
O
O
O
N
Si
N
Si
EtO
N
Si
H
H
NH
O
NH
EtO
O
NH
O
EtO
O2N
NO2
DNB-leu-ol
FIGURE 1 Structure of previously reported CSPs
O2N
NO2
N-ph-DNB-leu
O2N
NO2
DNB-leu
expected that the newly designed CSPs would show
much better resolutions than those of the previously
reported similar C3‐symmetric CSPs, owing to the addi-
tional N‐phenyl group. These three new symmetric
CSPs were synthesized and used to separate chiral sam-
ples by HPLC. In addition, the previously reported (R)‐
phenylglycinol‐derived C3‐symmetric CSP (CSP 1) was
prepared for direct comparison under the same separa-
tion conditions. In this study, 56 chiral samples (13
π‐acidic, 13 π‐basic, 11 aromatic, and 19 oxazolidinone
chiral compounds) were used. The structures of the 56
chiral samples are shown in Figures 2, 3, 4, and 5.
O
O
H2N
O
O
OH
O
O
O
O
HN
HN
HN
O2N
NO2
O2N
NO2
O2N
NO2
methyl-2-(3,5-dinitrobenzamide)- methyl-4-amino-(3,5-dinitrobenzamide)- N-(3,5-dinitrobenzoyl)-
4-oxobutanoate (A2)
leucinol (A3)
3-phenylpropanoate (A1)
OH
OH
O
OH
O
HN
O
HN
HN
Cl
O2N
NO2
O2N
NO2
O2N
NO2
-(3,5-dinitrobenzoyl)-
-(3,5-dinitrobenzoyl)-
N
N
N-(3,5-dinitrobenzoyl)-
2-phenylglycinol (A4)
p-chlorophenylalaninol (A5)
-1-amino-2-indanol (A6)
cis
OH
O
O
OH
O
OH
O
HN
HN
HN
O
HN
O
O2N
NO2
O2N
NO2
O2N
NO2
O2N
NO2
-(3,5-dinitrobenzoyl)-
phenylalaninol (A7)
N
N -(3,5-dinitrobenzoyl)-
N-(3,5-dinitrobenzoyl)-
methyl 2-(3,5-dinitrobenzamido)-
2-phenylacetate (A10)
2-amino-1-propanol (A8) 2-amino-1-phenylethanol (A9)
O
O
O
O
O
O
HN
O
HN
O
HN
O
O2N
NO2
O2N
NO2
O2N
NO2
ethyl 2-(3,5-dinitrobenzamido)-
2-phenylacetate (A11)
butyl 2-(3,5-dinitrobenzamido)-
2-phenylacetate (A12)
hexyl 2-(3,5-dinitrobenzamido)-
2-phenylacetate (A13)
FIGURE 2 Structure of π‐acidic chiral samples (group a)

YU ET AL.
3
FIGURE 3 Structure of π‐basic chiral samples (group B)
FIGURE 4 Structure of aromatic chiral samples (group C)
ThermoFisher (Waltham, MA) Flash 2000 elemental
analyzer.
2 | MATERIALS AND METHODS
2.1 | General methods
2.2 | Reagents
1H–NMR spectra were measured with a Bruker (Billerica,
MA) AVANCE digital 400 spectrometer (400 MHz), and
All chemicals used in this study were purchased from
elemental analysis data were obtained using
a
Sigma‐Aldrich Korea (Seoul, Korea) or Tokyo Chemical

4
YU ET AL.
FIGURE 5 Structure of oxazolidinone chiral samples (group D)
Industry Co. (Tokyo, Japan). The organic solvents were
obtained from Junsei Chemical Co. (Tokyo, Japan), and
5 μm spherical silica gel (Daiso gel) was provided by Daiso
Co. (Osaka, Japan).
then n‐hexane was added to it. A white solid product
was obtained, which was recrystallized and dried in a vac-
uum oven.
1a: Yield 62.2%, 1H NMR (DMSO‐d6) δ: 3.63–3.67 (dd,
1H), 3.71–3.76 (dd, 1H), 5.07–5.12 (m, 2H), 7.21–7.25 (tt,
1H), 7.30–7.33 (t, 2H), 7.38–7.40 (d, 2H), 8.49 (s, 1H),
9.09–9.12 (d, 1H).
2.3 | Preparation of chiral selectors 1b and
2b
1
2a: Yield 76.5%, H NMR (DMSO‐d6) δ: 0.86–0.89 (d,
3H), 0.90–0.91 (d, 3H), 1.36–1.42 (m, 1H), 1.46–1.53 (m,
1H), 1.57–1.70 (m, 1H), 3.36–3.40 (dd, 1H), 3.43–3.49
(dd, 1H), 4.06–4.15 (m, 1H), 4.72–4.75 (t, 1H), 8.29–8.31
(d, 1H), 8.39 (s, 1H).
2.3.1 | N1,N3,N5‐Tris(2‐hydroxy‐1‐
phenylethyl)benzene‐1,3,5‐tricarboxamide
(1a) and N1,N3,N5‐tris(1‐hydroxy‐4‐
methylpentan‐2‐yl)benzene‐1,3,5‐
tricarboxamide (2a)
2.3.2 | N1,N3,N5‐Tris(2‐hydroxy‐1‐
A
solution of 1,3,5‐benzenetricarbonyl trichloride
phenylethyl)benzene‐1,3,5‐tricarboxamide
(triethoxysilyl) propyl carbamate (1b) and
N1,N3,N5‐tris(1‐hydroxy‐4‐methylpentan‐2‐
yl)benzene‐1,3,5‐tricarboxamide
(3.24 mmol) in dichloromethane was slowly added to a
stirred solution of the aminoalcohol [(R)‐phenylglycinol
or (S)‐leucinol)] (10.72 mmol) and triethylamine (1 mL)
in dichloromethane at 0 °C. The reaction mixture was
stirred at room temperature under nitrogen for 36 h, and
then washed with 1.2 N HCl (aq.) and saturated NaHCO3
(aq.). After drying over anhydrous MgSO4, the solvent was
removed. The crude reaction mixture was dissolved in a
mixed solution of methanol and dichloromethane and
(triethoxysilyl) propyl carbamate (2b)
A
solution of 3‐(triethoxysilyl)propyl isocyanate
(5.81 mmol) and compound 1a or 2a (1.76 mmol) in pyr-
idine (100 mL) was stirred at reflux for 12 h. After cooling,
the mixture was directly used in the next step.

YU ET AL.
5
4b: Yield 80.9%, 1H NMR (CDCl3) δ: 0.24–0.36 (d, 3H),
0.68–0.80 (d, 3H), 1.14–1.24 (t, 1H), 1.51–1.67 (m, 1H),
1.92–2.12 (m, 2H), 4.41–4.61 (m, 2H), 5.02–5.22 (m, 4H),
6.04–6.29 (m, 1H), 7.31–7.55 (m, 5H), 8.57 (s, 1H).
2.4 | Preparation of chiral selectors 3b and
4b
2.4.1 | tert‐Butyl‐2‐(allyl(phenyl)amino)‐2‐
oxo‐1‐phenylethylcarbamate (3a) and tert‐
butyl‐1‐(allyl(phenyl)amino)‐4‐methyl‐1‐
oxopentan‐2‐ylcarbamate (4a)
2.4.3 | N1,N3,N5‐Tris(2‐((3‐
(ethoxydimethylsilyl)propyl)(phenyl)
amino)‐2‐oxo‐1‐phenylethyl)benzene‐1,3,5‐
tricarboxamide (3c) and N1,N3,N5‐tris(1‐((3‐
(ethoxydimethylsilyl)propyl)(phenyl)
amino)‐4‐methyl‐1‐oxopentan‐2‐yl)ben-
zene‐1,3,5‐tricarboxamide (4c)
A
solution
of
2‐ethoxy‐1‐ethoxycarbonyl‐1,2‐
dihydroquinoline (22.0 mmol) in dichloromethane
(30 mL) was added dropwise to a stirred solution of the
N‐protected amino acid (20.0 mmol) in dichloromethane
(70 mL) at 0 °C and stirred for 5 min. Then N‐allylaniline
was added and the reaction mixture was stirred for 12 h.
The mixture was then washed with 1.2 N HCl (aq.) and
saturated NaHCO3 (aq.). After drying over anhydrous
MgSO4, the solvent was removed. The residue was
purified by column chromatography on silica gel (ethyl
acetate/hexane: 1/4, v/v) to afford 3a or 4a as white solids,
which were dried in a vacuum oven.
A solution of dimethylchlorosilane (288.1 mmol) and
H2PtCl6·6H2O (6 mg) in THF (1 mL) was added to a solu-
tion of 3b or 4b (1.80 mmol) in dichloromethane (30 mL).
The reaction mixture was heated to reflux for 2 h. The sol-
vent and excess dimethylchlorosilane were evaporated
under vacuum and the residue was dissolved in dichloro-
methane. To this solution triethylamine (10 mL) and eth-
anol were slowly added and the mixture stirred for 1 h.
After drying, the product was used in the next step with-
out further purification.
1
3a: Yield 75.3%, H NMR (CDCl3) δ: 1.35 (s, 12H),
4.18–4.26 (dd, 1H), 4.27–4.37 (dd, 1H), 5.02–5.10 (m,
2H), 5.20–5.27 (d, 1H), 5.76–5.88 (m,1H), 6.90–6.98 (d,
4H), 7.14–7.27 (m, 4H), 7.27–7.32 (m, 2H).
4a: Yield 52.8%, 1H NMR (CDCl3) δ: 0.35–0.37 (d, 3H),
0.70–0.72 (d, 3H), 1.19–1.22 (m 1H), 1.35 (s, 12H), 4.17–
4.21 (dd, 1H), 4.32–4.37 (m, 2H), 5.08–5.11 (m, 3H),
5.78–5.87 (m, 1H), 7.23–7.43 (m, 5H).
2.5 | CSPs
To a 500 mL round bottom flask were added 5 μm silica
gel (4.00 g) and toluene (100 mL). From the resulting
slurry, water was removed azeotropically using a Dean‐
Stark trap. After the complete removal of water, a solution
of the prepared silyl compounds (1b, 2b, 3c, or 4c) in pyr-
idine (100 mL) was added to the slurry and the mixture
was heated to reflux for 12 h with magnetic stirring. The
modified silica gel was sequentially washed with toluene,
dichloromethane, ethyl acetate, ethyl alcohol, and ethyl
ether, and then dried in a vacuum oven. The dried silica
gel was packed into a 250 × 4.6 mm ID stainless steel col-
umn via conventional methods.13
2.4.2 | N1,N3,N5‐Tris(2‐(allyl(phenyl)
amino)‐2‐oxo‐1‐phenylethyl)benzene‐1,3,5‐
tricarboxamide (3b) and N1,N3,N5‐tris(1‐
(allyl(phenyl)amino)‐4‐methyl‐1‐
oxopentan‐2‐yl)benzene‐1,3,5‐
tricarboxamide (4b)
Compound 3a or 4a (10.6 mmol) was dissolved in 4 N
HCl/dioxane at 0 °C and the solution was stirred for
30 min at room temperature. The solvent was then evap-
orated under vacuum to yield a residue, which was used
2.6 | 1.6 Chromatography
directly without purification.
A solution of 1,3,5‐
benzenetricarbonyl trichloride (2.26 mmol) in dichloro-
methane was slowly added to the stirred solution of the
residue in dichloromethane, and the mixture was stirred
for 12 h at room temperature under nitrogen. The reac-
tion mixture was then washed with 1.2 N HCl (aq.) and
saturated NaHCO3 (aq.). After drying over anhydrous
MgSO4, the solvent was removed. The residue was puri-
fied by column chromatography on silica gel and dried
in a vacuum oven.
The HPLC system was a Waters 2690 Separation Module,
which consists of a Waters 996 photodiode array detector
and an auto sampler (Waters, Milford, MA). HPLC‐grade
solvents were obtained from J.T. Baker (Center Valley,
PA). The mobile phases consisted of 2‐propanol/n‐hexane
(10/90) and 2‐propanol/n‐hexane/trifluoroacetic acid (10/
90/0.1) with a flow rate of 1.2 mL/min. The column void
volume was checked by injecting 1,3,5‐tri‐tert‐
butylbenzene, a presumed unretained solute14 obtained
from Aldrich Chemical Co. (Milwaukee, WI).
1
3b: Yield 71.4%, H NMR (CDCl3) δ: 4.22–4.25 (m,
3H), 5.00–5.06 (dd, 2H), 5.75–5.84 (m, 1H), 6.92–6.94 (m,
3H), 7.16–7.18 (m, 7H), 8.52 (s, 1H).
Fifty‐six chiral samples were used in this study, which
were classified into four groups for convenience. Group A

6
YU ET AL.
contained 13 π‐acidic amino acid derivatives, group B
consisted of 13 π‐basic samples, group C contained 11 aro-
matic chiral compounds, and group D comprised 19
oxazolidinone series of chiral medicines. The sample
structures are shown in Figure 1. The chiral samples were
obtained from Sigma‐Aldrich Korea (Seoul, Korea) and
Prof. Armstrong's Lab (UTA, Arlington, TX).
3b and 4b, their syntheses were easier to handle com-
pared to those of 1a and 2a due to the solubility issues
of the latter.
By using pyridine as the organic solvent in steps 2 and
3, the synthetic procedures of CSP 1 and 2 could be
improved compared to the previously reported proce-
dure.1 This modification reduced the reaction time and
increased the amount of chiral selector introduced into
the silica gel (the amount of loaded chiral selector of the
previously reported CSP 1 was 0.167 mmol/g).10 The
crude intermediates 3c and 4c were not purified because
some of the impurities could be removed in the course
of silica gel washing.
The obtained compounds were identified by compar-
ing the 1H–NMR data with those of the previously
reported similar compounds. In addition, elemental anal-
yses were carried out to determine the extent of bonding
between the chiral selectors and silica gel. These results
are shown in Table 1.
3 | RESULTS AND DISCUSSION
The synthetic procedures for CSP 1 and CSP 2 are shown
in Scheme 1, and those for CSP 3 and CSP 4 in Scheme 2.
CSP 1 and CSP 2 were prepared by slightly modifying a
previously reported method,1 while CSP 3 and CSP 4 were
synthesized by referring to the literature.12,15-17 The syn-
theses of CSP 1 and CSP 2 included three steps, while
CSP 3 and CSP 4 required six steps for synthesis. Interme-
diates 1a, 2a, 3b, and 4b were obtained as white solids.
The solubilities of 1a and 2a were very poor in most
organic solvents (except for dimethylformamide [DMF],
dimethyl sulfoxide [DMSO], and pyridine). On the other
hand, 3b and 4b demonstrated good solubilities in the
usual organic solvents such as dichloromethane.
Although more synthetic steps were required to obtain
As shown in Table 1, the silica gels in CSP 1 and CSP 2
contain more than 0.25 mmol/g of chiral selectors,
whereas those in CSP 3 and CSP 4 contain about
0.15 mmol/g. It is assumed that the lower amount of
loaded chiral selectors in CSP 3 and CSP 4 is due to the
low yield of the hydrosilylation step in CSP 3 and CSP 4.
SCHEME 1 Synthetic procedure of CSP 1 and CSP 2

YU ET AL.
7
SCHEME 2 Synthetic procedure of CSP 3 and CSP 4
TABLE 1 Elemental analysis of CSPs
C (%)
N (%)
Based on C (mmol/g)
Based on N (mmol/g)
CSP 1
CSP 2
CSP 3
CSP 4
15.30
2.15
0.283
0.255
12.22
12.46
10.42
1.96
1.26
1.14
0.261
0.157
0.145
0.233
0.150
0.136
The chiral analytes used in this study were classified
into four groups. Group A is composed of π‐acidic chiral
samples of amino acid derivatives containing the 3,5‐
dinitrobenzoyl (DNB) group, group B contains π‐basic
samples with multiple benzene rings, group C contains
chiral compounds with aromatic groups, and group D
includes the chiral samples of the drugs of the
oxazolidinone family used as antimicrobial agents. Chiral
separations of the 56 enantiomeric samples were per-
formed on CSP 1–4. The chromatograms of N‐(3,5‐
dinitrobenzoyl)‐phenylalaninol (A7) on all the CSPs are
shown in Figure 6.
of the analysis of the 56 chiral samples on CSPs 1–4 are
summarized in Table 2.
In the case of π‐acidic chiral compounds (Group A),
CSP 1 resolved three chiral samples and CSP 2
The enantiomeric ratio of (R)‐ to (S)‐N‐(3,5‐
dinitrobenzoyl)‐phenylalaninol (A7) was 2:1, which
helped confirm the order of elution.
The separation factor (α) of A7 on CSP 1 was low
(1.06), whereas high separation factors were obtained on
CSP 2 (1.65), CSP 3 (1.27), and CSP 4 (2.27). The results
FIGURE 6 Separation of N‐(3,5‐dinitrobenzoyl)‐phenylalaninol
(A7) on the prepared CSP 1, 2, 3, and 4. Detection at 254 nm.
Flow rate, 1.2 mL/min. Eluent, IPA/hexane/trifluoroacetic acid
(TFA): 10/90/0.1, IPA/hexane: 10/90 For CSP 4

8
YU ET AL.
TABLE 2 Resolution of chiral samples on CSPs
CSP1
CSP2
k1
CSP3
k1
2.65a
CSP4
k1
α
α
α
k1
α
A1
A2
A3
A4
A5
A6
A7
A8
A9
A10
A11
A12
A13
B1
4.34
1.00
1.00
1.00
1.06a
1.11a
1.27
1.00
1.00
1.24a
1.00a
1.00a
1.71a
1.00a
1.55a
1.00a
1.65a
1.18
1.50
1.00
1.00
1.00
1.12
1.11
2.50
2.27
1.90
1.00
4.28
4.52
1.28
2.42a
1.49
3.47
2.34
6.88
3.86(R)a
3.62a(S)
9.81a
8.48a
9.64a
8.89a
ND
ND
8.25a
ND
1.29a
ND
1.06a
9.07(R)
6.75(R)a
9.01(R)
6.91
1.10
1.27a
1.12
1.00
1.28
1.00
1.28
1.29
1.00
8.83a
4.90(R)
7.57(S)a
6.86a(S)
1.73(S)
7.70a
5.79(R)
9.53(R)
15.10
ND
ND
ND
ND
1.62a
2.82
1.00
5.87
2.50
1.00
ND
ND
ND
ND
ND
ND
ND
ND
0.51
5.69
1.00
ND
ND
ND
ND
ND
ND
5.73
ND
ND
1.00
5.79
5.69
5.38
12.25
6.72
1.97
8.20
1.07
4.61
6.47
1.13
2.66
1.19
1.22
1.03
1.00
1.12
1.07
1.07
3.21
1.00
B2
9.53
4.59
1.33
8.71
1.17
1.06
1.09
1.07
ND
ND
5.01
2.32
0.90
1.04
1.43
1.22
B3
2.62a
0.94a
2.67
0.29
0.84
3.23a
0.34a
1.58
1.28
1.59
1.00a
1.00a
1.00
1.00
1.00
1.17a
1.00a
1.00
1.00
1.00
B4
B5
ND
ND
B6
ND
ND
0.39
1.15
0.50
0.42
1.00
1.00
1.00
1.00
B7
1.74
0.59
0.43
3.16a
1.41
1.00
1.00
1.10a
B8
ND
ND
B9
0.74
2.78
1.37
3.45
1.72
0.99
1.18
1.60
0.72
1.10
2.25
2.54
1.27
2.63
1.60
1.99
4.34
5.91
5.86
1.00
1.16
1.17
1.31
1.04
1.00
1.00
1.00
1.00
1.00
1.09
1.00
1.00
1.08
1.00
1.00
1.08
1.00
1.00
B10
B11
B12
B13
C1
ND
ND
ND
3.34
ND
ND
1.47
ND
1.38
ND
3.81
2.74
ND
ND
1.07
ND
ND
1.00
ND
1.00
ND
1.13
1.00
ND
0.63
1.86
1.14
1.00
ND
ND
ND
ND
0.79a
0.48a
0.97a
0.51
1.00a
1.00a
1.00a
1.00
1.31
2.03
1.00
2.58
C2
C3
ND
ND
C4
0.89
0.87
1.39
2.45
0.85
1.28
0.83
0.79
3.68
4.45
5.35
1.00
1.00
C5
0.88
1.00
C6
1.97a
2.17a
0.51a
1.34a
0.87a
0.58a
3.67a
8.37a
7.90a
1.00a
1.00a
1.00a
1.00a
1.00a
1.00a
1.06a
1.14a
1.20a
1.00
C7
1.00
C8
1.00
C9
2.76
1.82
1.00
1.00
1.18
C10
C11
D1
D2
D3
1.29
ND
ND
1.35
6.58
8.67
6.21a
1.00
1.03
1.03a
1.08a
1.00
1.08a
(Continues)

YU ET AL.
9
TABLE 2 (Continued)
CSP1
k1
CSP2
k1
8.25a
CSP3
k1
8.39a
CSP4
k1
6.30a
α
α
α
α
D4
2.21
5.76
1.00
1.00
1.08a
1.00
1.09a
1.00
1.00
1.00
1.05a
1.00
1.00a
1.25a
1.00
1.00
1.05
1.07a
1.12a
1.00
1.08
1.08a
1.13a
D5
4.01
5.38
8.03
5.75
5.51a
5.65
2.34a
2.47
1.56
7.00
4.60
3.53a
3.71a
5.23
5.96
3.63
3.35
5.24
4.43
4.20
4.49
2.32
6.27
7.17
6.15
3.21
1.72
2.52
3.12
4.08
2.11
1.00
1.05
1.00
1.00
1.00
1.00a
1.11
1.00
1.00
1.00
1.10
1.00
1.00
1.19
1.00
D6
ND
ND
1.00
7.72a
8.59a
7.00a
7.00
1.08a
1.13a
1.13a
1.12
D7
7.01
4.44
4.00
D8
1.14
1.00
D9
D10
D11
D12
D13
D14
D15
D16
D17
D18
D19
ND
6.78
ND
ND
7.95
ND
1.00
ND
ND
1.17
ND
ND
3.60
ND
ND
1.00
9.85
5.40a
2.24
3.91
6.34
4.83a
3.48a
1.05
1.29a
9.05
8.08
7.26a
1.11
1.11
1.11a
1.00
1.00
1.16a
1.07a
1.21a
ND
ND
ND
ND
Detection at 210 to 400 nm; flow rate, 1.2 mL/min; eluent, IPA/n‐hexane: 10/90.
aIPA/hexane/TFA: 10/90/0.1.
separated the largest number of samples, i.e., eight (A3,
A5, A6, A7, A8, A10, A11, A12, and A13), whereas CSP
3 and CSP 4 separated five chiral samples. In particular,
CSP 4 showed an impressive separation factor for A6,
A7, and A8, with a value of around 2. The order of elu-
tion was identified for analytes A3, A6, A7, and A8: the
factor. Moreover, CSP 3 showed good chiral recognition
for nine oxazolidinone samples, while CSP 1 and CSP 4
separated only seven analytes.
The summary of this study is shown in Table 3.
Thus, CSP 1 separated 19 chiral samples from the 36
detected samples (53%), CSP 2 resolved 22 chiral samples
from the 51 detected samples (43%), CSP 3 separated 26
chiral samples from the 54 detected samples (48%), and
CSP 4 resolved 20 chiral samples from the 48 detected
samples (42%). As shown in Table 3, CSP 4 exhibits the
highest average separation factor and CSP 2 performed
the highest number of separations for the oxazolidinone
samples (Group D). Particularly, in the case of CSP 4, rel-
atively high resolutions were obtained for certain samples
such as A6, A7, A8, and C2.
CSP 1 and CSP 3 have a phenyl group close to their
chiral centers, which derives from phenylglycinol or
phenylglycine, whereas CSP 2 and CSP 4 have an isobutyl
group derived from leucinol or leucine. The π‐basic chiral
compounds (Group B) resolved much better on the phe-
nyl‐derived CSPs than on the isobutyl‐derived ones. Thus,
it is assumed that face‐to‐edge π–π interactions occur
between the phenyl group in the chiral selectors and the
π‐basic chiral compounds.
(R)‐phenylglycine‐based CSPs (CSP
1 and CSP 3)
interacted with the (R)‐form more strongly than with
the (S)‐form, while the (S)‐leucine‐based CSPs (CSP 2
and CSP 4) exhibited a reverse order. This implies that
the (R)‐selectors interact strongly with the (R)‐analytes
and (S)‐selectors interact strongly with the (S)‐analytes.
In Group B, CSP 1 and CSP 3 separated eight and ten π‐
basic chiral compounds, respectively. Particularly, CSP 3
separated the most chiral samples used in this study (all
except B4, B8, and B9), and showed better resolutions com-
pared to the other CSPs, which suggest that the amide‐
bonded phenyl group on CSP 3 plays a more important role
in chiral resolution than the carbamate group on CPS 1
does. On the other hand, CSP 2 resolved only one chiral
analyte (B8) and CSP 4 separated four. In the case of aro-
matic compound samples (Group C), only a few samples
could be separated on all the CSPs; however, CSP 4 showed
the best resolution for C2 (α = 2.58).
In the case of oxazolidinone samples (Group D), CSP 2
resolved 13 samples with the highest average separation
CSP 1 and CSP 2 associate with the silica gel through
carbamate linkages, whereas CSP 3 and CSP 4 associate

10
YU ET AL.
TABLE 3 Summary of chiral separation results on newly prepared symmetric CSPs
The number of separated samples
Ratio of
resolved
samplesa
Average of
separation
factorb
π‐Acidic compounds
π‐Basic compounds
Aromatic compounds
(group C)
Oxazolidinones
(group D)
(group A)
(group B)
CSP 1
CSP 2
CSP 3
CSP 4
3
8
5
5
8
1
0
2
4
7
19/36(53%)
22/51(43%)
26/54(48%)
20/48(42%)
1.06
1.07
1.10
1.16
1
13
9
10
4
7
a(The number of resolved samples) / (the number of detected samples).
b(The sum of separation factors of detected samples) / (the number of detected samples).
exchangers implementing alkyne‐azide cycloaddition immobili-
zation chemistry. J Chromatogr A. 2014;1337:85‐94.
through N‐phenyl amide linkages. CSP 3 and CSP 4 showed
better average separation factors than CSP 1 and CSP 2 did.
3. Ahn SA, Hyun MH. Liquid chromatographic resolution of
amino acid esters of acyclovir including racemic valacyclovir
on crown ether‐based chiral stationary phases. Chirality.
2015;27:268‐273.
Therefore, it is assumed that the phenyl group attached to
the amide linkage in the chiral selectors interrupts the
hydrogen bonding interactions between the amide hydro-
gen, which is replaced by the phenyl group, and the chiral
compounds, thereby inducing an effective interaction at a
position near the chiral center. Nevertheless, the exact role
of this amide N‐phenyl group is not yet understood.
4. Song WJ, Wei JP, Wang SY, Wang HS. Restricted access chiral
stationary phase synthesized via reversible addition‐fragmenta-
tion chain‐transfer polymerization for direct analysis of
biological samples by high performance liquid chromatography.
Anal Chim Acta. 2014;832:58‐64.
5. Tan Y, Han J, Lin C, Yu H, Zheng S, Zhang W.
Synthesis and enantioseparation behaviors of novel immobilized
3,5‐dimethylphenylcarbamoylated polysaccharide chiral station-
ary phases. J Sep Sci. 2014;37:488‐494.
4 | CONCLUSION
We prepared three new C3‐symmetric CSPs (CSP 2, 3, and
4) and used them for the chiral resolution of 56 chiral sam-
ples. Even though they have no π‐acidic or π‐basic groups
in their structures, the newly prepared CSP 3 and 4 exhib-
ited better resolutions than those of the previously reported
π‐acidic phenylglycinol‐ or leucinol‐derived CSPs. In addi-
tion, even though the exact role of this N‐phenyl amide
group is not yet understood, it was revealed that the phenyl
group attached to the amide linkage in the chiral selectors
played an important role in chiral separation.
6. Yao X, Tan TTY, Wang Y. Thiol‐ene click chemistry derived
cationic cyclodextrin chiral stationary phase and its enhanced
separation performance in liquid chromatography. J Chromatogr
A. 2014;1326:80‐88.
7. Sun P, Wang C, Breitbach ZS, Zhang Y, Armstrong DW.
Development of new HPLC chiral stationary phases based on
native
and
derivatized
cyclofructans.
Anal
Chem.
2009;81:10215‐10226.
8. Ward TJ, Ward KD. Chiral separations: fundamental review
2010. Anal Chem. 2010;82:4712‐4722.
9. Ward TJ, Ward KD. Chiral separations: a review of current
topics and trends. Anal Chem. 2012;84:626‐635.
ACKNOWLEDGMENTS
This work was supported by a grant from the Korean
Research Foundation (KRF‐2016R1D1A1B03935000).
10. Yu JJ, Ryoo DH, Lee JM, Ryoo JJ. Synthesis and application of
C2 and C3 symmetric (R)‐phenylglycinol‐derived chiral station-
ary phases. Chirality. 2016;28:186‐191.
11. Ryoo JJ, Kim TH, Im SH, et al. Enantioseparation of racemic
ORCID
N‐acylarylalkylamines on various amino alcohol derived
π‐acidic chiral stationary phases.
2003;987:429‐438.
J
Chromatogr A.
Jae Jeong Ryoo http://orcid.org/0000-0002-8741-6820
12. Hyun MH, Lee JB, Kim YD. An improved (S)‐leucine‐
derived π‐acidic HPLC chiral stationary phase for the resolu-
REFERENCES
tion of π‐acidic racemates.
1998;21:69‐71.
J High Resol Chromatogr.
1. Han HK, Hong JH, Carey JR, Kang JS, Lee WJ. Chiral recogni-
tion using a conformationally rigid chiral stationary phase
derived from α‐amino‐ϵ‐carprolactam. J Liq Chromatogr Relat
Technol. 2014;37:2725‐2732.
13. Ryoo JJ, Im SH, Lee KP, Park JH, Hyun MH. Resolution
of racemic N‐Acyl‐1‐arylaminoalkanes on (3,5‐
Dinitrobenzoyl)phenylglycinol‐derived high performance
a
2. Hettergger H, Kohout M, Mimini V, Lindner W. Novel
carbamoyl type quinine and quinidine based chiral anion
liquid chromatography chiral stationary phase. Microchem
J. 1999;63:128‐133.

YU ET AL.
11
14. Pirkle WH, Welch CJ, Lamm B. Design, synthesis, and evalua-
tion of an improved enantioselective naproxen selector. J Org
Chem. 1992;57:3854‐3860.
17. Han G, Tamaki M, Hruby VJ. Fast, efficient and selective
deprotection of the tert‐butoxycarbonyl (Boc) group using HCl/
dioxane (4M). J Peptide Res. 2001;58:338‐341.
15. Uray G, Lindner W. (S,S)‐Diphenyl‐ethanediamine (DPEDA)
derivatives as chiral selectors. Part Ι. Undecenoyl bound
dinitrobenzoyl‐DPEDA as a broadly applicable chiral stationary
phase. Chromatographia. 1990;30:323‐327.
How to cite this article: Yu J, Armstrong DW,
Ryoo JJ. Synthesis of new C3 symmetric amino
acid‐ and aminoalcohol‐containing chiral
stationary phases and application to HPLC
enantioseparations. Chirality. 2017;1–11. https://
doi.org/10.1002/chir.22766
16. Hyun MH, Kim KJ, Jyung KK. Preparation of a new HPLC chiral
stationary phase from (S)‐naproxen and application in elucidat-
ing chiral recognition models. Bull Korean Chem Soc.
1997;18:1085‐1089.