Resolution of enantiomers using sugar-carrying polyisocyanides as chiral stationary phases for HPLC

Article abstract of DOI:10.1246/bcsj.75.2681

3,5-Dimethylphenylcarbamate derivatives of α/β-glucose and α/β-galactose-carrying helical poly(phenyl isocyanide)s were used as chiral stationary phases (CSPs) for HPLC to estimate their chiral recognition abilities. CD spectroscopy suggested that the hel

Full text of DOI:10.1246/bcsj.75.2681

c. Jpn.
© 2002 The Chemical Society of Japan
Bull. Chem. Soc. Jpn., 75, 2681–2685 (2002) 2681
e Chemical
ciety of Ja-
n
e Chemical
ciety of Ja-
n
Resolution of Enantiomers Using Sugar-Carrying Polyisocyanides as
Chiral Stationary Phases for HPLC
Sugar-Carrying Poly(phenyl isocyanide)s
A. Tsuchida et al.
Akiko Tsuchida, Teruaki Hasegawa,† Kazukiyo Kobayashi,† Chiyo Yamamoto, and Yoshio Okamoto*
02
81
85
SJA8
09-2673
177
Department of Applied Chemistry, Graduate School of Engineering, Nagoya University, Chikusa-ku, Nagoya 464-8603
†Department of Molecular Design and Department of Biotechnology, Graduate School of Engineering,
Nagoya University, Chikusa-ku, Nagoya 464-8603
(Received June 24, 2002)
02
3,5-Dimethylphenylcarbamate derivatives of α-/β-glucose and α-/β-galactose-carrying helical poly(phenyl isocya-
nide)s were used as chiral stationary phases (CSPs) for HPLC to estimate their chiral recognition abilities. CD spectro-
scopy suggested that the helix sense in these rigid helical polymers may be regulated by the chirality of the α- or β-ano-
meric center of the sugar moieties. Some 10 different types of racemates with functional groups were completely or
partially resolved depending on the stereostructure of the pendant sugars. The enhanced chiral recognition ability is
attributable to the three-dimensionally regulated sugar arrays along the helical backbone; that is, the CSP of the α-glu-
cose-carrying helical poly(phenyl isocyanide) exhibited more effective enantioseparation than that of the corresponding
flexible poly(N-phenylacrylamide). The CSPs of the galactose-carrying poly(phenyl isocyanide)s showed the resolving
ability for broader racemates than those of the glucose-type poly(phenyl isocyanide)s. Especially, the CSP of the α-ga-
lactose-carrying poly(phenyl isocyanide) separated the largest number of racemates.
02
2.1
Saccharide chains on cell surfaces play important roles in
cell–cell recognition as well as in infection by viruses and bac-
teria.1 Synthetic polymers carrying various kinds of pendant
saccharide chains are useful tools to investigate saccharide-
recognition events.2 The clustered saccharide chains along
flexible polystyrene and polyacrylamide backbones were re-
ported to strongly bind to a variety of carbohydrate-binding
proteins (lectins).3 However, glycosylated poly(phenyl isocya-
nide)s exhibited few specific interactions with lectins, in spite
of the multivalent saccharide arrays.4 The saccharide arrays of
the glycosylated poly(phenyl isocyanide)s were regulated in a
helical manner so tightly that they could neither access nor in-
duce-fit to the binding sites of the lectins. A polymer with a
rigid backbone, such as polyisocyanide, may be suitable for
the recognition of small compounds.
glucose- and galactose-carrying polyisocyanides (1–4) (Fig. 1)
and evaluated their chiral recognition abilities as CSPs for
HPLC, in comparison with that of the flexible backbone of
poly(N-phenylacrylamide) (5).
Materials and Methods
1
Measurements. The H and 13C NMR spectra were re-
corded on Varian Gemini-200 (200 MHz for 1H) and VXR-500
1
(500 MHz for H and 125 MHz for 13C) NMR spectrometers.
The chemical shifts are reported in ppm (δ) relative to Me4Si
or residual non-deuterated solvent. The IR spectra were re-
corded on a JASCO FT/IR-230 Fourier-transform infrared
spectrometer. Circular dichroism (CD) spectroscopy was mea-
sured by a JASCO J-720L spectropolarimeter using a 10 mm
quartz cell.
The phenylcarbamate derivatives of naturally occurring
chiral polymers, cellulose and amylose, have been commer-
cialized as chiral stationary phases (CSPs) for high-perfor-
mance liquid chromatography (HPLC) to separate many race-
mates including drugs.5 The chiral separation abilities of these
polysaccharide derivatives originate from the crowded polar
phenylcarbamate residues along the highly stereoregular poly-
mer backbones. CSPs consisting of synthetic polymers, such
as polyamides, polyacrylamides, and one-handed helical poly-
methacrylates, have been extensively studied.6 Although poly-
isocyanides are also non-natural helical polymers, few papers
have reported on their chiral recognition ability.7
Materials. Macroporous spherical silica gel (Daiso gel
SP-1000) with a mean particle size of 7 µm and a mean pore
diameter of 100 nm was supplied by Daiso Chemical (Osaka,
Japan). Analytical-grade solvents were carefully dried and dis-
tilled before preparing the CSPs.
Synthesis of Poly(phenyl isocyanide)s Carrying 3,5-Dim-
ethylphenyl-carbamoylated Sugars. The synthesis of the
glycosylated poly(phenyl isocyanide)s was carried out by the
previously described method,4 as shown in Scheme 1, starting
from p-nitrophenyl 2,3,4,6-tetra-O-acetyl-α-D-galactopyrano-
side, as an example. Briefly, the hydrogenation of the nitro
group to the amine, followed by formylation with acetic formic
anhydride in ethyl formate, gave the N-formyl-amino group,
which was then converted to an isocyano group by phosphoryl
chloride and triethylamine. The isocyano derivative was poly-
merized with nickel(Ⅱ) chloride in a mixture of chloroform
In this respect, it is of interest to investigate the chiral recog-
nition ability of glycosylated poly(phenyl isocyanide)s carry-
ing the chirality of both the saccharide side chain and polyiso-
cyanide main chain. We prepared the phenylcarbamates of the

2682 Bull. Chem. Soc. Jpn., 75, No. 12 (2002)
Sugar-Carrying Poly(phenyl isocyanide)s
Fig. 1. Rigid helical poly(phenyl isocyanide)s carrying phenylcarbamoylated sugars (1–4), together with the corresponding flexible
analogue poly(N-phenylacrylamide) polymer (5) as a control.
Scheme 1. Synthesis of rigid helical poly(phenyl isocyanide) carrying monosaccharide.
and methanol.8 The polymer was deprotected with a catalytic
amount of sodium methoxide in a mixture of methanol, tet-
rahydrofuran, and tetrachloromethane. Complete deacetyla-
tion of the polymer was carried out by the addition of water to
the reaction solution.
s, H-1), 9.3 (br, s, –OCONH–). IR (KBr) 3317, 3301, 1722,
1616, and 1552 cm−1
.
Preparation of Chiral Columns.9 Each poly(phenyl iso-
cyanide) derivative (0.75 g) was dissolved in THF (15 mL) to
coat the silica gel. Macroporous silica gel was treated with a
large excess of (3-aminopropyl)triethoxysilane in dry benzene
in the presence of a catalytic amount of dry pyridine at 80 °C
overnight. The silanized silica gel (3.0 g) was wetted with one
portion of the polymer solution as uniformly as possible; the
solvent was then evaporated under reduced pressure. The wet-
ting/drying procedure was repeated several times to coat the
silica gel with the polymer. The resulting silica gel was packed
into a stainless-steel tube (25 cm × 0.46 (i.d.) cm) by a con-
ventional high-pressure slurry-packing technique using a Mod-
Carbamoylation was performed by the conventional proce-
dure.9
The α-glucose-carrying poly(phenyl isocyanide)
(0.39 g, 1.39 mmol) was dispersed in dry pyridine (8 mL), and
an excess amount of 3,5-dimethylphenyl isocyanate (1.22 mL,
8.3 mmol) was added to the suspended solution. The mixture
was then stirred at 80 °C for 72 h. The resulting phenylcar-
bamoylated product was isolated as a methanol–water (5:1,
v/v) insoluble part (1.20 g, yield 98%). 1H NMR (500 MHz,
Me2SO-d6) δ 4.1–4.9 (br, m, other protons from sugar), 5.1 (br,

A. Tsuchida et al.
Bull. Chem. Soc. Jpn., 75, No. 12 (2002) 2683
el CCP-085 Econo packer pump (Chemco, Osaka, Japan).
1,3,5-Tri-t-butylbenzene was used as a non-retained compound
to estimate the dead time (t0).
polyisocyanide-bearing acetylated and nonprotected saccha-
ride moieties become more slender and more stretched than the
one estimated from the 41 helices reported for the other poly-
isocyanides.11 This is probably due to steric repulsion among
the highly crowded, bulky saccharide chains. In the same way,
the more bulky phenylcarbamoyl substituents may make a he-
lical structure of these polymers more slender. In addition, it is
also speculated that the number of helical reversals in a poly-
mer chain may be increased; that is, the length of the one-
handed helical segment in the polymer would become shorter.
Green et al.12 pointed out that in a statistical thermodynamic
study of polyisocyanates, in principle, dynamic helix reversals
in the polymer chain can randomly occur in the backbone of
the polymer chain. The existence of such backbone motions
can also be postulated for the present polyisocyanide.
Chromatographic Enatioseparation. Figure 3 shows a
chromatogram of the resolution of racemic 1,2,2,2-tetraphen-
ylethanol (6) on CSP-1 (1, PPI(Glcα-R)). Enantiomers elute
at retention times of t1 and t2. The capacity factors, k1ꢀ [= (t1
− t0)/t0] and k2ꢀ [= (t2 − t0)/t0], were 0.39 and 0.62, respective-
ly. The separation factor, α [= k2ꢀ/k1ꢀ], and the resolution fac-
tor, Rs [= 2(t2 − t1)/(w1 + w2)], were found to be 1.59 and
1.83, respectively. The α value expresses the chiral recogni-
tion ability of a CSP, and the Rs value depends on both the α
value and the sharpness of the peaks. The band widths of the
first- and second-eluted enantiomers on the base line are
shown as w1 and w2. Rs can be determined when the two peaks
are almost completely separated. The chromatogram in Fig. 3
shows the expeditious complete separation of enantiomers
with rather sharp peaks. Table 1 summarizes the results of the
Apparatus. Chromatographic experiments were per-
formed on a JASCO PU-980 chromatograph equipped with a
UV (JASCO UV-970) and a polarimetric (JASCO OR-990) de-
tector at room temperature. A solution of a racemate
(1–10 µL) was injected into the chromatographic system with
a Rheodyne Model 7125 injector. A hexane-2-propanol (98/2)
mixture was used as the eluent at a flow rate of 0.5 mL/min,
unless otherwise specified.
Results and Discussion
Circular Dichroism (CD) Spectra. The CD spectra of
the 3,5-dimethylphenylcarbamates in THF are compared in
Fig. 2. The α-glycoside polymers 1 and 3 showed a negative
Cotton effect, whereas the β-glycoside polymers 2 and 4
showed a positive Cotton effect for the absorption due to phen-
yl isocyanide residues above 250 nm. Since the negative and
positive Cotton effects of polyisocyanides are respectively as-
signable to right-handed and left-handed helices,10 the helix
sense in these polymers may be regulated by the chirality of
the α- or β-anomeric center. Each intensity of the CD spectra
was smaller than those expected for the completely right- or
left-handed helices, indicating that these poly(phenyl isocya-
nide)s are a mixture of the right- and left-handed helical poly-
mers or sequences.
The helix senses of the present phenylcarbamoylated poly-
mers were the same as those of the acetylated precursors,4 al-
though the intensities [θ] of the present polymers became
slightly smaller than those of the precursors. Based on molec-
ular dynamics calculations, we assumed that the helices of the
Fig. 3. Separation of 1,2,2,2-tetraphenylethanol (6) on CSP-
1. Eluent: hexane/2-propanol (98/2); flow rate: 0.5 mL/
min.
Fig. 2. CD spectra of 3,5-dimethylphenylcarbamoylated gly-
copolyisocyanides in THF.

2684 Bull. Chem. Soc. Jpn., 75, No. 12 (2002)
Sugar-Carrying Poly(phenyl isocyanide)s
Table 1. Resolution on CSP of 3,5-Dimethylphenylcarbamate Derivatives of Rigid Poly(phenyl isocyanide)s (1–4) and of Flexible
Poly(N-phenylacrylamide) (5)a)
PPI(Glcα-R), 1
k1ꢀ
0.39 (+) 1.59 1.83
0.80 (−) 1.15 1.16
PPI(Glcβ-R), 2
PPI(Galα-R), 3
k1ꢀ
0.30 (−) ~1
1.59 (+) ~1
0.73 (+) ~1
2.43 (+) 1.09
0.17 (+) ~1
0.71 (+) 1.34 1.14
1.41 (+) 1.37 0.76
0.57 (−) 1.07
0.61 (+) 1.47 1.41
9.19 (+) 1.02
PPI(Galβ-R), 4
k1ꢀ
PAP(Glcα-R), 5
k1ꢀ Rs
0.73 (+) 1.14 0.72
2.65 (−) ~1
α
Rs
k1ꢀ
0.49
α
1.00
Rs
α
Rs
α
Rs
α
6
7
8
9
0.54 (−) ~1
1.95 (+) 1.07
1.02 (+) ~1
2.77 (−) 1.08
0.24 (+) 1.16
1.05 (+) 1.10
3.50 (+) 1.26
0.82 (−) ~1
1.62 (+) 1.24 1.29
0.90 (+) ~1
2.92 (−) 1.15
0.21 (−) ~1
0.38
1.56
1.00
1.00
0.92
1.22
~1
1.00
10 0.12 (−) ~1
0.26 (−) ~1
11 0.55
12 0.95
1.00
~1
1.11 (−) ~1
1.07
0.88
~1
~1
4.47b)
1.00
13 0.29 (+) ~1
14 0.29 (−) ~1
0.71 (−) ~1
0.48 (−) 1.13
0.79 (+) ~1
0.52 (+) ~1
2.17 (+) 1.22 0.77
0.70
4.03
~1
1.02
15 2.09
1.00
7.83b)
1.00
a) Eluent: hexane/2-propanol (98/2), flow-rate: 0.5 mL/min. The signs in parentheses represent the optical rotation of the first-elut-
ing enantiomers. b) Flow-rate: 1.0 mL/min.
Fig. 4. Structures of racemates 6–15.
enantioseparation of the following 10 racemates with function-
al groups (Fig. 4): 6, benzoin (7), 2-phenylcyclohexanone (8),
1-(9-anthryl)-2,2,2-trifluoroethanol (9), ( )-trans-stilbene ox-
ide (10), tris(acetylacetonate) cobalt(Ⅲ) (11), trans-cyclopro-
pane-1,2-dicarboxanilide (12), flavanone (13), Tröger’s base
(14), and 2,2ꢀ-dihydroxy-6,6ꢀ-dimethyl-1,1ꢀ-biphenyl (15).
The results were evaluated by the separation factor α.
The effects of the helical backbone on the enantioseparation
can be clearly demonstrated by comparing CSP-1 and CSP-5,
which have the same α-glucopyranoside structures along the
helical and flexible backbones, respectively. In the resolution
of racemate 6, 7, and 15 on CSP-1 and CSP-5, there existed a
distinct difference in the separation factors (α). The helical
CSP-1 almost completely separated racemates 7 and 6 with α
separation factors of 1.15 and 1.59, respectively. In contrast,
the non-helical CSP-5 only partially separated racemate 6 and
15 with α values of 1.14 and 1.22. The chiral recognition abil-
ity of CSP-1 is clearly different from that of CSP-5, which may
be attributable to the helical saccharide arrays along the regu-
lated poly(phenyl isocyanide) backbone.
rated some racemates with larger α and Rs; especially, the α
and Rs for racemate 6 on CSP-1 were the highest among the
combinations listed in Table 1. In contrast, the galactose-type
CSP-3 and CSP-4 showed chiral recognition ability for a
broader range of compounds than the glucose-type CSP-1 and
CSP-2; that is, six racemates were completely or partially re-
solved on both CSP-3 and CSP-4. Especially, CSP-3 separated
the largest number of racemates.
The chiral recognition ability also depended on the anomer-
ic configuration (α/β) of the sugar moieties; in other words, the
helical structure of the poly(phenyl isocyanide)s induced by
various sugars. By comparing the most efficient separations
on each CSP, it is found that the α-glycoside-type CSPs have
higher enantioseparation abilities than the β-glycoside-type
CSPs; the α value (1.59) for 6 on CSP-1 and those for 11
(1.34), 12 (1.37), and 14 (1.47) on CSP-3 were larger than
those for the corresponding racemates on CSP-2 and CSP-4.
Additionally, it was found that the elution order of some enan-
tiomers was reversed on these CSPs. For example, (−)-9 elut-
ed first on CSP-2 and CSP-4, but second on CSP-3, and (−)-14
eluted first on CSP-2, and second on CSP-3. Also, (−)-7 elut-
ed first on CSP-1 and (+)-7 on CSP-2. On the other hand, the
selectivities of enantiomers on CSP-1 and CSP-5 were similar.
The chiral recognition ability is dependent on the stereo-
chemistry (C1 and C4) of the sugar moieties in the poly(phen-
yl isocyanide)s (1–4). The glucose-type CSPs-1 and -2 sepa-

A. Tsuchida et al.
Bull. Chem. Soc. Jpn., 75, No. 12 (2002) 2685
References
a) M. Fukuda and O. Hindsgaul, “Molecular and Cellular
Glycobiology,” ed by Oxford University Press, Oxford, (2000). b)
R. A. Dwek, Chem. Rev., 96, 683 (1999).
Therefore, the selectivity of enantiomers on poly(phenyl isocy-
anides) seems to be mainly dependent on the conformation and
the anomeric configuration of the tethered monosaccharide de-
rivative.
The chiral discrimination mechanism of these glyco-
poly(phenyl isocyanide)s is still obscure. The discrimination
may be ascribed to the different chiral environment construct-
ed by the different sugar units. The phenylcarbamoylated
monosaccharide residues arranged along the main chain may
induce a prevailing one-handed helical structure, which can
contribute to enantioseparation.
The separation factors (α) on CSP-1 were smaller than those
on other CSPs for various racemates, except for 6. This may
be related to the lower one-handed helical structure of PPI(Gl-
cα-R) (1) than 2, 3, and 4, as suggested by the CD measure-
ment of the CSPs. The smaller CD intensity [θ] of 1 compared
with the others suggests that the obtained enantioseparation
data for various racemates may not show the inherent chiral
recognition ability of CSP-1. The present α-glucose carrying
polyisocyanide has helix reversals in the polymer backbone,
and the length of the one-sided helical segment in polymer
seems to be short. If the more regulated PPI(Glcα-R) (1) with
the one-handed helical sequence could be prepared, the CSP-1
might perform excellent enantioseparations. The future direc-
tion of this study will be an examination of appropriate condi-
tions for the more stereo-regulated polymerization and the sup-
pression of dynamic helix reversal in a polymer.
1
2
a) L. L. Kiessling and N. L. Pohl, Chem. Biol., 3, 71
(1996). b) K. Kobayashi, A. Kobayashi, and T. Akaike, “Methods
in Enzymology, Vol. 247, Neoglycoconjugates Part B,
Application,” ed by Y.C. Lee and R.T. Lee, Academic Press, San
Diego (1994), p. 409.
3
a) K. Kobayashi, A. Tsuchida, T. Usui, and T. Akaike,
Macromolecules, 30, 2016 (1997). b) A. Tsuchida, K. Kobayashi,
N. Matsubara, T. Muramatsu, T. Suzuki, and Y. Suzuki,
Glycoconjugate J., 15, 1047 (1998). c) H. Dohi, Y. Nishida, M.
Mizuno, M. Shinkai, T. Kobayashi, H. Uzawa, and K. Kobayashi,
Bioorg. Med. Chem., 7, 2053 (1999).
4
a) T. Hasegawa, S. Kondoh, K. Matsuura, and K.
Kobayashi, Macromolecules, 32, 6595 (1999). b) T. Hasegawa, K.
Matsuura, K. Ariga, and K. Kobayashi, Macromolecules, 33, 2772
(2000).
5
a) E. Yashima, J. Chromatogr., A, 906, 105 (2001). b) Y.
Okamoto and E. Yashima, Angew. Chem., Int. Ed., 37, 1021
(1998). c) E. Yashima, C. Yamamoto, and Y. Okamoto, Synlett,
1998, 344.
6
a) G. Blaschke, J. Liq. Chromatogr., 9, 341 (1986). b) Y.
Okamoto and T. Nakano, Chem. Rev., 94, 349 (1994). c) K. Saigo,
Prog. Polym. Sci., 17, 35 (1992). d) T. Nakano and Y. Okamoto,
Chem. Rev., 101, 4013 (2001). e) T. Nakano, J. Chromatogr., A,
906, 205 (2001).
In conclusion, the CSPs of the 3,5-dimethylphenylcarbam-
oyled glycosyl poly(phenyl isocyanide)s effectively separated
some enantiomers, depending on the saccharide structures.
The highly crowded saccharide arrays along the rigid helical
poly(phenyl isocyanide) backbones may play an important role
for chiral recognition. Their chiral separation ability would be
significantly enhanced if the one-handed helical poly(phenyl
isocyanide)s could be prepared.
7
A. Yamagishi, I. Tanaka, M. Taniguchi, and M. Takahashi,
J. Chem. Soc., Chem. Commun., 1994, 1113.
R. J. M. Nolte, J. A. J. van Zomerern, and J. W. Zwikker, J.
Org. Chem., 43, 1972 (1978).
Y. Okamoto, M. Kawashima, and K. Hatada, J.
8
9
Chromatogr., 363, 173 (1986).
10 R. J. M. Nolte, Chem. Soc. Rev., 1994, 11.
11 M. Clericuzio, G. Alagona, C. Ghio, and P. Salvadori, J.
Am. Chem. Soc., 119, 1059 (1997).
12 M. M. Green, J.-W. Park, T. Sato, A. Teramoto, S. Lifson,
R. L. B. Selinger, and J. V. Selinger, Alagona, C. Ghio, and P.
Salvadori, Angew. Chem., Int. Ed., 38, 3138 (1999).
One (A.T.) of us thanks for a Grant-in-Aid for JSPS Fellows
(No. 567) from the Ministry of Education, Science, Sports and
Culture.