Enantiomeric separations of chiral sulfonic and phosphoric acids with barium-doped cyclofructan selectors via an ion interaction mechanism

Article abstract of DOI:10.1021/ac403686a

New cyclofructan-6 (CF6)-based chiral stationary phases (CSPs) bind barium cations. As a result, the barium-complexed CSPs exhibit enantioselectivity toward 16 chiral phosphoric and sulfonic acids in the polar organic mode (e.g., methanol or ethanol mobile phase containing a barium salt additive). Retention is predominantly governed by a strong ionic interaction between the analyte and the complexed barium cation as well as hydrogen bonding with the cyclofructan macrocycle. The log k versus log [X], where [X] = the concentration of the barium counteranion, plots for LARIHC-CF6-P were linear with negative slopes demonstrating typical anion exchange behavior. The nature of the barium counteranion also was investigated (acetate, methanesulfonate, trifluoroacetate, and perchlorate), and the apparent elution strength was found to be acetate > methanesulfonate > trifluoroacetate > perchlorate. A theory based upon a double layer model was proposed wherein kosmotropic anions are selectively adsorbed to the cyclofructan macrocycle and attenuate the effect of the barium cation. van't Hoff studies for two analytes were conducted on the LARIHC-CF6-P for three of the barium salts (acetate, trifluoroacetate, and perchlorate), and the thermodynamic parameters governing retention and enantioselectivity are discussed. Interestingly, for the entropically driven separations, enantiomeric selectivity can increase at higher temperatures, even with decreasing retention.

Full text of DOI:10.1021/ac403686a

Article
pubs.acs.org/ac
Enantiomeric Separations of Chiral Sulfonic and Phosphoric Acids
with Barium-Doped Cyclofructan Selectors via an Ion Interaction
Mechanism
Jonathan P. Smuts,† Xin-Qi Hao,‡,§ Zhaobin Han,‡ Curran Parpia,† Michael J. Krische,‡
and Daniel W. Armstrong*,†
†Department of Chemistry and Biochemistry, University of Texas at Arlington, TX 76019, United States
‡Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, Texas 78712, United States
§College of Chemistry and Molecular Engineering, Zhengzhou University, Zhengzhou 450052, P. R. China
S
* Supporting Information
ABSTRACT: New cyclofructan-6 (CF6)-based chiral stationary phases
(CSPs) bind barium cations. As a result, the barium-complexed CSPs
exhibit enantioselectivity toward 16 chiral phosphoric and sulfonic acids
in the polar organic mode (e.g., methanol or ethanol mobile phase
containing a barium salt additive). Retention is predominantly governed
by a strong ionic interaction between the analyte and the complexed
barium cation as well as hydrogen bonding with the cyclofructan
macrocycle. The log k versus log [X], where [X] = the concentration of
the barium counteranion, plots for LARIHC−CF6-P were linear with
negative slopes demonstrating typical anion exchange behavior. The
nature of the barium counteranion also was investigated (acetate,
methanesulfonate, trifluoroacetate, and perchlorate), and the apparent
elution strength was found to be acetate > methanesulfonate >
trifluoroacetate > perchlorate. A theory based upon a double layer model was proposed wherein kosmotropic anions are
selectively adsorbed to the cyclofructan macrocycle and attenuate the effect of the barium cation. van’t Hoff studies for two
analytes were conducted on the LARIHC−CF6-P for three of the barium salts (acetate, trifluoroacetate, and perchlorate), and
the thermodynamic parameters governing retention and enantioselectivity are discussed. Interestingly, for the entropically driven
separations, enantiomeric selectivity can increase at higher temperatures, even with decreasing retention.
separation of sulfonic acid enantiomers is important to this
he separation of chiral phosphoric and sulfonic acid
Tenantiomers is of great importance to synthetic and field.
analytical chemists. The independent work of Akiyama1 and
Cyclofructan 6 (CF6) is a cyclic oligosaccharide produced by
Terada2 in developing BINOL-phosphates to catalyze enantio-
selective C−C bond formation reactions in 2004 set the stage
for asymmetric catalysis in a significant way.3 Since then, great
strides have been made in developing asymmetric catalysis for
many reactions, including organocatalytic aryl−aryl bond
formation,4 spiroketalizations,5 Mannich reactions,6 hydro-
cyanation of hydrazones,7 reductive amination,8 and in
crotylation reactions.9,10 Chiral sulfonic acids, on the other
hand, while also being strong Brønsted acids, have not enjoyed
the same success in asymmetric synthesis but have rather been
the focus of chiral-resolving agents in the separation of basic
compounds. Recently, Kellog et al. broadened the concept of
diastereomeric recrystallization from using a single chiral
selector for purifying basic racemates to simultaneously using
a family of enantiomers.11 This technique has been dubbed
“Dutch resolution” and preferably requires three enantiomeri-
cally pure sulfonic acids from the same family (e.g., R-
camphorsulfonic acid and R-bromocamphorsulfonic acid) to
recrystallize basic compounds.11−13 Thus, the synthesis and
the enzymatic digestion of inulin. It consists of an 18-crown-6
core, spiro-annulated with D-fructofuranose pendant groups.
Since the discovery of cyclofructans in 1989 by Kawamura et
al.,14 the interest in these macrocycles slowly increased.
Research was primarily focused on assessing the ionophoric
ability of CF6,15−18 and nothing was conducted from a
chromatographic perspective. In contrast to related cyclic
oligosaccharides, the cyclodextrins, and even to synthetic crown
ethers, cyclofructans have yet to reach the cusp of their research
potential. In 2009, we published the first major chiral
separations work with cyclofructans.19 Subsequently, this
work has been further developed and greatly expanded.20−28
It was clear that native CF6 bonded to silica gel was usually a
poor chiral selector. However, upon derivitization with an
isopropyl carbamate group, it exhibited pronounced and broad
Received: November 13, 2013
Accepted: December 17, 2013
Published: December 27, 2013
© 2013 American Chemical Society
1282
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
Figure 1. Structures of the chiral sulfonic and phosphoric acids used in this study. See text for the names of the structures.
selectivity for primary amines in organic and supercritical fluid
mobile phases.19 However, this selectivity did not extend to
anionic or acidic compounds. It is known that the ionophoric
proclivity of CF6 is particularly pronounced for barium and
lead cations.29 Thus, the question arises as to whether this
cation affinity might be exploited somehow to induce a different
kind of enantioselectivity. This question is interesting from
both a mechanistic point of view and from the practical
standpoint that these CSPs have excellent preparative
capabilities. Herein, we describe the first highly effective
metal ion interaction approach for chiral separations since the
classic ligand exchange work of Davankov.30,31 However, the
nature of the metal ion, the mechanism, and the selectivity in
this work are shown to be unique. More specifically, this
transforms a neutral chiral selector into one with a great affinity
for anionic (sulfonic and phosphoric acid) chiral compounds. It
should be noted that the analytes in this work are retained
either by: (1) ion pairing with the barium cations in the mobile
phase or (2) by dynamic anion exchange (i.e., where the
cations, Ba2+ in this case, are so strongly bound that they can be
considered an integral part of the stationary phase) or (3) a
combination of both ion pairing and dynamic anion exchange.
Historically, the term ion-interaction chromatography has been
coined to encompass but not distinguish between all of the
aforementioned possibilities, particularly in reversed-phase
chromatography.32
stationary phases (CSPs) .33,34 Thus far we know of no other
enantioselective anion-exchange-type CSPs commercially avail-
able, which makes this an important area of research.
EXPERIMENTAL SECTION
■
Barium acetate (Ba(OAc)2), barium hydroxide monohydrate,
barium perchlorate (Ba(ClO4)2) (NOTE: all barium salts, with
the exception of barium sulfate, are toxic and need to be
handled with care), ammonium acetate (NH4OAc), ammonium
trifluoroacetate (NH4TFA), triethylamine (TEA), trifluoro-
acetic acid (TFA), acetic acid (HOAc), methane sulfonic acid
(MSA), ethylene diamine, (S)- and (R)-(−)-VAPOL hydrogen-
phosphate (CPA-1), (S)- and (R)-(−)-1,1′-binaphthyl-2,2′-diyl
hydrogenphosphate (CPA-2), (S) and (R)-(−)-3,3′-bis-
(triphenylsilyl)-1,1′-binaphthyl-2,2′-diyl hydrogenphosphate
(CPA-3), (S)- and (R)-3,3′-bis[3,5-bis(trifluoromethyl)-
phenyl]-1,1′-binaphthyl-2,2′-diyl hydrogenphosphate (CPA-4),
(S)- and (R)-(+)-2-hydroxy-4-(2-methoxyphenyl)-5,5-dimeth-
yl-1,3,2-dioxaphosphorinane 2-oxide (CPA-5), (S)- and (R)-
(+)-4-(2-chlorophenyl)-2-hydroxy-5,5-dimethyl-1,3,2-dioxa-
phosphorinane 2-oxide (CPA-6), camphor sulfonic acid (CSA-
6), trans-chalcone, 4-chlorochalcone, and 4-nitrochalcone were
purchased from Sigma-Aldrich (St. Louis, MO). The total
syntheses for compounds CSA-1−CSA-5 and CPA-7 together
1
with the H- and 13C NMR, ESI-MS, and FT-IR spectra are
available in the Supporting Information. Compounds CSA-7−
CSA-9 were synthesized from the corresponding chalcones
according to literature procedures.11,33 The chiral sulfonic and
phosphoric acids (see Figure 1) may be further subdivided as
(1) chiral aryl sulfonic acids with either one (CSA-1 and CSA-
In previous work, Lindner et al. have shown using methanol
mobile phases with acid/base additives that excellent
separations of α,β, and γ-aminophosphonic acids and chiral
sulfonic acids (CSAs) can be attained on cinchona alkoid chiral
1283
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
4) or two (CSA-2, CSA-3, and CSA-5) chiral axes, (2) β-
ketosulfonic acids (CSA-7 − 9), and (3) γ-ketosulfonic acid
(CSA-6). The CPAs may also be further classified as (1) biaryl
phosphoric acids (CPA-1 − 4), (2) cyclic phosphoric acids
(CPA-5 and 6), and (3) acyclic phosphoric acid (CPA-7).
Barium trifluoroacetate (Ba(TFA)2) was prepared by neutraliz-
ing barium hydroxide monohydrate with an excess of
trifluoroacetic acid and then allowed to evaporate to dryness.
The resulting residue was further dried in a vacuum oven
overnight and used without further purification. When not in
use, the material was stored in a desiccator. Barium
methanesulfonate was prepared in an analogous way using
methanesulfonic acid and barium hydroxide. HPLC grade
methanol (MeOH), ethanol (EtOH), acetonitrile (ACN), and
heptane (Hept) were purchased from VWR (Sugarland, TX).
Water was purified by a Milli-Q-water purification system
(Millipore, Billerica, MA).
HPLC. The HPLC used was an Agilent 1200 HPLC (Agilent
Technologies, Palo Alto, CA) consisting of a diode array
detector, a temperature-controlled column chamber, auto
sampler, quaternary pump, and fraction collector. Data
acquisition and analysis was controlled by ChemStation
software (Rev. B.03.02[341], Agilent Technologies 2001−
2008) in Microsoft Windows XP Professional OS. Unless stated
otherwise, all HPLC separations were carried out at 25 °C with
an injection volume of 5 μL and a flow rate of 1.0 mL/min
(isocratic). The following UV wavelengths were monitored:
230, 254, 265, 280, and 287 nm.
The cyclofructan loading and phase ratio, ϕ, where ϕ = Vs/Vm
(Vs = volume of the selector, Vm = volume of the mobile phase)
was required for the determination of entropy from the ln k
versus 1000/T plot and was calculated as published in the
literature.35,36 The values for the density of the LARIHC−CF6-
P stationary phase and the unbonded chiral selector were
measured by pycnometry (10 mL pycnometer), which was
calibrated gravimetrically with DI water.
Barium Loading and Removal. Analytical columns were
preconditioned with a 5 mM Ba(OAc)2 or Ba(TFA)2 solution
containing 20% methanol by volume for 30 min. The
preparative FRULIC-N column was preconditioned overnight
with a 40 mM Ba(OAc)2 solution containing 20% methanol by
volume. The columns were then either evaluated directly in the
polar organic mode or stored in methanol.
The barium loading was estimated by the procedure
described by Durmaz et al.37 Briefly, untreated columns were
conditioned with water and then methanol, 10-column volumes
each. An HPLC method was developed where the column was
washed with methanol at 0.5 mL/min for 5 min and then
switched to a 5 mM Ba(OAc)2 solution in methanol at 5.01
min at 0.5 mL/min. The method was continued until the
breakthrough point (monitored at 204 and 210 nm) was
observed. To regenerate the column, barium cations were
eluted from the stationary phase using a 100 mM NH4OAc
solution (pH = 4.1) for 30 min. Barium elution was monitored
until the effluent showed no sign of turbidity upon addition of a
few drops of concentrated sodium sulfate solution.
Enantiomeric separations were evaluated and optimized on
barium-doped LARIHC−CF6-P (isopropyl carbamate), LAR-
IHC−CF6-RN (R-1-(1-naphthyl)ethylcarbamate), and FRUL-
IC-N (native cyclofructan) chiral CSPs 250 mm × 4.6 mm × 5
μm (Azyp, LLC, Arlington, TX) in the polar organic mode.
The polar organic mode (sometimes referred to as the polar
ionic mode) mobile phase consisted of methanol (or ethanol)
containing a barium salt and when necessary, varying mixtures
of acidic (HOAc, TFA, or MSA) and basic (TEA, ethylene
diamine) additives or ammonium salts. The void time, t0, was
measured by injecting 1,3,5-tri-tert-butylbenzene and moni-
tored at 254 nm. All analytes were dissolved in methanol at
∼1−2 mg/mL and stored in a freezer when not in use. CSA-6,
however, was prepared in the 8−10 mg/mL range in water on
account of its low UV absorbance and was monitored at 287
nm. All mobile phases were degassed by sonication under
vacuum. Columns had to be conditioned by circulating the
mobile phase overnight.
RESULTS AND DISCUSSION
■
While attempting to resolve enantiomers of CSA-5 by fractional
recrystallization, simultaneous development of an HPLC
method was undertaken to more accurately monitor the
enantiomeric excess. Initially, a partial separation (Rs = 1.2) was
obtained on the LARIHC−CF6-P in the reversed-phase mode
(see Figure 2A). This separation could not be optimized
further, though different acidic and basic additives as well as
ammonium salts were explored (not shown). Finally, knowing
that cyclofructans bind barium ions strongly, 5 mM Ba(OAc)2
was added to the mobile phase with the hope that it might
induce better enantioselectivity for CSA-5. However the
presence of this additive in the reversed-phase mode worsened
Preparative HPLC. Preparative HPLC was performed on a
Jasco 2000 series HPLC using a FRULIC-N column (Azyp,
LLC, Arlington, TX; 250 mm × 21 mm × 5 μm) treated with
barium (software: Jasco ChromNav version 1.17.01). The
pump (PU-2086) was set at 20 mL/min with the mobile phase
consisting of methanol with 0.1% TFA and 0.1% TEA.
Detection was monitored at 280 nm with a high-pressure
UV−vis VWD (UV-2075) cell. The enantiomeric fractions
were manually collected using the Jasco SCF-Vch-Bp 6-valve
change unit. Sample injection was performed using an
autosampler (AS-2059-SFC) with a 1 mL injection loop in
the partial fill loop mode (injection volume =800 μL).
van’t Hoff Plots. The van’t Hoff experiments were
conducted at 10, 20, 30, 40, and 50 °C in a thermostatted
column chamber equipped with a Peltier cooler. When the
temperature was changed, the column was allowed to
equilibrate for at least 30 min before injecting the samples.
Figure 2. Original separation of CSA-5 on LARIHC−CF6-P. Mobile
phase: (A) 80/20/0.1/0.1 (v/v %) water/MeOH/TEA/HOAc; (B) 5
mM Ba(OAc)2 in 80/20 (v/v %) water/MeOH; (C) 50/50/0.1/0.1
ACN/MeOH/TEA/TFA; flow rate: 1 mL/min; detector: UV at 254
nm.
1284
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
the selectivity and efficiency (see Figure 2B). When the normal
phase was restored, 80/20/0.1 heptane/EtOH/TEA (not
shown), CSA-5 was retained longer than before with strong
peak tailing. Subsequently, the polar organic mode was used,
and an excellent separation was obtained (Rs = 2.4) with
superior efficiency and peak shape compared to all other
approaches (see Figure 2C). The separation was easily scaled to
preparative amounts on the FRULIC-N (see Figure 3). When
Table S-1 (see Supporting Information) shows the clear
effect of barium treatment on the cyclofructan CSPs for all of
the analytes in Figure 1. The untreated column was evaluated in
the polar organic mode using methanol containing 10 mM
NH4OAc. Under these conditions, all analytes were eluted as
sharp peaks close to the void volume with no enantioselectivity.
If no additive was present in the mobile phase, analytes were
still poorly retained but with poor peak efficiency. On a barium
pretreated column and in the presence of a barium additive,
however, there was a significant increase in retention often
accompanied by an increase in enantioselectivity. Addition of
low levels of barium salt to the mobile phase was necessary to
maintain a constant level of barium on the stationary phase and
to afford highly reproducible elution volumes and good peak
efficiency. The order of retention strength for the three Ba2+-
doped columns, in this particular mobile phase, was FRULIC-N
> LARIHC−CF6-RN > LARIHC−CF6-P. Retention is
governed by (1) the loading of cyclofructan, which also dictates
the amount of barium that can be loaded, (2) the nature of the
selector, and (3) the nature of the analyte. The cyclofructan
loading on the CSPs was determined to be 0.23, 0.12, and 0.18
mmol/g, respectively. Thus, even though the LARIHC−CF6-
RN stationary phase had somewhat lower overall loading than
the LARIHC−CF6-P, its ability to utilize π−π interactions with
aromatic analytes resulted in greater overall retention in many
cases.
Effect of the Counteranion of the Ba2+ Salt. The
barium-doped stationary phases in this study can be treated as
anion-exchange phases. However, to prevent slow leaching of
the barium from the stationary phase and to maintain its level
of saturation, a small amount of a barium salt is put into the
mobile phase. Since a small amount of barium salt is in the
mobile phase, the question arises as to the effect, if any, of its
counteranion. It is known that retention in ion-exchange
systems is believed to be a combination of selective adsorption,
ion pairing, and accumulation in the diffuse layer.38 As has
already been shown (Table S-1), in the absence of barium
Figure 3. Preparative HPLC separation of racemic CSA-5 on 21.2 mm
× 250 mm FRULIC-N treated with barium. Mobile phase: MeOH
with 0.1/0.1 (v/v %) TEA/TFA; flow rate: 20 mL/min; UV: 280 nm;
injection volume: 800 μL; sample concentration: 48 mg/mL.
the barium is eluted from the stationary phase, there is no
retention or enantioselectivity for CSA-5 enantiomers in the
polar organic mode. When the column was reconditioned with
a barium solution, retention and enantioselectivity were
restored.
Table 1. Comparison of Different Barium Salt Additives (∼ 5 mM in Each Mobile Phase) on the Retention, Selectivity, and
a
Resolution of Chiral Sulfonic/Phosphoric Acids on LARIHC−CF6-P in the Polar Organic Mode
5.0 mM Ba(OAc)2
in MeOH
5.0 mM Ba(ClO4)2
in MeOH
5.0 mM Ba(MSA)2
in MeOH
5.2 mM Ba(TFA)2
in MeOH
5.2 mM Ba(TFA)2 in EtOH
analyte
k
α
Rs
k
α
Rs
k
α
Rs
k
α
Rs
k
α
Rs
CSA-1
CSA-2
CSA-3
CSA-4
CSA-5
CSA-6
CSA-7
CSA-8
CSA-9
CPA-1
CPA-2
CPA-3
CPA-4
CPA-5
CPA-6
CPA-7
0.57
0.57
0.54
0.90
0.68
0.76
0.71
0.67
0.66
0.58
0.91
0.38
0.20
1.21
1.08
0.28
1.00
1.00
1.00
1.00
1.11
1.07
1.00
1.00
1.00
1.09
1.00
1.00
1.00
1.00
1.00
1.00
<0.2
<0.2
<0.2
<0.2
1.0
4.48
4.10
4.18
5.29
4.86
3.68
6.12
5.55
5.42
2.35
8.40
1.69
1.00
15.95
12.93
1.65
1.04
1.02
1.00
1.04
1.28
1.24
1.00
1.00
1.03
1.24
1.03
1.00
1.05
1.03
1.02
1.00
0.6
<0.5
<0.2
0.7
1.92
1.76
1.86
2.15
2.06
1.46
2.34
2.26
2.35
1.56
3.04
1.53
0.92
3.96
3.59
1.27
1.00
1.00
1.00
1.00
1.16
1.12
1.00
1.00
1.00
1.11
1.00
1.00
1.00
1.03
1.00
1.00
<0.4
<0.2
<0.2
<0.4
2.4
2.89
2.73
2.82
3.30
3.20
2.29
3.66
3.42
3.37
1.95
4.84
1.70
1.02
7.18
6.32
1.47
1.03
1.03
1.00
1.03
1.21
1.15
1.00
1.00
1.02
1.16
1.00
1.00
1.00
1.04
1.00
1.00
0.4
0.4
3.94
2.35
2.92
4.11
2.91
2.79
6.28
6.13
6.91
2.16
8.44
2.23
2.25
11.64
10.83
2.83
1.06
1.10
1.00
1.06
1.22
1.25
1.00
1.00
1.04
1.37
1.09
1.18
1.19
1.08
1.00
1.10
0.6
0.7
<0.2
0.4
<0.2
0.6
4.2
3.3
1.8
0.4
1.8
1.3
1.9
1.6
<0.2
<0.2
<0.2
0.7
<0.2
<0.2
0.5
<0.2
<0.2
0.4
<0.2
<0.2
0.4
<0.2
<0.2
0.5
3.4
1.5
2.3
3.1
<0.2
<0.2
<0.2
<0.2
<0.2
<0.2
0.5
<0.2
<0.2
<0.2
0.5
<0.4
<0.2
<0.4
0.6
1.1
<0.2
0.6
2.0
2.8
<0.4
<0.4
<0.2
0.6
<0.2
<0.2
<0.2
<0.2
<0.2
1.4
a
−
Mobile phase: methanol (or ethanol) with 5 mM BaX2 where X = acetate (OAc−), perchlorate (ClO4 ), methane suflonate (MSA−), and
trifluoroacetate (TFA−).
1285
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
cations, analytes elute at the dead volume or are poorly retained
on these stationary phases in the polar organic mode. However,
they are retained on the same stationary phases that have been
doped with barium. The effect of three other barium salts
(methanesulfonate, perchlorate, and trifluoroacetate) on
retention and enantioselectivity was explored using the
LARIHC−CF6-P column (Table 1). It is clear that the order
of elution strength for the four counteranions is: OAc− > MSA−
potential, ψ0,Ads, by the adsorption of acetate anions to the
cyclofructan macrocycle resulting in decreased retention of
analytes. In extreme cases, this selective adsorption can lead to
a reversal of the surface potential.38 Interestingly, this behavior
was observed when oxalic acid/TEA and phosphoric acid/TEA
were evaluated as mobile phase additives with no barium
present in the mobile phase. The result (data not shown) was
that chiral phosphoric and sulfonic acids eluted at the void
volume (exclusion) and racemic 1-(1-naphthyl)ethylamine was
retained and separated! Selective adsorption to the stationary
phase particle will lessen as the nature of the anion becomes
more chaotropic as with perchlorate.
Effect of the Mobile Phase Dielectric Constant. Also in
Table 1, we note the important role that the dielectric constant
of the solvent plays. Methanol has a dielectric constant of 32.7,
and ethanol has a value of 24.5. The replacement of methanol
for ethanol in Table 1, while using the same additive, 5.2 mM
Ba(TFA)2, shows (i) an increase in retention for every
compound except CSA-2 and CSA-5, (ii) an improvement in
the enantioselectivity, and (iii) additional enantioselectivity that
did not occur in the methanol mobile phase (CPA-2−4 and 7).
The number of compounds exhibiting enantioselectivity when
changing from methanol to ethanol is improved from 8 to 12,
respectively. The rationale for these improvements is that a
lower solvent dielectric constant affords tighter ion pairing with
the barium cation and therefore longer retention. Tighter ion
pairing necessitates being closer to the chiral macrocycle, thus
enhancing enantioselectivity as well. Although ethanol was the
preferable polar organic solvent, the kosmotropic barium salts
(OAc− and MSA−) were insoluble in it. The use of even lower
dielectric constant systems was investigated for poorly retained
compounds such as CPA-3 and CPA-4 by using heptane−
ethanol mobile phases. This improved both the retention and
the enantioselectivity; however, it caused excessive retention of
the other analytes.
−
> TFA− > ClO4 . As retention increases, the number of
analytes that show selectivity (Rs > 0.4) also increases as
−
follows: 11 compounds for ClO4 , 8 compounds for TFA−, 4
compounds for MSA−, and 3 compounds for OAc−. However,
the observed enantioselectivities are not simply a function of
retention. Figure 4 shows the separation of CSA-5 and CPA-1
Figure 4. Overlay of four chromatograms on LARIHC−CF6-P. In
each case, the first pair of enantiomers corresponds to CPA-1 and the
latter to CSA-5; flow rate: 1 mL/min, UV = 230 nm. Mobile phase
conditions and enantioselectivity: (A) 5.11 mM Ba(ClO4)2 in
methanol, αCPA‑1 = 1.25, αCSA‑5 = 1.29; (B) 4.16 mM Ba(TFA)2 in
methanol, αCPA‑1 = 1.16, αCSA‑5 = 1.21; (C) 2.00 mM Ba(MSA)2 in
methanol, αCPA‑1 = 1.12, αCSA‑5 = 1.18; (D) 1.00 mM Ba(OAc)2 in
methanol, αCPA‑1 = 1.09, αCSA‑5 = 1.12.
Retention Order. In methanol mobile phases containing
using the four different barium salts at concentrations that
afford similar retention. ClO4 has a profound effect on the
MSA−, TFA−, and ClO4 (Table 1), the retention order on
−
−
LARIHC−CF6-P appeared to follow the geometry and
accessibility of the anionic moiety. The more sterically hindered
the ionizable group of the analyte was, the less it was retained.
The most retained compounds, CPA-2, CPA-5, and CPA-6, are
among the smallest and least sterically hindered of the CPAs,
and they are retained the longest. Conversely, the largest
analytes (CPA-3, CPA-4, and CPA-7) are the least retained,
presumably because of the steric hindrance of the phosphate
group.
enantioselectivity for these two compounds that is clearly
greater than TFA−, MSA−, and OAc−.
Moyer et al. used the phrase “Hofmeister effect or bias” in
their research on cesium extraction by crown ethers.39 The
phrase was used to describe the observation that in liquid−
liquid extraction systems, the partitioning of cesium improved
when chaotropic counteranions were present. It appears that
there are analogous effects or trends in these chromatographic
results. The well-known Hofmeister series ranks the counter-
anions used in this publication as: OAc− > MSA− > TFA− >
ClO4− with OAc− being the most kosmotropic (water structure
Effect of Varying Concentration of the Ba2+ Salt in the
Mobile Phase. The retention of CPA-1 and CSA-5 was
studied as a function of the barium salt concentration by
systematically changing it from 1−5 mM for each of the barium
salts (∼1.0, 2.0, 3.0, 4.0, and 5.0 mM). Plots of the logarithm of
the retention factor for the first eluting enantiomer (log k)
versus the logarithm of the concentration of the counteranion
of the barium salt (log [X], where X = the barium salt
−
or hydrogen bond making) and ClO4 the most chaotropic
(water structure or hydrogen bond breaking).39,40 While these
terms, water structure making or breaking, may seem to have
no direct bearing on the polar organic mobile phase used here,
other recent publications have indicated Hofmeister effects in
nonaqueous solvents or aqueous organic mixtures.41,42 Indeed,
the current trend for explaining the Hofmeister series for
anions is not in terms of their effect on bulk water structure but
rather their direct interaction with the interface of the
substrate.43
−
counteranion, OAc−, MSA−, TFA−, and ClO4 ) were all linear
(r2 > 0.99) according to the equation
log k = −slog[X] + constant
(1)
We postulate that kosmotropic anions, such as acetate, will
have a tendency to selectively adsorb to cyclofructan through its
many hydroxyl groups. The effect of such a selective adsorption
is illustrated in Scheme 1. The original surface potential of the
barium-complexed cyclofructan, ψ0, is lowered to a new surface
where k is the retention factor, s is the slope which is usually
found to be approximately equal to zA/zX, where zA is the
charge of the analyte anion and zX is the charge of the barium
counteranion. The negative slopes (see Table S-2) indicated a
decreasing retention time with increasing barium salt
1286
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
a
Scheme 1. Effect of Specific Adsorption of Eluent Anions on the Potential Curve of a Charged Surface
a
The blue sphere represents a charged object or particle. The shaded, hemispherical area with dashed boundary is the Stern layer. The ovoid shapes
with a negative charge indicate eluent anions, and the spheres containing a positive charge indicate cations. The boundary to the diffuse layer is not
shown. (A) No selective adsorption of anions to the charged surface, potential is ψ0. (B) Selective adsorption of anion to the charged surface, ψ0,Ads
ψ0.
<
concentration typical of that found in ion-exchange chromatog-
raphy.33,38,44 In classical ion-pairing techniques, analyte
retention initially increases with increasing concentration of
ion-pairing reagent up to a maximum and thereafter decreases.
Assuming linear behavior, a positive slope for log k versus log
[X] should be seen for the former case and a negative one for
the latter. At greater barium salt concentrations, where
solubility permitted (e.g., 100 mM Ba(ClO4)2), retention
continued to decrease. When the concentration of the barium
salt in the mobile phase is much greater than that on the
stationary phase, the analytes predominantly ion pair in the
mobile phase and are not retained. Retention can then be
increased by decreasing the barium salt concentration to afford
separation.
Lamb et al. have shown in their work on C18 reversed phase
columns coated with a lipophilic crown ether/cryptand using a
KOH hydro-organic mobile phase that below ∼0.8 mM KOH,
retention increased with increasing KOH concentration, but
thereafter it decreased linearly.45 Such a turning point must
exist for our stationary phase as well but we did not observe it.
Fritz has observed that some ion-pairing agents can be attached
so strongly to the stationary phase that they may be considered
as permanently coated.46 In our case, this seems to have some
merit because the binding of Ba2+ to cyclofructan is stronger in
a pure organic solvent as is also common for crown ethers in
general.29,47
cyclofructan complex must be addressed. As with the binding of
alkali and alkaline earth metals to synthetic crown ethers and
cryptands, the binding of barium to cyclofructan is an
exothermic process.47−50 Although there is no quantitative
thermodynamic data on the binding of barium to cyclofructan,
it is nevertheless undisputed that it binds strongest to
cyclofructan compared to the rest of the Group I and II
metal cations.29 Takai and Sewada have generated thermody-
namic data for the binding of potassium, rubidium, and cesium
cations to permethylated cyclofructan and have shown that
ΔHbinding becomes increasingly exothermic.15 We therefore
infer that an analogous enthalpy will be exhibited for barium
binding to cyclofructan derivatives. A question arises as to how
much barium will be decomplexed from the cyclofructan
macrocycle as the equilibrium shifts to dissociation with
increasing temperature. The van’t Hoff equation expresses the
changing equilibrium constant as a function of temperature as
follows49
⎛
⎞
K2
K1
−ΔH
R
1
1
ln
=
−
⎜
⎝
⎟
⎠
T
T
(2)
2
1
where K1 and K2 represent the equilibrium constants at
absolute temperatures T1 and T2, respectively, ΔH is the
change in enthalpy (in J·mol−1), and R is the gas constant
(8.314 J·K−1·mol−1). Thus, if ΔH, K1, and T1 are known, the
equilibrium constant, K2, at some other temperature, T2, can be
readily calculated. Furthermore, the ratio of Ba2+−CF6 complex
van’t Hoff Plots. Before delving into the thermodynamics
of the chromatography, the thermodynamics of the barium−
1287
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
on the stationary phase to uncomplexed CF6 on the stationary
phase may be expressed as follows45
The retention and enantioselectivity of CPA-1 and CSA-5 on
LARIHC−CF6-P were monitored as a function of mobile
phase temperature according to the well-known van’t Hoff
expressions for chromatography
[Ba2+···CF6]S.P.
= K[Ba2+
]
M.P.
[CF6]S.P.
(3)
ΔH
RT
ΔS
R
ln k = −
+
+ ln ϕ
where K is the equilibrium constant, [Ba2+]M.P. represents the
concentration of barium in the mobile phase, [Ba2+···CF6]S.P.
represents the concentration of the barium−CF6 complex on
the stationary phase, and [CF6]S.P. represents the concentration
of CF6 on the stationary phase. We chose a K value for Ba2+−
CF6 of 19 000 M−1 taken from Takai and Sewada et al.,15,16
which is in agreement with research done by Na and
Padivitage.29 For ΔH, we used the value reported by Takai
and Sewada for cesium complexed to permethylated CF615
(i.e., ΔH = −38 kJ.mol−1). The calculations for eqs 2 and 3 are
represented graphically in Figure 5. The net decrease in the
(4)
ΔΔH
RT
ΔΔS
R
ln α = −
+
(5)
where k is the retention factor, ΔH is the change in enthalpy,
ΔS is the change in entropy, R is the gas constant (8.314 J·K−1·
mol−1), T is the absolute temperature, ϕ is the phase ratio (Vs/
Vm where Vs is the volume of the selector (i.e., excluding the
silica support material) and Vm is the volume of the mobile
phase), and α is enantioselectivity. The thermodynamic
parameters calculated from the van’t Hoff plots (see Figures
S-1−S-4 in Supporting Information) of CPA-1 and CSA-5 are
shown in Table 2. The ln k versus 1000/T van’t Hoff plot
governs the thermodynamics of retention as the analyte moves
from the mobile phase to the stationary phase. Here, we
observed two extremes: Ba(ClO4)2 was an enthalpy driven
system (ΔH < 0 and ΔH < TΔS), while Ba(OAc)2 was purely
entropy driven (ΔH > 0 and ΔH < TΔS). Ba(TFA)2 was an
intermediate between these two extremes in that although ΔH
< 0, the ΔG was dominated by the TΔS contribution.
The solvation of the cyclofructan macrocycle, the barium
cation, and the analyte should be more or less constant in each
mobile phase system. Furthermore, the number of barium
counteranions that undergo ionic interactions with the
complexed barium should also be constant. The remaining
variable is the nature of the anion and its ability to associate
with the stationary phase via hydrogen bonding. We postulate
that analyte retention is governed both by ionic interaction as
well as hydrogen bonding. The former interaction will be
Figure 5. Change of K (blue diamond) as a function of temperature
overlaid with the change in the % Ba2+−CF6 complexes (red square).
−
present in both OAc− and ClO4 mobile phase systems;
however, the latter interaction will be unfavorable for the
amount of barium on the CSP over the temperature range
studied is estimated to be ∼3.5% or in other words, 96.5% of
the barium remains on the CSP. This very small change in the
Ba2+ concentration with increasing temperature is not expected
to have a significant impact on retention.
strongly hydrogen bonded OAc− system and absent in the
ClO4 system. The removal of tightly hydrogen bonded OAc−
−
anions from the stationary phase to the mobile phase affords an
increase in its degrees of freedom and a consequent increase in
entropy.
ln k vs 1000/T
ln α vs 1000/T
Table 2. Thermodynamic Parameters Calculated from the van’t Hoff Plots in Figures S-1−S-4 (See Supporting Information)
ln(k) vs 1000/T
ΔS (J·K‑1·mol‑1)
ln(α) vs 1000/T
ΔΔS (J·K‑1·mol‑1)
2.20
5.2 mM Ba(TFA)2
ΔH (kJ·mol‑1)
ΔG298 K (kJ·mol‑1)
ΔΔH (kJ·mol‑1)
ΔΔG298 K (kJ·mol‑1)
−0.38
CPA-1
1st enantiomer
−2.06
−1.78
−3.29
−4.07
15.30
17.50
15.28
14.33
−6.62
−7.00
−7.85
−8.34
0.28
2nd enantiomer
1st enantiomer
2nd enantiomer
CSA-5
−0.78
−0.95
−0.49
5.11 mM Ba(ClO4)2
CPA-1
1st enantiomer
−5.59
−5.08
−6.76
−7.63
2.50
6.07
4.44
3.61
−6.34
−6.89
−8.08
−8.71
0.51
3.57
−0.55
−0.63
2nd enantiomer
1st enantiomer
2nd enantiomer
CSA-5
−0.87
−0.83
1.01 mM Ba(OAc)2
CPA-1
1st enantiomer
3.07
3.26
3.28
2.65
31.27
32.65
34.04
33.03
−6.86
−7.20
−6.24
−6.47
0.19
1.38
−0.23
−0.33
2nd enantiomer
1st enantiomer
2nd enantiomer
CSA-5
−0.63
−1.01
1288
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
a
Table 3. Optimized Separations of Chiral Sulfonic/Phosphoric Acids on Cyclofructan Columns
analyte
column
mobile phase
k1
α
Rs
CSA-1
CSA-2
CSA-3
CSA-4
CSA-5
CSA-6
CSA-7
CSA-8
CSA-9
CPA-1
CPA-2
CPA-3
CPA-4
CPA-5
CPA-6
CPA-7
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-P
LARIHC−CF6-RN
LARIHC−CF6-RN
FRULIC-N
EtOH with 0.05/0.05 (v/v %) TEA/TFA and 0.5 mM Ba(TFA)2
EtOH with 0.05/0.05 (v/v %) TEA/TFA and 0.5 mM Ba(TFA)2
EtOH with 0.08/0.02 (v/v %) TEA/TFA and 1.0 mM Ba(TFA)2
EtOH with 0.05/0.05 (v/v %) TEA/TFA and 0.5 mM Ba(TFA)2
MeOH with 5 mM Ba(ClO4)2
17.49
11.86
7.22
1.04
1.09
1.12
1.06
1.28
1.24
1.05
1.04
1.06
1.23
1.23
1.14
1.27
1.06
1.05
1.08
0.7
1.5
1.9
1.1
4.2
1.8
1.0
1.0
1.3
4.4
2.2
1.7
4.2
1.6
1.5
1.8
19.59
4.86
MeOH with 5 mM Ba(ClO4)2
3.68
EtOH with 0.1/0.1 (v/v %) TEA/TFA and 1 mM Ba(TFA)2
MeOH with 5 mM Ba(TFA)2
13.88
26.49
27.94
4.01
FRULIC-N
MeOH with 5 mM Ba(TFA)2
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-RN
LARIHC−CF6-P
MeOH with 5 mM Ba(TFA)2
EtOH with 5 mM Ba(ClO4)2
13.90
4.95
EtOH with 5.2 mM Ba(TFA)2
EtOH with 5.2 mM Ba(TFA)2
3.01
MeOH with 1 mM Ba(OAc)2
12.10
10.86
6.78
MeOH with 1 mM Ba(OAc)2
EtOH with 5.2 mM Ba(TFA)2*
a
In all cases, the flow rate was set to 1 mL/min, except for *, where it was set at 0.5 mL/min.
With regard to the Gibbs free energy (Table 2) for the
enantioselectivity for CSA-7. In fact, the FRULIC-N was the
only CSP to show appreciable selectivity for these two
compounds. Single chiral axis sulfonic acids (CSA-1 and
CSA-4) displayed longer retention but smaller enantioselectiv-
ity than the two chiral axes analogues (CSA-3 and CSA-5).
Figure S-6 in the Supporting Information shows the overall
performance of the three CSPs in separating chiral sulfonic and
phosphoric acids (Figure 1). The LARIHC−CF6-RN was
clearly the most successful in terms of the number of baseline
separations. It displayed enantioselectivity for every analyte
except CSA-9, which was best separated on the FRULIC-N.
enantioselectivity, we once again observed two extreme cases:
Ba(ClO4)2 exhibits the most favorable ΔΔG (−0.55 and −0.63
kJ·mol−1 for CPA-1 and CSA-5, respectively), while Ba(OAc)2,
though still favorable, exhibits the smallest ΔΔG (−0.23 and
−0.33 kJ·mol−1 for CPA-1 and CSA-5, respectively). Ba(TFA)2
is once again intermediate between the two extremes (−0.38
and −0.49 kJ·mol−1 for CPA-1 and CSA-5, respectively). A
possible explanation for the poor enantioselectivity observed in
the Ba(OAc)2 system is that adsorbed acetate anions would
impede hydrogen bonding interactions between the analyte and
the chiral selector. Three points of interaction are required for
chiral recognition,51 and one of these is ionic. The remaining
interactions can be hydrogen bonding, dipole−dipole, and/or
steric interactions. In the case of Ba(ClO4)2, though, the analyte
would be free to hydrogen bond and would therefore exhibit
better chiral recognition, which was observed.
The data in Table 2 demonstrate that the enantiomeric
separation of CPA-1 is entropically driven, while CSA-5 is
enthalpically driven for each mobile phase system. This means,
in the case of Ba(ClO4)2 and Ba(TFA)2 systems, that as
temperature is increased, enantioselectivity for CPA-1 improves
even while retention is decreasing (see Figure S-5 in the
Supporting Information). Such remarkable enantioselective
behavior is rarely observed. Since it is unlikely that the
cyclofructan macrocycle is changing and since the thermody-
namic behavior for each analyte is consistent regardless of the
barium additive used, the observed behavior must be related to
the analyte structure. Both CPA-1 and CSA-5 are sterically
hindered molecules; however, the sulfonate group has some
rotational degree of freedom, whereas the phosphate group
does not.
CONCLUSIONS
■
The selectivity of cyclofructan-based CSPs has been signifi-
cantly expanded to an entirely new class of anionic compounds.
Preliminary tests for carboxylates are also promising. This has
been achieved by exploiting a unique ion-pairing behavior of a
simple inorganic metal ion (Ba2+). Once the CSPs have been
conditioned and evaluated in the polar organic mode
(methanol or ethanol) with a barium salt in the mobile
phase, log k vs log [X] plots are linear with negative slopes and
behave as anion exchangers. Sterically hindered sulfonates or
phosphates elute earlier than the less sterically hindered ones.
The presence of additional hydrogen bond acceptors on the
analyte also increases retention. Ba(OAc)2 mobile phase
additive affords shorter retention, which for CPA-1 and CSA-
5, is entropy driven, while Ba(ClO4)2 affords longer retention,
which is enthalpy driven and Ba(TFA)2 is intermediate.
Enantioselectivity improves in the order of OAc− < TFA− <
−
ClO4 . The scale up of a separation of CSA-5 to preparative
scale LC has also been successfully demonstrated. Barium-
complexed cyclofructans behave as hydrophilic analogues of
reversed-phase ion-pairing LC systems.
Optimized Separations. Table 3 shows the optimized
separations for all the analytes in Figure 1. For the β-
ketosulfonic acids (CSA-7−9), the LARIHC−CF6-RN and the
FRULIC-N showed complementary behavior. CSA-7, chalcone
sulfonic acid containing no aromatic substituents, displayed
enantioselectivity on the LARIHC−CF6-RN column, whereas
the para-substituted CSA-8 and CSA-9 showed no selectivity.
Evidently, the presence of a para substituent was unfavorable
for enantioselectivity. The FRULIC-N showed reasonable
enantioselectivity for CSA-8 and CSA-9 but only slight
ASSOCIATED CONTENT
* Supporting Information
Additional information as noted in the text. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: sec4dwa@uta.edu.
■
1289
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290

Analytical Chemistry
Article
(29) Na, Y.-C. P.; Nilusha, L. T.; Dissanayake, M. K.; Armstrong, D.
W. Supramol. Chem. 2014, DOI: 10.1080/10610278.2013.85.
(30) Rogozhin, S. V.; Davankov, V. A. J. Chem. Soc. D 1971, 490a−.
(31) Roumeliotis, P.; Unger, K. K.; Kurganov, A. A.; Davankov, V. A.
J. Chromatogr. A 1983, 255, 51−66.
(32) Bidlingmeyer, B. A.; Deming, S. N.; Price, W. P., Jr.; Sachok, B.;
Petrusek, M. J. Chromatogr. A 1979, 186, 419−434.
(33) Pell, R.; Schuster, G.; Laemmerhofer, M.; Lindner, W. J. Sep. Sci.
2012, 35, 2521−2528.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge the Welch Foundation (Y-0026) for
their continued support. The Krische group gratefully acknowl-
edges the Shanghai Institute of Organic Chemistry (SIOC)
Postdoctoral Fellowship and the Welch Foundation (F-0038).
(34) Gargano, A. F. G.; Kohout, M.; Macikova, P.; Laemmerhofer,
M.; Lindner, W. Anal. Bioanal. Chem. 2013, 405, 8027−8038.
(35) Qiu, H.; Armstrong, D. W.; Berthod, A. J. Chromatogr. A 2013,
1272, 81−89.
REFERENCES
■
(1) Akiyama, T.; Itoh, J.; Yokota, K.; Fuchibe, K. Angew. Chem., Int.
Ed. 2004, 43, 1566−1568.
(36) Gritti, F.; Guiochon, G. Anal. Chem. 2006, 78, 4642−4653.
(37) Durmaz, F.; Memon, F. N.; Memon, N. A.; Memon, S.; Memon,
S.; Kara, H. Chromatographia 2013, 76, 909−919.
(38) Okada, T. Anal. Chem. 1998, 70, 1692−1700.
(39) Levitskaia, T. G.; Maya, L.; Van Berkel, G. J.; Moyer, B. A. Inorg.
Chem. 2006, 46, 261−272.
(2) Uraguchi, D.; Sorimachi, K.; Terada, M. J. Am. Chem. Soc. 2004,
126, 11804−11805.
(3) Terada, M. Chem. Commun. 2008, 4097−4112.
(4) Li, G.-Q.; Gao, H.; Keene, C.; Devonas, M.; Ess, D. H.; Kurti, L.
J. Am. Chem. Soc. 2013, 135, 7414−7417.
(5) Sun, Z.; Winschel, G. A.; Borovika, A.; Nagorny, P. J. Am. Chem.
Soc. 2012, 134, 8074−8077.
(40) Roberts, J. M.; Diaz, A. R.; Fortin, D. T.; Friedle, J. M.; Piper, S.
D. Anal. Chem. 2002, 74, 4927−4932.
(6) Zhang, H.; Wen, X.; Gan, L.; Peng, Y. Org. Lett. 2012, 14, 2126−
2129.
(41) Peruzzi, N.; Ninham, B. W.; Lo, N. P.; Baglioni, P. J. Phys. Chem.
B 2012, 116, 14398−14405.
(7) Zamfir, A.; Tsogoeva, S. B. Org. Lett. 2010, 12, 188−191.
(8) Storer, R. I.; Carrera, D. E.; Ni, Y.; MacMillan, D. W. C. J. Am.
Chem. Soc. 2006, 128, 84−86.
̌ ́
(42) Bilanicova, D.; Salis, A.; Ninham, B. W.; Monduzzi, M. J. Phys.
Chem. B 2008, 112, 12066−12072.
(43) Zhang, Y.; Cremer, P. S. Curr. Opin. Chem. Biol. 2006, 10, 658−
663.
(9) McInturff, E. L.; Yamaguchi, E.; Krische, M. J. J. Am. Chem. Soc.
2012, 134, 20628−20631.
(44) Staahlberg, J. Anal. Chem. 1994, 66, 440−449.
(45) Lamb, J. D.; Drake, P. A. J. Chromatogr. 1989, 482, 367−380.
(46) Fritz, J. S. J. Chromatogr. A 2005, 1085, 8−17.
(47) Izatt, R. M.; Bradshaw, J. S.; Nielsen, S. A.; Lamb, J. D.;
Christensen, J. J.; Sen, D. Chem. Rev. 1985, 85, 271−339.
(48) Lamb, J. D.; Smith, R. G. Talanta 1992, 39, 923−930.
(49) Smith, R. G.; Drake, P. A.; Lamb, J. D. J. Chromatogr. 1991, 546,
139−149.
(10) Zbieg, J. R.; Yamaguchi, E.; McInturff, E. L.; Krische, M. J.
Science 2012, 336, 324−327.
(11) Kellogg, R. M.; Nieuwenhuijzen, J. W.; Pouwer, K.; Vries, T. R.;
Broxterman, Q. B.; Grimbergen, R. F. P.; Kaptein, B.; La Crois, R. M.;
de Wever, E.; Zwaagstra, K.; van der Laan, A. C. Synthesis 2003, 1626−
1638.
(12) Loh, J. S. C.; van Enckevort, W. J. P.; Vlieg, E.; Gervais, C.;
Grimbergen, R. F. P.; Kaptein, B. Cryst. Growth Des. 2006, 6, 861−865.
(13) Bathori, N. B.; Nassimbeni, L. R. Cryst. Growth Des. 2012, 12,
2501−2507.
(14) Kawamura, M.; Uchiyama, T.; Kuramoto, T.; Tamura, Y.;
Mizutani, K. Carbohydr. Res. 1989, 192, 83−90.
(15) Shizuma, M.; Takai, Y.; Kawamura, M.; Takeda, T.; Sawada, M.
J. Chem. Soc., Perkin Trans. 2 2001, 1306−1314.
(16) Takai, Y.; Okumura, Y.; Tanaka, T.; Sawada, M.; Takahashi, S.;
Shiro, M.; Kawamura, M.; Uchiyama, T. J. Org. Chem. 1994, 59, 2967−
2975.
(50) Lamb, J. D.; Smith, R. G.; Jagodzinski, J. J. Chromatogr. 1993,
640, 33−40.
(51) Pirkle, W. H.; Pochapsky, T. C. Chem. Rev. 1989, 89, 347−362.
(17) Uchiyama, T.; Kawamura, M.; Uragami, T.; Okuno, H.
Carbohydr. Res. 1993, 241, 245−248.
(18) Yoshie, N.; Hamada, H.; Takada, S.; Inoue, Y. Chem. Lett. 1993,
2, 353−356.
(19) Sun, P.; Wang, C.-L.; Breitbach, Z. S.; Zhang, Y.; Armstrong, D.
W. Anal. Chem. 2009, 81, 10215−10226.
(20) Aranyi, A.; Bagi, A.; Ilisz, I.; Pataj, Z.; Fueloep, F.; Armstrong, D.
W.; Peter, A. J. Sep. Sci. 2012, 35, 617−624.
(21) Gondova, T.; Petrovaj, J.; Kutschy, P.; Armstrong, D. W. J.
Chromatogr. A 2013, 1272, 100−105.
(22) Kozlik, P.; Simova, V.; Kalikova, K.; Bosakova, Z.; Armstrong, D.
W.; Tesarova, E. J. Chromatogr. A 2012, 1257, 58−65.
(23) Padivitage, N. L. T.; Armstrong, D. W. J. Sep. Sci. 2011, 34,
1636−1647.
(24) Padivitage, N. L. T.; Dissanayake, M. K.; Armstrong, D. W. Anal.
Bioanal. Chem. 2014, 405, 8837−8848.
(25) Qiu, H.; Loukotkova, L.; Sun, P.; Tesarova, E.; Bosakova, Z.;
Armstrong, D. W. J. Chromatogr. A 2011, 1218, 270−279.
(26) Sun, P.; Armstrong, D. W. J. Chromatogr. A 2010, 1217, 4904−
4918.
(27) Sun, P.; Wang, C.; Padivitage, N. L. T.; Nanayakkara, Y. S.;
Perera, S.; Qiu, H.; Zhang, Y.; Armstrong, D. W. Analyst 2011, 136,
787−800.
(28) Vozka, J.; Kalikova, K.; Roussel, C.; Armstrong, D. W.;
Tesarova, E. J. Sep. Sci. 2013, 36, 1711−1719.
1290
dx.doi.org/10.1021/ac403686a | Anal. Chem. 2014, 86, 1282−1290