Chiral ice chromatography

Article abstract of DOI:10.1021/ja1055214

Water-ice particles simultaneously doped with β-cyclodextrin and a salt enabled chromatographic separation of enantiomers without synthetic processes, and enhanced chiral recognition occurring in the liquid-water phase coexistent with the solid-ice phase.

Full text of DOI:10.1021/ja1055214

Published on Web 09/08/2010
Chiral Ice Chromatography
Taiki Shamoto, Yuiko Tasaki, and Tetsuo Okada*
Department of Chemistry, Tokyo Institute of Technology, Meguro-ku, Tokyo 152-8551, Japan
Received June 23, 2010; E-mail: tokada@chem.titech.ac.jp
Abstract: Water-ice particles simultaneously doped with ꢀ-cy-
clodextrin and a salt enabled chromatographic separation of
enantiomers without synthetic processes, and enhanced chiral
recognition occurring in the liquid-water phase coexistent with the
solid-ice phase.
Despite the fact that it is a common material, water ice has
attracted a number of researchers in a wide variety of scientific
fields because ice is related to various natural phenomena and plays
important roles in the global environment.1-4 Some atmospheric
reactions proceed in so-called cryogenic reactors, including the
quasi-liquid layer on the surface of ice and micropockets developed
at the triple junction of ice grain boundaries.4 It is also known that
some reactions are accelerated when confined in ice.5 These
examples suggest that a liquid-water phase in ice (WPI) acts as a
new class of reaction medium.
In contrast to extensive studies to reveal the characteristic features
of water ice, there have been very few attempts to utilize its unique
functionalities for productive developments in chemistry and related
disciplines. Ice chromatography, in which ice particles are used as
a chromatographic stationary phase, was first attempted by Dasgupta
and Mo.6 We have devised an effective way to prepare ice particles
suitable for this purpose and have revealed the interfacial chemistry
of water ice.7-9 In this method, the presence of the WPI allows
the mechanistic shift from adsorption mode to partition mode.
This shift is enhanced by doping a salt into the ice stationary
phase.9
This communication discusses the chiral recognition reaction in
the WPI studied using chiral ice chromatography measurements.
This approach not only reveals the suitability of the WPI as a
platform for the chiral recognition reaction but also will lead to
the preparation of a new class of chiral stationary phase. Chiral
chromatography constitutes a dominant sector of chiral separations
conducted in various branches of chemical, biological, pharmaceuti-
cal, and medical science.10,11 In most of the contemporary stationary
phases, chiral selectors are chemically bonded on a solid matrix
such as silica gel or polymer beads. Chiral ice chromatography may
allow us to skip the synthetic processes required for the preparation
of the current chiral solid phases.
In order to realize chiral ice chromatographic separation,
ꢀ-cyclodextrin (ꢀ-CD) was incorporated as a chiral selector into
an ice stationary phase throughout this work. The mobile phase
was hexane containing a small amount of tetrahydrofuran (THF);
the concentration of THF was adjusted for the retentivity of a
solute. In general, native ꢀ-CD is not suitable for chiral
separation in a nonpolar medium such as hexane because solvent
molecules are accommodated in the cavity of ꢀ-CD and interfere
with its chiral recognition.10 Various structures have been
introduced onto the CD to reduce this effect in commercialized
CD-bonded normal-phase chiral stationary phases. In the present
Figure 1. Chiral ice chromatographic separation of (A) hexobarbital and
(B) 1,1′-bis-2-naphthol with using (top) UV and (bottom) circular dichroism
detection. (A) Stationary phase, ice doped with 0.5 mM ꢀ-CD and 75 mM
KCl; mobile phase, 0.5% (v/v) THF in hexane; temperature, -8 °C;
detection at 230 nm. (B) Stationary phase, ice doped with 10 mM ꢀ-CD
and 100 mM KCl; mobile phase, 0.1% (v/v) THF in hexane; temperature,
-8 °C; detection at 280 nm.
system, a salt as well as ꢀ-CD is doped in ice to facilitate the
development of the WPI, in which the chiral recognition occurs;
native CD can thus be used. We previously studied chromatog-
raphy with salt-doped ice and found that the growth of the WPI
can be probed chromatographically.9 KCl or NaCl was selected
as a dopant in this work because the phase diagrams of the
water-KCl and water-NaCl systems are well-known. The
literature data for the phase boundaries in these systems were
fitted using an appropriate polynomial, as shown in Figures S1
and S2 in the Supporting Information. The temperature-molar
ratio diagrams were used for the calculation of the volume of
the WPI at a given temperature and salt concentration in the
entire system (i.e., its concentration before freezing), whereas
the temperature-molar concentration diagrams were utilized for
the estimation of the salt concentration in the WPI equilibrated
with the ice phase at a given temperature.
Figure 1 shows chiral separations of hexobarbital and 1,1′-bis-
2-naphthol using the ice stationary phase doped with both ꢀ-CD
and KCl. The circular dichroism spectrometric detection clearly
indicated that the enantiomers were successfully separated by the
doped ice stationary phase. The effects of the concentrations of
ꢀ-CD and the salt as well as the effect of temperature on the
retention of the enantiomers of hexobarbital are shown in Figures
S3-S5. We can summarize some characteristics of chiral ice
chromatographic separation from these figures: (1) the effect of
temperature is smaller than those of other factors; (2) doping with
a salt is essential for separation of the enantiomers (no separation
occurred with ꢀ-CD-doped ice that was not doped with salt); (3)
increasing the ꢀ-CD concentration enhances the chiral separation,
as expected.
The effects of the experimental variables on the chiral separation
can be comprehensively interpreted on the basis of the following
9
10.1021/ja1055214  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 13135–13137 13135

C O M M U N I C A T I O N S
model. The retention factor (k) of a solute in the present system is
represented by
k ) kads + kpar
(1)
where the contributions to k from adsorption on the ice stationary
surface and from partitioning into the WPI are given by kads and
kpar, respectively. When ꢀ-CD is incorporated into the ice stationary
phase, complexation of ꢀ-CD with a solute (S) occurs in the WPI;
this enhances kpar. Substitution of the volume of the WPI (VWPI),
the volume of the mobile phase in the column (Vmob), and the
concentrations of complexed and uncomplexed solute in the two
phases into eq 1 yields
Figure 2. Relation between the corrected ∆k for the enantiomers of
hexobarbital and the concentration of ꢀ-CD in the WPI in KCl-doped (O)
and NaCl-doped (4) ice stationary phases. The solid curve was obtained
by curve fitting and is given by y ) (1.97 × 103)[CD]WPI + (1.56 ×
105)[CD]2WPI. The dashed curve was drawn using the values determined for
bulk complexation.
VWPI [S]WPI + [S-CD]WPI + [S-(CD)2]WPI
k ) kads
+
+
Vmob
VWPI
[S]mob
kads
(1 + K1[CD]WPI + ꢀ2[CD]W2 PI
)
)
Kh/wVmob
(2)
curve, the chiral ice chromatographic retention data imply that ∆K1
) 1.97 × 103 M-1 and ∆ꢀ2 ) 1.56 × 105 M-2 for the enantiomers
of hexobarbital in the WPI. In contrast, as shown in Figure S7,
solvent extraction experiments at -5 °C gave ∆K1 ≈ 500 M-1 and
∆ꢀ2 ) 1.26 × 105 M-2. These values give the broken curve depicted
in Figure 2. An enhancement of ꢀ-CD chiral recognition in the
WPI relative to that in bulk solutions is suggested. Another point
that should be noted here is the solubility of ꢀ-CD in the WPI.
The solubility of ꢀ-CD in water is reported to be ∼16 mM but is
enhanced up to ∼70 mM in the presence of an electrolyte.12 We
found that its solubility in the hexane-saturated aqueous KCl
solution (1.1-2.8 M) was almost constant at 2-2.5 mM over the
temperature range -10 to 10 °C. If the concentration of ꢀ-CD in
the WPI were constant irrespective of the extent of its growth,
Kh/w∆k should be proportional to VWPI according to eq 3. However,
the ∆k-VWPI plot was entirely scattered and did not suggest any
trends. This strongly implies that the solubility of ꢀ-CD in the WPI
is enhanced in comparison with that in bulk solutions. Although
we attempted chiral separation using silica gel impregnated with
an aqueous ꢀ-CD solution, no separation was confirmed. This again
indicates the specific physicochemical properties of the WPI, which
plays an essential role in successful chiral separation in the present
scheme. The present results indicate that the WPI well dissolves
ꢀ-CD or excludes hexane and facilitates formation of the inclusion
complex of ꢀ-CD. If the WPI has properties similar to those of
high-density water, which is under debate in relation to the
polyamorphism of water, the present findings may be interpreted.13
Hence, the present method has several advantages. First, the
separation performance can be adjusted by (1) changing VWPI
through changing temperature and/or the salt dopant concentration
and (2) varying Kh/w through changing the organic solvent or
modifier. Second, the specific medium property of the WPI can
result in enhanced chiral separation; the function of the WPI and
its molecular origin should be revealed in more detail. Although
separation using ꢀ-CD has been discussed here, any water-soluble
chiral selector can be incorporated into the ice stationary phase in
a similar fashion. We believe that ice chromatography opens a new
era for both ice chemistry and chiral separation.
where K1 (M-1) and ꢀ2 (M-2) are the association constants for 1:1
(S-CD) and 1:2 [S-(CD)2] complexes between the solute and
ꢀ-CD, respectively, and Kh/w ) [S]mob/[S]WPI is the partition
coefficient of uncomplexed solute between the mobile phase and
the WPI phase.
The adsorption of a solute on the doped-ice stationary phase is
the result of several processes, namely, adsorption on the solid-ice
surface, adsorption on the surface of the WPI, and the interaction
with ꢀ-CD at these interfaces. It should be noted that these
mechanisms are not enantioselective; the first two have no enan-
tioselective origins, and the last one is also non-enantioselective
because the formation of inclusion complexes of ꢀ-CD is very weak
in the presence of a nonpolar solvent such as hexane, as discussed
above.10 The contribution from these processes to kads should vary
with experimental conditions such as the temperature and the
concentrations of the salt and ꢀ-CD. The quantitative discussion
of kads is not straightforward. However, since the enantioselectivity
in chiral ice chromatography appears only in the kpar term, we can
discuss the enantioselectivity of the ice stationary phase on the basis
of the difference between the enantiomer k values, which allows
the cancellation of the contribution from the non-enantioselective
kads term:
VWPI
∆k ) k2 - k1 )
(∆K1[CD]WPI + ∆ꢀ2[CD]W2 PI
)
Kh/wVmob
(3)
where ∆K1 and ∆ꢀ2 are the differences in the complexation
constants for the two enantiomers. Although the temperature
dependence of Kh/w is negligibly small, its dependence on the salt
concentration is so large that this effect must be taken into account
to interpret the separation data (Figure S6). Thus, we can correct
∆k in terms of VWPI, Vmob, and Kh/w. VWPI can be calculated from
the phase diagram, and Vmob can be determined from the void
volume of the column; Kh/w can be obtained from Figure S6 using
the salt concentration in the WPI at the working temperature.
Obviously, the corrected ∆k values are described by a quadratic
equation in [CD]WPI
.
Acknowledgment. This work was supported by a Grant-in-Aid
for Scientific Research from the Japan Society for the Promotion
of Science.
Figure 2 shows the relation between the corrected ∆k values
and [CD]WPI, which was calculated by assuming that ꢀ-CD added
in the stationary phase is completely dissolved in the WPI. Although
there is some scatter in the data, all of the corrected ∆k values
obtained with NaCl- and KCl-incorporated ice stationary phases
fall on the solid quadratic curve. From the coefficients of the fitting
Supporting Information Available: Experimental details and results
of chromatographic experiments. This material is available free of
charge via the Internet at http://pubs.acs.org.
9
13136 J. AM. CHEM. SOC. VOL. 132, NO. 38, 2010

C O M M U N I C A T I O N S
(8) Okada, T.; Tasaki, Y. Anal. Bioanal. Chem. 2010, 396, 221.
(9) Tasaki, Y.; Okada, T. Anal. Sci. 2009, 25, 177.
References
(1) Petrenko, V. F. ; Whitworth, R. W. Physics of Ice; Oxford University Press:
New York, 1999.
(10) Mitchell, C. R.; Armstrong, D. W. Methods Mol. Biol. 2004, 243, 61.
(11) Andersson, S. In Chiral Separation Techniques, 3rd ed.; Subramanian, G.,
Ed.; Wiley-VCH: Weinheim, Germany, 2007; pp 585-600.
(2) Sadtchenko, V.; Ewing, G. E. J. Chem. Phys. 2002, 116, 4686.
(3) Tasaki, Y.; Okada, T. J. Phys. Chem. C 2008, 112, 2618.
(4) Huthwelker, T.; Ammann, M.; Peter, T. Chem. ReV. 2006, 106, 1375.
(5) Takenaka, N.; Bandow, H. J. Phys. Chem. A 2007, 111, 8780.
(6) Dasgupta, P. K.; Mo, Y. Anal. Chem. 1997, 69, 4079.
(7) Tasaki, Y.; Okada, T. Anal. Chem. 2006, 78, 4155.
´
(12) Buva´ri, A.; Barcza, L. J. Inclusion Phenom. 1989, 7, 379.
(13) Suzuki, Y. Chem. Phys. Lett. 2001, 335, 357.
JA1055214
9
J. AM. CHEM. SOC. VOL. 132, NO. 38, 2010 13137