Download The Reciprocal Principle of Selectand-Selector-Systems in

molecules
Review
The Reciprocal Principle of Selectand-Selector-Systems
in Supramolecular Chromatography †
Volker Schurig
Institute of Organic Chemistry, University of Tübingen, Auf der Morgenstelle 18, 72076 Tübingen, Germany;
[email protected]; Tel.: +49-7071-62605
† This contribution is dedicated to Professor Stig Allenmark at his 80th anniversary.
Academic Editor: Yoshio Okamoto
Received: 5 October 2016; Accepted: 7 November 2016; Published: 15 November 2016
Abstract: In selective chromatography and electromigration methods, supramolecular recognition
of selectands and selectors is due to the fast and reversible formation of association complexes
governed by thermodynamics. Whereas the selectand molecules to be separated are always present
in the mobile phase, the selector employed for the separation of the selectands is either part of the
stationary phase or is added to the mobile phase. By the reciprocal principle, the roles of selector and
selectand can be reversed. In this contribution in honor of Professor Stig Allenmark, the evolution
of the reciprocal principle in chromatography is reviewed and its advantages and limitations are
outlined. Various reciprocal scenarios, including library approaches, are discussed in efforts to
optimize selectivity in separation science.
Keywords: supramolecular chromatography; electromigration methods; selectand; selector;
reciprocal principle; combinatorial libraries; cyclopeptides; single-walled nanotubes (SWCNTs);
fullerenes; cyclodextrins; calixarenes; congeners; enantiomers; molecular recognition; chiral
recognition; chiral stationary phases
1. Introduction
Separation approaches in highly selective supramolecular chromatography [1,2], with
their most refined expression of stereochemical resolution [3], rely on the fast and reversible
non-covalent association equilibrium between functionally and/or structurally complementary
partners. The designations selector and selectand were introduced by Mikeš into chromatography
to avoid ambiguities that may arise from the use of the notations solvent-solute, ligand-substrate,
or host-guest [4]. The new terms were coined in analogy to Ashby’s cybernetic operator-operand
terminology [5]. This terminology extends to all supramolecular interactions such as antibody-antigen,
receptor-substrate, metal ion-ligand etc. In chromatographic partitioning systems, selectors are
employed to separate mixtures of selectands which are added to the mobile phase. The selector
is either present as a stationary phase or as an additive to the liquid mobile phase. In order to find
an optimized selector for a pair of selectands, the role of selector/selectand can be reversed by the
reciprocal principle. One molecule of the selectand is now used as the stationary phase and an array
of potential structural related selectors are now employed as analytes in the mobile phase. The best
identified candidate is then used as the optimized lead stationary phase for the target selectands to
be separated. Thus the reciprocal recognition principle applies to selective separation system when
selectors and selectands change their role and belong to two or more forms of congeners, analogues,
isomers etc. In the special case of chiral recognition, both enantiomers of selectand and selector must
be available. This excludes proteins [6], carbohydrates [7] and antibodies [8], although they represent
versatile chiral stationary phases (CSPs). Pertinent examples will be described in the following Sections.
Molecules 2016, 21, 1535; doi:10.3390/molecules21111535
www.mdpi.com/journal/molecules
Molecules2016,
2016,21,
21,1535
1535
Molecules
of35
35
22of
2. Origin of the Reciprocal Principle in Chromatography
2. Origin of the Reciprocal Principle in Chromatography
Following the enantioseparation of [5]–[14]helicenes on the optically active charge-transfer agent
Following the enantioseparation of [5]–[14]helicenes on the optically active charge-transfer agent
R-2-(2,4,5,7-tetranitro-9-fluorenylideneamino-oxy)propionic acid (TAPA, Figure 1) or its 2-butyric acid
R-2-(2,4,5,7-tetranitro-9-fluorenylideneamino-oxy)propionic acid (TAPA, Figure 1) or its 2-butyric acid
homologue (TABA) [9–11], Mikeš speculated that the function of the selector and the selectand could
homologue (TABA) [9–11], Mikeš speculated that the function of the selector and the selectand could
be reversed, i.e., optically active helicenes could be used as resolving agents for racemic compounds
be reversed, i.e., optically active helicenes could be used as resolving agents for racemic compounds [4].
[4]. Thus, in a reciprocal fashion, if the selectands A1 and A2 are chromatographically separated on the
Thus, in a reciprocal fashion, if the selectands A1 and A2 are chromatographically separated on the
selector B1 (or B2), the selectands B1 and B2 are separated on the selector A1 (or A2) as well, whereby
selector B1 (or B2), the selectands B1 and B2 are separated on the selector A1 (or A2) as well, whereby
the respective pairs A1/A2 and B1/B2 refer to isomers (e.g., stereoisomers) or belong to members of
the respective pairs A1/A2 and B1/B2 refer to isomers (e.g., stereoisomers) or belong to members
homologous series of compounds (Figure 2). Indeed, it was subsequently demonstrated that P-(+)of homologous series of compounds (Figure 2). Indeed, it was subsequently demonstrated that
hexahelicene-7,7′-dicarboxylic
acid disodium salt could be used as the chiral stationary phase (CSP)
P-(+)-hexahelicene-7,70 -dicarboxylic acid disodium salt could be used as the chiral stationary phase
as π-donor selector for the enantioseparation of N-2,4-dinitrophenyl(DNP)-α-amino acid esters as
(CSP) as π-donor selector for the enantioseparation of N-2,4-dinitrophenyl(DNP)-α-amino acid esters
π-acceptor selectands by HPLC [12].
as π-acceptor selectands by HPLC [12].
Figure
(R)-(−)-2-(2,4,5,7-tetranitro-9Figure 1.
1. Enantiomeric
Enantiomeric separation
separation of
of racemic
racemic [5]–[14]helicenes
[5]–[14]helicenes on
on (R)-(
−)-2-(2,4,5,7-tetranitro-9fluorenylideneamino-oxy)propionic
fluorenylideneamino-oxy)propionicacid
acid(TAPA)
(TAPA) linked
linked to
to silica
silica gel
gel [4].
[4]. (a)
(a) odd
odd numbers;
numbers; (b)
(b) even
even
numbers.
numbers.Insertions:
Insertions:[7]helicene
[7]heliceneand
andTAPA.
TAPA.
Figure
Figure2.2.The
Thereciprocal
reciprocalsupramolecular
supramolecularrecognition
recognitionprinciple.
principle.IfIfthe
theselector
selectorBBseparates
separatesthe
the selectands
selectands
A11and
andA2A
,
than
the
selector
A
is
expected
to
separate
the
selectands
B
and
B
.
A
and
Bi are
A
, than
the
selector
A
is
expected
to
separate
the
selectands
B
1
and
B
2
.
A
i
and
i
are
homologous
2
1
2
i
homologouscongeners
compounds,
congeners
or isomers
(including enantiomers).
compounds,
or isomers
(including
enantiomers).
The principle of reciprocity by inverting the role of selectand and selector had first been
demonstrated via the differentiation of enantiomers by chiral solvating agents (CSA) in NMR
Molecules 2016, 21, 1535
3 of 35
Molecules 2016, 21, 1535
35
The principle of reciprocity by inverting the role of selectand and selector had first3 of
been
demonstrated via the differentiation of enantiomers by chiral solvating agents (CSA) in NMR
spectroscopy[13].
[13].Pirkle
Pirkleetetal.
al.stated:
stated:the
theenantiomers
enantiomersofofsecondary
secondaryand
andtertiary
tertiaryalcohols
alcoholsmay
maybebecaused
causedtoto
spectroscopy
havenonequivalent
nonequivalentnmr
nmrspectra
spectrathrough
throughthe
theuse
useofofappropriate
appropriateoptically
opticallyactive
activeamines
aminesasassolvents.
solvents.The
Theconverse
converse
have
of
this
phenomenon
also
obtains;
amine
enantiomers
may
have
nonequivalent
nmr
spectra
in
optically
active
of this phenomenon also obtains; amine enantiomers may have nonequivalent nmr spectra in optically active
alcohols
[14]
and
Pirkle
and
Hoover
concluded
that
the
roles
of
an
enantiomeric
solute
and
a
CSA
may
alcohols [14] and Pirkle and Hoover concluded that the roles of an enantiomeric solute and a CSA may
interchangeablefor
foraagiven
givenpair
pairofofcompounds
compounds[15].
[15].The
Thereciprocal
reciprocalprinciple
principlewas
waslater
laterextended
extendedby
by
bebeinterchangeable
Pirkle
et
al.
to
enantioselective
liquid
chromatography
[16,17].
The
researchers
pointed
out
that
one
Pirkle et al. to enantioselective liquid chromatography [16,17]. The researchers pointed out that one
cantake
takeone
onechiral
chiralstationary
stationaryphase
phase(CSP)
(CSP)totodevelop
developothers:
others:ififaaCSP
CSPderived
derivedfrom
from(+)-A
(+)-Aretains
retains(+)-B,
(+)-B,
can
thenaaCSP
CSPcomprised
comprisedofofimmobilized
immobilized(+)-B
(+)-Bshould,
should,using
usingthe
thesame
sameinteractions,
interactions,selectively
selectivelyretain
retain(+)-A
(+)-A[18,19].
[18,19].
then
Or
put
in
other
words
by
Pirkle
and
Pochapsky:
This
‘bootstrapping’
method
of
designing
reciprocal
CSPs
Or put in other words by Pirkle and Pochapsky: This ‘bootstrapping’ method of designing reciprocal CSPs
fromcompounds
compoundsthat
thatthemselves
themselvesresolve
resolveon
onexisting
existingCSPs
CSPsisisbased
basedon
onthe
thepremise
premisethat
thatififtwo
twomolecules
moleculesshow
show
from
mutual
chiral
recognition,
then
it
does
not
matter
which
of
the
two
is
bound
to
a
stationary
support
for
that
mutual chiral recognition, then it does not matter which of the two is bound to a stationary support for that
recognition
to
occur.
This
is
in
practice
not
strictly
correct,
for
the
nature
of
attachment
to
the
CSP
often
affect
recognition to occur. This is in practice not strictly correct, for the nature of attachment to the CSP often
chiralchiral
recognition.
Within
limits,
however,
reciprocity
is is
a auseful
[20]. The
Thelatter
latter
affect
recognition.
Within
limits,
however,
reciprocity
usefulguide
guidetoto CSP
CSP design
design [20].
important
restrictions
was
repeatedly
stressed
by
Pirkle
et
al.:
The
two
situations
are
not
‘mirror
images’
important restrictions was repeatedly stressed by Pirkle et al.: The two situations are not ‘mirror images’
andthe
therelationship
relationshipis isonly
only
approximate.
The
manner
of immobilization,
for example,
has some
bearing
and
approximate.
The
manner
of immobilization,
for example,
has some
bearing
uponupon
the
the
efficacy
of
the
chiral
recognition
process
[19].
Thus
deviations
from
the
reciprocal
principle
may
arise
efficacy of the chiral recognition process [19]. Thus deviations from the reciprocal principle may arise from
from
the environment
the selector
the stationary
e.g.,
the result
of immobilization
and
the
environment
of the of
selector
in thein
stationary
phase,phase,
e.g., as
theasresult
of immobilization
and the
the presence
of a spacer
linking
the selector
to a chromatographic
matrix
presence
of a spacer
linking
the selector
to a chromatographic
matrix
[21]. [21].
A
vivid
example
of
the
de
novo
design
of
a
CSP
for
a
particular
targetchiral
chiralselectand
selectandby
bythe
the
A vivid example of the de novo design of a CSP for a particular target
reciprocal
principle
involved
a
series
of
CSPs
developed
for
the
separation
of
the
enantiomers
of
the
reciprocal principle involved a series of CSPs developed for the separation of the enantiomers of the
non-steroidalanti-inflammatory
anti-inflammatorydrug
drug(NSAID)
(NSAID)naproxen
naproxen[22,23].
[22,23].By
Bythe
theso-called
so-calledimmobilized
immobilizedguest
guest
non-steroidal
method,
a
single
enantiomer
of
naproxen
(Figure
3a)
was
immobilized
to
form
a
naproxen-derived
method, a single enantiomer of naproxen (Figure 3a) was immobilized to form a naproxen-derived CSP
CSP (Figure
3b) which
was used
then used
to identify
a potential
candidate
an array
of test
racemates
(Figure
3b) which
was then
to identify
a potential
candidate
fromfrom
an array
of test
racemates
as
as
a
de
novo
enantioselective
naproxen
selector
[23].
This
reciprocal
study
also
identified
the
structural
a de novo enantioselective naproxen selector [23]. This reciprocal study also identified the structural
requirementsfor
forenantioselective
enantioselectiverecognition
recognitionfor
forthe
thetarget
targetracemate
racematenaproxen.
naproxen.The
Thetailor-made
tailor-made
requirements
Whelk-O1 CSP
CSP (Figure
only
showed
the highest
enantioseparation
factor for
naproxen
(α = 2.25)
Whelk-O1
(Figure3c)
3c)not
not
only
showed
the highest
enantioseparation
factor
for naproxen
but
also
enantioseparated
structurally
related
NSAIDs
as
well
as
a
host
of
different
racemates
by a
(α = 2.25) but also enantioseparated structurally related NSAIDs as well as a host of different
combination
of
face-to-edge,
and
face-to-face
π-π
interaction
as
well
as
hydrogen
bonding
[22,23].
racemates by a combination of face-to-edge, and face-to-face π-π interaction as well as hydrogen
The Whelk-O1
a broad
spectrum
for enantioseparations
of aromatic selectands.
The
bonding
[22,23]. CSP
The offered
Whelk-O1
CSP offered
a broad
spectrum for enantioseparations
of aromatic
enantioseparation
of
bis-,
trisand
hexakis-adducts
of
C
60 has also been achieved by HPLC on the
selectands. The enantioseparation of bis-, tris- and hexakis-adducts of C60 has also been achieved by
Whelk-O1-CSP
(Section 4). (Section 4).
HPLC
on the Whelk-O1-CSP
(a)
(b)
(c)
Figure3.3.(a)
(a)Structure
Structureof
ofnaproxen;
naproxen;(b)
(b)AAsingle
singleenantiomer
enantiomerofofnaproxen
naproxenattached
attachedcovalently
covalentlyororionically
ionically
Figure
onsilica;
silica;(c)(c)The
Theoptimized
optimized
CSP
Whelk-O1
(commercialized
Regis
Chemical
Morton
Grove,
on
CSP
Whelk-O1
(commercialized
by by
Regis
Chemical
Co.,Co.,
Morton
Grove,
IL,
IL,
USA).
Reprinted
(adapted)
with
permission
from
[23].
USA). Reprinted (adapted) with permission from [23].
Some preliminary work on the use of the reciprocal principle has been reported for the design of
Some preliminary work on the use of the reciprocal principle has been reported for the design
pyrethroid specific CSPs [24]. In a preliminary attempt to design novel CSPs for the resolution of the
of pyrethroid specific CSPs [24]. In a preliminary attempt to design novel CSPs for the resolution
stereoisomers of the insecticide cypermethrin, the principle of reciprocal interaction has been considered
of the stereoisomers of the insecticide cypermethrin, the principle of reciprocal interaction has been
via the following rationale: ‘If a CSP based on a pure enantiomer (+)-A resolves a racemate (+/−)-B, then a
considered via the following rationale: ‘If a CSP based on a pure enantiomer (+)-A resolves a racemate
single pure enantiomer (+)- or (−)-B should resolve (+/−)-A. This is true if no major point of interaction is
(+/−)-B, then a single pure enantiomer (+)- or (−)-B should resolve (+/−)-A. This is true if no major point
changed when attaching B to silica’. It should be mentioned that by employing the CSP (−)-A instead of
of interaction is changed when attaching B to silica’. It should be mentioned that by employing the CSP
Molecules 2016, 21, 1535
4 of 35
(−)-A instead of (+)-A, a reversal of the elution order of (−/+)-B would obtain. Likewise by employing
the CSP (−)-B instead of (+)-B, again a reversal of the elution order of (−/+)-A would obtain.
The biologically most active (1R cis S) stereoisomer of cypermethrin has been structurally modified
by an allylic group, followed by free radical addition or hydrosilylation reaction on two different silanes
and linking (i) a short-chain monofunctional unit and (ii) a long-chain trifunctional unit, respectively,
to silica. As cypermethrin can be resolved on N-(3,5-dinitrobenzoyl)-phenylglycine CSP, the reciprocal
interaction of the cypermethrin CSP with racemic N-3,5-dinitrobenzoyl-phenylglycinepropylamide
was claimed [24]. The actual objective to use the cypermethrin-derived CSPs to identify and optimize
the interaction with racemic test compounds from which a single enantiomer will then be bound to
silica to afford a cypermethrin-specific CSP still awaits its realization [24].
2.1. Capillary Electromigration Methods (CE)
In enantioselective capillary electromigration methods, the non-racemic chiral selector is added
to the background electrolyte as part of the mobile phase. The method of enantioseparation of charged
selectands with neutral selectors is referred to capillary zone electrophoresis (CZE), whereas the
method of enantioseparation of neutral selectands with charged selectors, entailed with self-mobility,
is referred to electrokinetic chromatography (EKC) [25]. Chankvetadze argued that from a mechanistic
point of view there is no principle difference whether the selectand or the selector is charged in
electromigration methods [26]. The philosophy behind this concept has been stated as follows: if a
neutral chiral selector can resolve the enantiomers of a charged chiral analyte, then the enantioseparation of the
neutral chiral analyte should also be possible with the charged chiral selector [26]. Thus another type of the
reciprocal principle applies to electromigration methods. Whereas the enantioseparation of neutral
chiral selectands on charged chiral selectors were at first considered not possible in earlier studies
of EKC, it was later successfully realized based on the reciprocal concept. Thus positively charged
cyclodextrin derivatives [25,27] and negatively charged cyclodextrin derivatives [27–30] were used for
the enantioseparation of a multitude of neutral chiral compounds [31,32].
Another reciprocal principle in the achiral electromigration domain has not yet been realized, but
may be anticipated. Native α-, β-, γ-cyclodextrin [33] and, more recently, δ-cyclodextrin [34] have been
used as an added mobile phase selector for molecular association with various positively or negatively
charged selectands in CZE. By inverting their role, a charged selector may be able to separate various
cyclodextrin congeners preceding δ-CD (α-, β-, γ-) and, more importantly, congeners following up
δ-CD (>δ).
In enantioselective capillary electrochromatography (enantio-CEC), a non-racemic chiral selector is
used as CSP coated on the inner capillary wall. As the first examples, racemic 1-phenylethanol [35,36]
and some non-steroidal anti-inflammatory drugs (NSAIDs, Figure 4) were enantioseparated on
Chirasil-Dex (permethylated β-cyclodextrin chemically linked to polydimethylsiloxane and thermally
immobilized on the inner wall of a 80 cm (effective) × 50 µm I.D. capillary column).
In order to find the best cyclodextrin selector for a given racemic analyte by the reciprocal approach,
the target molecule may now be used as CSP, whereas the modified cyclodextrin is added to the mobile
phase. However, as the cyclodextrins are only available in one enantiomeric form (comprised of
D -glucose building blocks) two columns coated with the L - and D -target molecule, respectively, are
required. As shown in Figure 5 the highest enantioselectivity for phenylalanilol was observed with
hydroxypropyl-β-cyclodextrin due to a large retention time difference on the two mirror image CSPs [37].
Although not related to the reciprocal principle, the following approach is of interest in its own
right. A dual chiral recognition system can be obtained when one cyclodextrin selector is used
as the CSP whereas another one is added to the mobile phase (combination of CEC and EKC).
As the enantioselectivities of the bonded and free CDs were opposite, the incremental addition
of the sulfonated β-cyclodextrin to a Chirasil-Dex-coated capillary led to an inversion of the elution
order of racemic 1,1’-bi-2,2’-naphthol-hydrogenphosphate (20 mM borate/phosphate (pH 7) buffer,
at 30 kV) [29].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
5 of 35
5 of 35
5 of 35
Figure 4. Enantioseparation of
profens by
electrochromatography on
Chirasil-β-Dex
coated
and
Figure
of profens
profens
by electrochromatography
on Chirasil-β-Dex
Chirasil-β-Dexcoated
coatedand
and
Figure 4.4. Enantioseparation
Enantioseparation of
by
electrochromatography on
◦
immobilized
(d
f
=
0.20
μm)
on
an
80
cm
(effective)
×
50
μm
I.D.
fused-silica
capillary
at
20
°C
and
30
kV.
immobilized
(d
=
0.20
µm)
on
an
80
cm
(effective)
×
50
µm
I.D.
fused-silica
capillary
at
20
C
and
immobilized (dff = 0.20 μm) on an 80 cm (effective) × 50 μm I.D. fused-silica capillary at 20 °C and 30 kV.
Buffer:
20 mM Tris/HCl
at pH 7. Reproduced
from S. Mayer,
Doctoral
Thesis,
University
of
Tübingen,
30
kV. Buffer:
mM Tris/HCl
pH 7. Reproduced
from Doctoral
S. Mayer,Thesis,
Doctoral
Thesis, of
University
Buffer:
20 mM20
Tris/HCl
at pH 7. at
Reproduced
from S. Mayer,
University
Tübingen,of
Tübingen,
Germany,
1993.
Tübingen,
1993.
Tübingen,Tübingen,
Germany, Germany,
1993.
Figure 5. Retention time differences of three modified β-cyclodextrins (β-CD as mono-2.6Figure 5.
time time
differences
of three
β-cyclodextrins
(β-CD as mono-2.6Figure
5. Retention
Retention
differences
of modified
three modified
β-cyclodextrins
(β-CD as
D- and L-phenylalaninol by the CEC-experiment (50 mM
dimethylphenylcarbamate)
on silica-bound
D
and
L
-phenylalaninol
by
the
CEC-experiment
(50 mM
dimethylphenylcarbamate)
on
silica-bound
mono-2.6-dimethylphenylcarbamate)
on silica-bound
and
L-phenylalaninol by the CEC-experiment
ammonium acetate, pH 6.5, 15 kV, column:
68 cm eff.D×-50
μm
I.D.) [37].
ammonium
acetate,
pH
6.5,
15
kV,
column:
68
cm
eff.
×
50
μm
I.D.)
[37].
(50 mM ammonium acetate, pH 6.5, 15 kV, column: 68 cm eff. × 50 µm I.D.) [37].
2.2. Combinatorial Library Approaches in Achiral Supramolecular Chromatographic Systems
2.2. Combinatorial Library Approaches in Achiral Supramolecular Chromatographic Systems
2.2. Combinatorial Library Approaches in Achiral Supramolecular Chromatographic Systems
The reciprocal principle has been used for the rational design of an optimized selector for
The reciprocal principle has been used for the rational design of an optimized selector for
The
reciprocal
principle
has been
used for the
rationalisdesign
optimized
selector for molecular
molecular
recognition.
Another
optimization
principle
based of
onan
the
use of combinatorial
selector
molecular recognition. Another optimization principle is based on the use of combinatorial selector
recognition.
Another
optimization
principle
is based
on the
use of combinatorial
selector
libraries
libraries aimed
at identifying
the prime
selector
for a given
enantioselective
separation
system.
This
libraries aimed at identifying the prime selector for a given enantioselective separation system. This
approach can also be combined with the reciprocal principle [38,39]. It has generally been accepted
approach can also be combined with the reciprocal principle [38,39]. It has generally been accepted
Molecules 2016, 21, 1535
6 of 35
aimed at identifying the prime selector for a given enantioselective separation system. This approach
can also be combined with the reciprocal principle [38,39]. It has generally been accepted that the
screening of libraries of selectors anchored to solid supports allows the identification of key structural
elements responsible for host/guest interaction and molecular recognition through binding assays
with soluble selectands. Combinatorial methods have been classified into two major categories. One is
based on the synthesis of mixtures and another is based on the parallel synthesis of pure library
components (also called the high-throughput approach). In the former approach, a mixture of library
components is synthesized and screened simultaneously for the desired property. In the parallel
library approach, each library component is synthesized and screened individually. The advantage of
the parallel approach is that the library is generally better characterized, whereas the random library
technique is advantageous in that a large number of components can be synthesized and screened
more efficiently. A classification of the various combinatorial approaches for the preparation and use
of CSPs for enantioseparations has been summarized [40].
Peptide libraries have been used to optimize selectors that bind specifically to a desired target
protein in the area of protein purification. Solid phase ‘one-bead-one-peptide’ (OBOP) parallel
combinatorial libraries have been successfully employed to search for affinity ligands for large
proteins. Thus the immobilized linear hexapeptide Try-Asn-Phe-Glu-Val-Leu was identified as a
potential host for the purification of S-protein by affinity chromatography [41]. Also an affinity resin
containing the peptide ligand Phe-Leu-Leu-Val-Pro-Leu has been developed for the purification of
fibrinogen. The selector was identified by screening the solid-phase combinatorial peptide library
using an immunostaining technique [42]. In a reversed reciprocal fashion, affinity chromatography
employing the immobilized S-protein was used for the screening of affinity peptide selectands from a
soluble peptide library consisting of octamers with glycine (Gly) at both termini of each peptide, i.e.,
Gly-X-X-X-X-X-X-Gly. The six variable centre positions were constructed using random sequences of
the six L-amino acids Tyr, Asn, Phe, Glu, Val and Leu. Peptides that were retained specifically on the
immobilized S-protein column were eluted by 2% acetic acid. The peptides in the acid eluate were further
separated using reversed-phase HPLC. Each separated peptide fraction was collected and the peptide
sequences deconvoluted by mass spectrometry (MS/MS). The screenings of the peptide library resulted
in twelve affinity peptides and eight out of them contained the essential sequence Asn-Phe-Glu-Val [43].
2.3. Combinatorial Cyclopeptide Library Approaches in Chiral Capillary Electrophoresis
Combinatorial cyclohexapeptide libraries were used as novel multicomponent chiral additives
for enantioseparation in capillary electrophoresis (CE) by Jung and Schurig et al. in 1996 [44]. It was
inferred that the time-consuming screening of a multitude of individual potential chiral selectors can
be avoided by employing a whole selector library. The random library approach is then followed by
deconvolution steps. The sub-libraries with reduced heterogeneity can then be employed to identify
the most enantioselective selectors. In general, cyclopeptide libraries of high molecular diversity
can be obtained by including structural variations. By using the nearly 200 commercially available
amino acids of defined chirality, 2006 = 64 × 1012 individual cyclohexapeptides are in principle
accessible. The number of cyclopeptides can further be increased by varying the ring size and by
including non-peptidic components, derivatives with side chain and backbone modifications, and
selected building blocks with inverted chirality [44]. Three libraries consisting of 183 = 5832 individual
cyclohexapeptides were synthesized as a random mixture and were characterized by electrospray
mass spectrometry [44]. Each library had the composition cyclo(O-O-X-X-X-O) and consisted
of three defined positions (O) and three randomized positions (X) which represented a mixture
of 18 of the common L-α-amino acids except cysteine and tryptophan. To aid ring closure,
one position (O) consisted of unnatural D-alanine (ala). When added to the mobile phase in
CE, the libraries cyclo(Asp-Phe-X-X-X-ala), cyclo(Arg-Lys-X-X-X-ala) and cyclo(Arg-Met-X-X-X-ala),
respectively, enantioseparated Tröger’s base and N-2,4-dinitrophenyl (DNP)- or Fmoc-D,L-glutamic
acid (α = 1.04–1.13) (Figure 6) [44].
Molecules 2016, 21, 1535
7 of 35
Molecules 2016, 21, 1535
7 of 35
Figure 6. Enantioseparation of racemic Tröger’s base and DNP-D,L-glutamic acid with combinatorial
Figure 6. Enantioseparation of racemic Tröger’s base and DNP-D,L-glutamic acid with combinatorial
cyclohexapeptides by CE. Conditions: 10 mmol cyclopeptide library cyclo(OOXXXO) in phosphate buffer
cyclohexapeptides
by CE.
Conditions:
10 mmol×cyclopeptide
phosphate
pH 7.4 (20 mmol).
Capillary:
50 cm (effective)
50 μm I.D., 20 library
kV (left, cyclo(OOXXXO)
middle) and 10 kV in
(right),
buffer pH
7.4 (20
Capillary:
50 (middle
cm (effective)
50 µm I.D., 20 kV (left, middle) and 10 kV
detection
at mmol).
260 nm (left)
and 340 nm
and right)×[44].
(right), detection at 260 nm (left) and 340 nm (middle and right) [44].
The result raised the question whether all 183 individual components (which are present in the library
at high dilution and probably in a non-equimolar
ratio) possess the same or, more likely, different
Theenantioselectivities
result raised the
question
whether
all winners’.
183 individual
components
(which
aremembers
present in the
with
individual
‘losers and
It may even
be conceived
that some
library at
high
dilution
andorder
probably
in a non-equimolar
the same
or, more likely,
exert
an opposite
elution
of the selectand
thus reducing theratio)
overallpossess
enantioselectivity.
Cooperative
between the individual
selectors were
believed
be absent [44].
To identify
theconceived
prime selector
differenteffects
enantioselectivities
with individual
‘losers
andtowinners’.
It may
even be
that some
in
the
library,
Chiari
et
al.
used
an
interesting
deconvolution
strategy
[45].
members exert an opposite elution order of the selectand thus reducing the overall enantioselectivity.
approach was performed in two steps: (i) the synthesis of sublibraries with a progressively
CooperativeThis
effects
between the individual selectors were believed to be absent [44]. To identify the
increased number of defined positions and (ii) the evaluation of their ability to act as chiral selectors
prime selector in the library, Chiari et al. used an interesting deconvolution strategy [45].
in CE for a set of DNP-D,L-amino acids. First of all, Chiari et al. deconvoluted the library cyclo(ArgThis
approach wasby
performed
in Dtwo
thelibrary
synthesis
of sublibraries with
a progressively
Lys-X-X-X-β-Ala)
substituting
-Alasteps:
(ala) of(i)the
cyclo(Arg-Lys-X-X-X-ala)
of Jung
and
increased
number
of by
defined
positions(Table
and1,(ii)
the
evaluation
of their
ability to
actS.2.2
as chiral
Schurig
et al. [44]
achiral β-alanine
S.1.1)
[45].
Then a number
the sublibraries
S.2.1,
and
S.2.3
were
prepared,
whereby
proline
was
deliberately
selected
for
one
of
the
positions
X
as
selectors in CE for a set of DNP-D,L-amino acids. First of all, Chiari et al. deconvoluted theit library
facilitates hexapeptide cyclization.
However,
only
proline
in position
(Table 1, S.2.2) exerted aof Jung
cyclo(Arg-Lys-X-X-X-β-Ala)
by substituting
D -Ala
(ala)
of the
library four
cyclo(Arg-Lys-X-X-X-ala)
high enantioselectivity of the sublibrary.
and Schurig et al. [44] by achiral β-alanine (Table 1, S.1.1) [45]. Then a number the sublibraries S.2.1,
S.2.2 and S.2.3
were
prepared,
whereby
proline was
deliberately
selected
for one
of the
positions X as
Table
1. List
of sublibraries
of cyclohexapeptides
used
in the deconvolution
process.
X stands
for one
it facilitates of
hexapeptide
cyclization.
However,
proline
in position
fourwith
(Table
1, S.2.2) exerted
the 18 of the common
L-α-amino
acids exceptonly
cysteine
and tryptophan.
Reprinted
permission
from [45]. Copyright
(1998)
American Chemical Society.
a high enantioselectivity
of the
sublibrary.
Numbering
Cyclohexapeptide Library
S.2.3:
cyclo(Arg-Lys-X-X-Pro-β-Ala)
cyclo(Arg-Lys-Val-Pro-X-β-Ala)
Cyclohexapeptide Library
cyclo(Arg-Lys-Met-Pro-X-β-Ala)
cyclo(Arg-Lys-X-X-X-β-Ala)
cyclo(Arg-Lys-Ile-Pro-X-β-Ala)
cyclo(Arg-Lys-Leu-Pro-X-β-Ala)
cyclo(Arg-Lys-Pro-X-X-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-X-β-Ala)
cyclo(Arg-Lys-X-Pro-X-β-Ala)
cyclo(Arg-Lys-Phe-Pro-X-β-Ala)
cyclo(Arg-Lys-X-X-Pro-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Val-β-Ala)
cyclo(Arg-Lys-Val-Pro-X-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Met-β-Ala)
cyclo(Arg-Lys-Met-Pro-X-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Ile-β-Ala)
cyclo(Arg-Lys-Ile-Pro-X-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Leu-β-Ala)
cyclo(Arg-Lys-Leu-Pro-X-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-X-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Phe-β-Ala)
Table 1. List of sublibraries of cyclohexapeptides used in the deconvolution process. X stands for one
S.1.1:
cyclo(Arg-Lys-X-X-X-β-Ala)
of the 18 of the common L-α-amino
acids except
cysteine and tryptophan. Reprinted with permission
S.2.1:
cyclo(Arg-Lys-Pro-X-X-β-Ala)
from [45]. Copyright (1998) American
Society.
S.2.2: Chemical
cyclo(Arg-Lys-X-Pro-X-β-Ala)
S.3.1:
Numbering
S.3.2:
S.1.1:S.3.3:
S.2.1:S.3.4:
S.2.2:S.3.5:
S.2.3:S.3.6:
S.4.1:
S.3.1:S.4.2:
S.3.2:S.4.3:
S.3.3:S.4.4:
S.3.4:S.4.5:
S.3.5:S.4.6:
S.3.6:
S.4.1:
S.4.2:
S.4.3:
S.4.4:
S.4.5:
S.4.6:
cyclo(Arg-Lys-Phe-Pro-X-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Val-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Met-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Ile-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Leu-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Phe-β-Ala)
Molecules 2016, 21, 1535
8 of 35
The optimized library cyclo(Arg-Lys-X-Pro-X-β-Ala) (S.2.2) was then semipreparatively resolved
Molecules 2016, 21, 1535
8 of 35
into three fractions
by reversed-phase HPLC. The second fraction exerted the highest enantioselectivity
Molecules 2016, 21, 1535
8 of 35
and amino acid
revealed
that this fraction (S.2.2)
contained
the
hydrophobic
amino
Theanalysis
optimized library
cyclo(Arg-Lys-X-Pro-X-β-Ala)
was then
semipreparatively
resolved
intoacids Val,
The
optimized
librarygave
cyclo(Arg-Lys-X-Pro-X-β-Ala)
(S.2.2)
was
semipreparatively
resolved into
three
fractions
reversed-phase
HPLC.
The synthesis
second fraction
highest enantioselectivity
and (Table 1).
Met, Ile, Leu
and
Phe.by This
rise
to the
of exerted
thethen
sixthe
sublibraries
S.3.1–S.3.6.
three
fractions
by reversed-phase
HPLC.
The second
fraction
exerted
thefor
highest
enantioselectivity
and
amino
acid analysis
revealed that
this fraction
contained
hydrophobic
amino
acids deconvolution,
Val, Met,
Ile,
The sublibrary
cyclo(Arg-Lys-Tyr-Pro-X-β-Ala)
(S.3.5)
wasthe
selected
the
further
since
amino
acid
analysis
revealed
fraction
contained
the hydrophobic
amino
acids
Val,sublibrary
Met, Ile,
Leu and
Phe.
This gave
rise tothat
thethis
synthesis
of the
six sublibraries
S.3.1–S.3.6.
(Table
1). The
Tyr contains
an aromatic system useful
for
detection
purposes.
Now the
six
sublibraries
S.4.1–S.4.6
Leu
and Phe. This gave rise to the(S.3.5)
synthesis
the sixfor
sublibraries
(Table
1). The
cyclo(Arg-Lys-Tyr-Pro-X-β-Ala)
was of
selected
the furtherS.3.1–S.3.6.
deconvolution,
since
Tyrsublibrary
contains an
were synthesized
(Table
1).
In
a
similar
fashion
position
5
was
optimized
by
a
preceding
HPLC
cyclo(Arg-Lys-Tyr-Pro-X-β-Ala)
(S.3.5) was
selected
for the
deconvolution,
sincewere
Tyr contains
an
aromatic system useful for detection
purposes.
Now
the further
six sublibraries
S.4.1–S.4.6
synthesized
aromatic
system
useful
for
detection
purposes.
Now
the
six
sublibraries
S.4.1–S.4.6
were
synthesized
(Table
1).
In
a
similar
fashion
position
5
was
optimized
by
a
preceding
HPLC
separation
into
three
separation into three fractions and again identifying hydrophobic amino acids as being essential
(Table
1). In
a similar
fashion
position
5 was optimized
by aas
preceding
HPLC
separation
into three and hence
fractions
and
again
hydrophobic
amino
acids
being
essential
for enantioselectivity.
for enantioselectivity.
Phe identifying
and
Tyr
in
position
5
provided
the
highest
enantioselectivity
fractions
identifying
hydrophobic
amino enantioselectivity
acids as being essential
for enantioselectivity.
Phe andand
Tyragain
in position
5 provided
the highest
and hence
the final defined
the final defined
cyclohexapeptide
selector
(Figure
7)for
has
been selected
for the
enantioseparation
of
Phe
and Tyr in position
provided
the
highest
enantioselectivity
and henceof the
final
defined
cyclohexapeptide
selector5(Figure
7) has
been
selected
the enantioseparation
DNPD,L-glutamic
DNP-D,L-glutamic
acid
(α = 8)
1.24)
(Figure
8) [45].
cyclohexapeptide
selector
(Figure
7) has been
selected for the enantioseparation of DNP-D,L-glutamic
acid (α = 1.24)
(Figure
[45].
acid (α = 1.24) (Figure 8) [45].
Figure
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
Figure 7. The optimized cyclohexapeptide library component, deconvoluted by Chiari et al. [45].
7.Figure
The optimized
cyclohexapeptide
component,
deconvoluted
byetChiari
7. The optimized
cyclohexapeptide library
library component,
deconvoluted
by Chiari
al. [45]. et
al. [45].
Figure 8. Enantioseparation of D,L-DNP-amino acids in a running electrolyte containing cyclo(ArgLys-Tyr-Pro-Tyr-β-Ala)
(10 of
mM
20 mM sodium
buffer,
pH 7.0) containing
as chiral mobile
phase
8. Enantioseparation
-DNP-amino
acidsphosphate
in a running
electrolyte
cyclo(ArgFigure 8.Figure
Enantioseparation
ofD,Lin
D , L -DNP-amino acids in a running electrolyte containing
additive (CMPA) in CE.
cmsodium
(effective)
× 50 μm
I.D.; pH
−20 7.0)
kV,as
15chiral
°C. Reprinted
with
Lys-Tyr-Pro-Tyr-β-Ala)
(10 Column:
mM in 2057mM
phosphate
buffer,
mobile phase
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
(10 mM
in
20
mMChemical
sodium
phosphate
buffer, pH 7.0) as chiral mobile
permission
from [45].
Copyright
American
Society.
additive
(CMPA)
in CE.
Column:(1998)
57 cm
(effective)
× 50 μm
I.D.; −20 kV, 15 °C. Reprinted with
phase additive
(CMPA)
in CE.
Column:
cm (effective)
50 µm I.D.; −20 kV, 15 ◦ C. Reprinted with
permission
from [45].
Copyright
(1998)57
American
Chemical×
Society.
The
overall
deconvolution
process
required
the
synthesis
and the evaluation of 15 sublibraries
permission from [45]. Copyright (1998) American Chemical Society.
The of
overall
process by
required
the synthesis
the evaluation
of 15 [46].
sublibraries
instead
the 54deconvolution
syntheses considered
a classical
procedureand
of serial
deconvolution
instead
the 54 syntheses
considered
by a classical and
procedure
of serial deconvolution
Inof
a follow
up study, several
cyclohexapeptide
cylcoheptapeptide
analogues of[46].
the identified
The overall
deconvolution
process
required
synthesis
and
the
evaluation
of[47].
15 sublibraries
In amembers
follow
upwere
study,
several
and
cylcoheptapeptide
analogues
of the identified
library
studied
ascyclohexapeptide
chiral
selectors the
in
countercurrent
capillary
electrophoresis
At
were
studied
countercurrent
capillary
At
the operating
pH
of 7.0,
all of as
thechiral
selectors
bear ainpositive
charge,of
while
the electrophoresis
analyte
bears a [47].
negative
instead oflibrary
the
54members
syntheses
considered
by selectors
a classical
procedure
serial
deconvolution
[46].
the
operating
pH of 7.0,
all of the
selectors
a positive
charge, while the ether)
analyte
bears
a negative
charge.
A poly(acrylamide-co-allyl
amide ofbear
D-gluconic
was
coated
onto
theidentified
In a follow
up
study,
several
cyclohexapeptide
andacid-co-allylglycidyl
cylcoheptapeptide
analogues
of the
charge. A poly(acrylamide-co-allyl amide of D-gluconic acid-co-allylglycidyl ether) was coated onto the
library members were studied as chiral selectors in countercurrent capillary electrophoresis [47].
At the operating pH of 7.0, all of the selectors bear a positive charge, while the analyte bears
a negative charge. A poly(acrylamide-co-allyl amide of D-gluconic acid-co-allylglycidyl ether)
was coated onto the capillary to suppress the EOF to ensure that the analyte and selector would
migrate in opposite directions by counter-current CE. The following changes were made in the
deconvoluted cyclohexapeptide cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala). (i) Tyr in position five was
Molecules 2016, 21, 1535
9 of 35
substituted by Tyr(3-NO2 ), Phe, Phe(4-NO2 ) and Trp; (ii) Tyr in position three was substituted
by positively charged Lys, negatively charged Glu and neutral Leu; (iii) Ala was inserted into
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) in the subsequent positions 1–6 to yield conformationally flexible
cycloheptapeptides with positional variety. All positively charged cyclohexapeptides are effective
Molecules 2016, 21, 1535
9 of 35
chiral selectors for the enantioseparation of DNP-α-amino acids with resolutions factors Rs = 1–5.
The charge
in position
3 did
anythat
influence
onand
enantioselectivity
of DNP-AAs.
However, all of
capillary
to suppress
the not
EOFhave
to ensure
the analyte
selector would migrate
in opposite directions
by counter-current
CE. to
Theresolve
following
changes were
made
the finding
deconvoluted
cyclohexapeptide
the cycloheptapeptides
failed
DNP-amino
acids
[47].inThis
suggested
that the size and
(i) Tyrimportant
in position five
substituted
by Tyr(3-NO
2), Phe, Phe(4-NO2)
rigidity cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala).
of the cyclopeptide system was
forwas
ensuring
chiral
discrimination.
In subsequent
and Trp; (ii) Tyr in position three was substituted by positively charged Lys, negatively charged Glu
NMR studies it was recognized that the presence of aromatic moieties carrying electron-withdrawing
and neutral Leu; (iii) Ala was inserted into cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) in the subsequent
substituents,
i.e., nitro groups, were essential for enantiorecognition via π-π stacking interactions and
positions 1–6 to yield conformationally flexible cycloheptapeptides with positional variety. All
amide/aromatic
n-π interactions
[47,48].
positively charged
cyclohexapeptides
are effective chiral selectors for the enantioseparation of DNP-αBased
onacids
the optimized
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
cyclohexapeptide
selector
deconvoluted
amino
with resolutions
factors Rs = 1–5. The charge in position
3 did not have any
influence
on
enantioselectivity
of DNP-AAs.
However,etallal.
of the
cycloheptapeptides
failed to
resolve DNP-amino
by Chiari
et al. [45], Jung
and Schurig
also
screened related
cyclopeptides
with small
acids
[47]. This finding
suggested
that the
size and was
rigidity
of the cyclopeptide
system wasand
important
structural
variations
[49]. Thus
aromatic
tyrosine
replaced
by phenylalanine
tryptophane
for ensuring chiral discrimination. In subsequent NMR studies it was recognized that the presence of
in position 3 and 5. In both cases the enantioselectivity was reduced for several DNP-amino
aromatic moieties carrying electron-withdrawing substituents, i.e., nitro groups, were essential for
acids (Table
2, Figure 9)viaand
choice
of the aromatic
amino acid
had a pronounced
effect on
enantiorecognition
π-π the
stacking
interactions
and amide/aromatic
n-π interactions
[47,48].
enantioselectivity.
When
the five-membered
ring of proline
was opened,
the deconvoluted
rigid structure of
Based on the
optimized
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
cyclohexapeptide
selector
by Chiari et al. [45],
Jung
and Schurig
al. also screenedby
related
small
structural
the cyclohexapeptide
was
released
andetaccompanied
ringcyclopeptides
extension. with
Thus
when
proline was
variations
[49].
Thus
aromatic
tyrosine
was
replaced
by
phenylalanine
and
tryptophane
in
position
3 at all
substituted by 6-aminohexanoic acid in cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala), no enantioseparation
and
5.
In
both
cases
the
enantioselectivity
was
reduced
for
several
DNP-amino
acids
(Table
2,
Figure
9)
was observed for various DNP-amino acids [49]. Enantioselectivity also broke down when the amino
and the choice of the aromatic amino acid had a pronounced effect on enantioselectivity. When the
acid sequence in cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) was changed to cyclo(Arg-Pro-Lys-Tyr-Tyr-β-Ala)
five-membered ring of proline was opened, the rigid structure of the cyclohexapeptide was released
and cyclo(Arg-Pro-Tyr-Lys-Tyr-β-Ala).
Thuswhen
the sequence
amino
acid-proline-aromatic
and accompanied by ring extension. Thus
proline was‘aromatic
substituted
by 6-aminohexanoic
acid in amino
acid’ (Tyr-Pro-Tyr)
was
essential
for
chiral
recognition.
Finally
the
ring
size
of
the
cyclopeptide
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala), no enantioseparation at all was observed for various DNP-amino
acids [49].
Enantioselectivity
broke down
when the aminowas
acid observed
sequence inwith
cyclo(Arg-Lys-Tyrwas probed
(Table
3) [49]. Thealso
highest
enantioselectivity
cyclohexapeptides.
Pro-Tyr-β-Ala)
was
changed
to
cyclo(Arg-Pro-Lys-Tyr-Tyr-β-Ala)
and
cyclo(Arg-Pro-Tyr-Lys-Tyr-βIn agreement with results of De Lorenzi et al. [47], enantioselectivity was lost for conformationally
Ala). Thus the sequence ‘aromatic amino acid-proline-aromatic amino acid’ (Tyr-Pro-Tyr) was essential
flexible cycloheptapeptides. Surprisingly, this drop was even more pronounced with cyclopentapeptides
for chiral recognition. Finally the ring size of the cyclopeptide was probed (Table 3) [49]. The highest
(Table 3).
When the cyclohexapeptide cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) was opened up to the linear
enantioselectivity was observed with cyclohexapeptides. In agreement with results of De Lorenzi
peptideetArg-Lys-Tyr-Pro-Tyr-Lys
by lost
exchanging
β-Ala for flexible
Lys nocycloheptapeptides.
enantioselectivity
for DNP-amino
al. [47], enantioselectivity was
for conformationally
Surprisingly,
acids was
[49].more
Thus
the cyclohexapeptide
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
identified by
thisobserved
drop was even
pronounced
with cyclopentapeptides
(Table 3). When the cyclohexapeptide
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
opened upthe
to the
linear
peptide Arg-Lys-Tyr-Pro-Tyr-Lys
by not be
Chiari et
al. (Figure 7) [45] indeedwas
represents
most
favourable
selector which could
exchanging
β-Ala
for
Lys
no
enantioselectivity
for
DNP-amino
acids
was
observed
[49].
Thus
the
optimized further.
cyclohexapeptide cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) identified by Chiari et al. (Figure 7) [45] indeed
represents
the mostfactors
favourable
which could
optimized further.by CE (conditions as in
Table 2. Resolution
Rs ofselector
DNP-amino
acids not
on be
cyclohexapeptides
Figure 6) [49].
Table 2. Resolution factors Rs of DNP-amino acids on cyclohexapeptides by CE (conditions as in
Figure 6) [49].
Amino Acid/Cyclopeptide
Amino Acid/Cyclopeptide
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
cyclo(Arg-Lys-Phe-Pro-Phe-β-Ala)
cyclo(Arg-Lys-Phe-Pro-Phe-β-Ala)
cyclo(Arg-Lys-Trp-Pro-Trp-β-Ala)
cyclo(Arg-Lys-Trp-Pro-Trp-β-Ala)
DNP-Glu
DNP-Glu
4.97 4.97
2.33 2.33
1.69 1.69
DNP-Norleu
DNP-Norleu
3.80
3.80
1.63
1.63
1.31
1.31
DNP-Meth
DNP-Meth
5.46
5.46
1.96
1.96
1.21
1.21
DNP-Ala
DNP-Ala
2.27 2.27
1.03 1.03
1.47 1.47
DNP-Norval
DNP-Threo
DNP-Norval
DNP-Threo
3.64
2.10
3.64
2.10
1.66
0.59
1.66
0.59
1.93
0.90
1.93
0.90
Figure
9. Influence
of change
the change
Tyrfor
forPhe
Phe in
in the
the cyclohexapeptide
for enantioseparation
of
Figure 9.
Influence
of the
of of
Tyr
cyclohexapeptide
for enantioseparation
of
norvaline (conditions as in Figure 8) [49].
norvaline (conditions as in Figure 8) [49].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
10 of 35
10 of 35
Table
Resolution
factors
RsDNP-amino
of DNP-amino
on cyclopeptides
of ring
varying
ring
by CE
Table 3. Resolution
factors
Rs of
acids acids
on cyclopeptides
of varying
size by
CEsize
(conditions
(conditions
Figure 6) [49].
as in Figureas
6)in
[49].
Amino Acid/Cyclopeptide
DNP-Glu
DNP-Norleu
DNP-Meth
Amino Acid/Cyclopeptide
DNP-Glu
DNP-Norleu
DNP-Meth
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
4.97
3.80
5.46
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala)
3.80 1.41
5.461.53
cyclo(Lys-Lys-Tyr-Tyr-Tyr-Tyr-Lys) 4.97 0.36
cyclo(Lys-Lys-Tyr-Tyr-Tyr-Tyr-Lys)
0.36
1.41 0.62
1.53 0
cyclo(Lys-Tyr-Arg-Tyr-β-Ala)
0
cyclo(Lys-Tyr-Arg-Tyr-β-Ala)
0
0.62
0
cyclo(Arg-Lys-Tyr-Tyr-β-Ala)
0.23
0.73
cyclo(Arg-Lys-Tyr-Tyr-β-Ala)
0.23
0.73
0.430.43
DNP-Ala
DNP-Ala
2.27
2.27
1.05
1.05
0.35
0.35
0.34
0.34
DNP-Norval
DNP-Norval
3.64
3.64
1.46
1.46
0.37
0.37
0.55
0.55
DNP-Threo
DNP-Threo
2.10
02.10
00
0
00
2.4. Reciprocal Principle in Combinatorial Approaches Utilizing Cyclopeptides—A Failure and a Success
2.4. Reciprocal Principle in Combinatorial Approaches Utilizing Cyclopeptides—A Failure and a Success
Instead of a deconvolution strategy, a reciprocal LC approach has been considered by Jung, Schurig
Instead of a deconvolution strategy, a reciprocal LC approach has been considered by Jung,
et al. to identify an enantioselective hit component in the deconvoluted library (Figure 10) [49].
Schurig et al. to identify an enantioselective hit component in the deconvoluted library (Figure 10) [49].
Figure 10.
10. Illustration
Illustration of
of the
the identification
identification of
of an
an optimal
optimal selectand
selectand in
in aa small
small library
library exhibiting
exhibiting the
the
Figure
highest
enantioselectivity
toward
a
chiral
selector
used
either
in
the
Sor
R-configuration
as
revealed
highest enantioselectivity toward a chiral selector used either in the S- or R-configuration as revealed
by the
the largest
largest difference
differenceof
ofits
itsretention
retention(earmarked
(earmarkedby
bythe
thepale
paleblue
bluecolour
colouratatthe
theright).
right).
by
Since the cyclopeptide library exists only in one enantiomeric form (obtained from L-amino acids)
Since the cyclopeptide library exists only in one enantiomeric form (obtained from L-amino acids)
two different columns containing the target racemate separately in the L- and D-form, respectively, are
two different columns containing the target racemate separately in the L- and D-form, respectively, are
required. The sub-library component present in the mobile phase which exhibits the largest difference
required. The sub-library component present in the mobile phase which exhibits the largest difference
in its retention on the two mirror-image selectors, should also show the highest enantioselectivity in
in its retention on the two mirror-image selectors, should also show the highest enantioselectivity
the reciprocal system when used as chiral selector for the racemic non-bonded candidate used as
in the reciprocal system when used as chiral selector for the racemic non-bonded candidate used as
selectand (Figure 10).
selectand (Figure 10).
In initial experiments L- or D-DNP-alanine silica (Figures 11 and 12) was filled into
In initial experiments L- or D-DNP-alanine silica (Figures 11 and 12) was filled into 100 µm
100 μm I.D. capillaries. Cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) was selected as the analyte as it shows high
I.D. capillaries. Cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) was selected as the analyte as it shows high
enantioselectivity toward racemic DNP-alanine (Figure 8) [45]. Micro-HPLC measurements were
enantioselectivity toward racemic DNP-alanine (Figure 8) [45]. Micro-HPLC measurements were
performed in the reversed phase mode (H2O), in the organic polar mode (methanol/acetonitrile) and
performed in the reversed phase mode (H2 O), in the organic polar mode (methanol/acetonitrile) and
in the normal phase mode (toluene, n-hexane/2-propanol). However, no measurable retention time
in the normal phase mode (toluene, n-hexane/2-propanol). However, no measurable retention time
differences for the two mirror image stationary phases were observed. Also the coating of 50 μm I.D.
differences for the two mirror image stationary phases were observed. Also the coating of 50 µm I.D.
capillaries with L- or D-5-fluoro-2,4-DNP-alanine in the presence of aminopropyltrimethoxysilane in
capillaries with L- or D-5-fluoro-2,4-DNP-alanine in the presence of aminopropyltrimethoxysilane in
toluene and the investigation of cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) by pressure-supported open-tubular
toluene and the investigation of cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala) by pressure-supported open-tubular
capillary electrochromatography (OT-CEC) did not reveal differences in retention times as expected
capillary electrochromatography (OT-CEC) did not reveal differences in retention times as expected
from the reciprocal principle [49]. This unexpected result reinforces the limitation of the reciprocal
from the reciprocal principle [49]. This unexpected result reinforces the limitation of the reciprocal
principle as enantioselectivity may be reduced by non-specific interactions on silica. This might be
principle as enantioselectivity may be reduced by non-specific interactions on silica. This might be
circumvented by end-capping of silanol groups or elongation of the linker categories [38,39].
circumvented by end-capping of silanol groups or elongation of the linker categories [38,39].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
11 of 35
11 of 35
11 of 35
Figure 11. Linking of L- or D-DNP-alanine to silica (5 μm) via the amine route [49].
Figure
11.11.Linking
to silica
silica(5
(5μm)
µm)via
viathe
theamine
amineroute
route
[49].
Figure
LinkingofofLL- -or
orDD-DNP-alanine
-DNP-alanine to
[49].
Figure 12. Linking of L- or D-DNP-alanine to silica (5 μm) via the epoxide route [49].
Figure
LinkingofofLL- -or
orDD-DNP-alanine
-DNP-alanine to
[49].
Figure
12.12.Linking
to silica
silica(5
(5μm)
µm)via
viathe
theepoxide
epoxideroute
route
[49].
Since the cyclohexapeptide cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala), used as chiral mobile phase selector
Since the cyclohexapeptide cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala), used as chiral mobile phase selector
in CE,
enantioseparates
DNP-glutamic
acid (Figure 8), the reciprocal principle
was also
tested
by adding
Since
the cyclohexapeptide
cyclo(Arg-Lys-Tyr-Pro-Tyr-β-Ala),
used as chiral
mobile
phase
selector
in CE, enantioseparates DNP-glutamic acid (Figure 8), the reciprocal principle was also tested by adding
L-DNP-Glu
in a first run,
and D-DNP-Glu
in a second
run,
to the principle
background
electrolyte
inby
theadding
CE
in CE,
enantioseparates
DNP-glutamic
acid
(Figure
8),
the
reciprocal
was
also
tested
L-DNP-Glu in a first run, and D-DNP-Glu in a second run, to the background electrolyte in the CE
experiment.
The
corresponding
time in
of a
the
added run,
cyclohexapeptide
was determined
by indirect
L -DNP-Glu
in
a
first
run,
and
Delution
-DNP-Glu
second
to
the
background
electrolyte
in
the
experiment. The corresponding elution time of the added cyclohexapeptide was determined by indirectCE
UV-detection.
also
no striking
difference
of cyclohexapeptide
the elution time of the cyclohexapeptide
was
experiment.
TheUnexpectedly
corresponding
elution
time of
the added
determined by indirect
UV-detection.
Unexpectedly
also
no striking
difference
of the elution time of was
the cyclohexapeptide
was
observed
for
the
mirror-image
situation
of
the
two
experiments
between
15–20
°C
and
at
voltages
of
UV-detection.
also
no striking
of the elution
of°C
theand
cyclohexapeptide
observed for Unexpectedly
the mirror-image
situation
of thedifference
two experiments
betweentime
15–20
at voltages of
10–25
kV, again
reinforcing
the limitation
the
concept [49].
was
observed
for the
mirror-image
situationof
thereciprocal
two experiments
between 15–20 ◦ C and at voltages
10–25
kV, again
reinforcing
the limitation
ofofthe
reciprocal
concept [49].
For
screening
a
mixture
of
a
combinatorial
library
for
its
best
selector,
a reciprocal approach
of 10–25For
kV,screening
again reinforcing
of the
reciprocal
[49]. also
a mixture the
of alimitation
combinatorial
library
for itsconcept
best selector,
also a reciprocal approach
was developed by Li et al. in line with Figure 10 [50]. The model study concerned the optimized
Fordeveloped
screeningby
a mixture
combinatorial
itsmodel
best selector,
also a reciprocal
approach
was
Li et al. of
in aline
with Figure library
10 [50].for
The
study concerned
the optimized
enantioseparation of (1-naphthyl)leucine ester with a small peptide library (4 × 4 = 16) containing Leu,
was
developed by Li
al. in line with Figure
10 [50].
The
model
study(4concerned
the optimized
enantioseparation
ofet
(1-naphthyl)leucine
ester with
a small
peptide
library
× 4 = 16) containing
Leu,
Ala, Gly and Pro and four different N-acyl components. In a reciprocal fashion, L-(1-naphthyl)leucine
Ala, Gly and Proofand
four different N-acyl
components.
Inpeptide
a reciprocal
fashion,
enantioseparation
(1-naphthyl)leucine
ester
with a small
library
(4 ×L4-(1-naphthyl)leucine
= 16) containing Leu,
ester was first immobilized and used as CSP for the enantiomeric peptide libraries obtained from Lester
and used
as CSP
for the enantiomeric
peptide
libraries
obtained from LAla,
Glywas
andfirst
Proimmobilized
and four different
N-acyl
components.
In a reciprocal
fashion,
L -(1-naphthyl)leucine
and D-amino acids, respectively. L- and D-sublibraries, exhibiting the largest retention differences on
and
D
-amino
acids,
respectively.
L
and
D
-sublibraries,
exhibiting
the
largest
retention
differences
on
ester was first immobilized and used as CSP for the enantiomeric peptide libraries
obtained
Molecules 2016, 21, 1535
12 of 35
from L- and D-amino acids, respectively. L- and D-sublibraries, exhibiting the largest retention
Molecules 2016, 21, 1535
12 of 35
differences on the L-configurated CSP, were deconvoluted and the best selector for the racemate was
identified
[50]. The optimized
peptide Leu-Gly
wasbest
then
attached
to silica.
It enantioseparated
the L-configurated
CSP, were deconvoluted
and the
selector
for the
racemate
was identified [50].racemic
The
(1-naphthyl)leucine
ester
with
an
enantioseparation
factor
α
=
3.
Thus
it
was
indeed
demonstrated
optimized peptide Leu-Gly was then attached to silica. It enantioseparated racemic (1-naphthyl)leucine
that
CSPs
cananbeenantioseparation
developed by a reciprocal
approach
using enantiomeric
ester
with
factor α chromatographic
= 3. Thus it was indeed
demonstrated
that CSPslibraries.
can be
developed by a reciprocal chromatographic approach using enantiomeric libraries.
2.5. On-Bead Library Combinatorial Approaches
2.5.For
On-Bead
Library Combinatorial
Approaches
the enantioselective
screening
of a target racemate, Weingarten et al. used a resin library
containing
60 members
of a multimodal
receptor
wereWeingarten
bonded to polystyrene
(100 µm)
For the
enantioselective
screening of
a targetwhich
racemate,
et al. used a beads
resin library
containing
60 members
of a multimodal
receptor
which weresupported
bonded to polystyrene
(100 μm)
(Figure
13) [51].
The chemically
encoded,
solid-phase
library wasbeads
generated
by a
(Figure 13) [51].
The chemically
encoded,
solid-phase
supported
library
was (Figure
generated
by The
a mix-andmix-and-split
protocol
employing
three classes
of chiral
building
blocks
13).
rationale
split protocol
employing three classes
of chiral
building
13). IfThe
rationale
behind
behind
this non-chromatographic
screening
method
was blocks
stated (Figure
as follows:
a tightly
binding
andthis
highly
non-chromatographic
screening
method
was
stated
as
follows:
If
a
tightly
binding
and
highly to
enantioselective (e.g., Rs > 10) resin or other solid material were readily available, then it would be possible
enantioselective
Rs > 10)stirring
resin orthe
other
solid material
wereareadily
available,
thenThe
it would
be possible
to
resolve
compounds(e.g.,
by simply
racemate
with such
resin and
filtering.
screening
method
resolve
compounds
by
simply
stirring
the
racemate
with
such
a
resin
and
filtering.
The
screening
method
employed differently coloured enantiomeric target molecules that were labelled with differently
employed
differently
enantiomeric
targetthe
molecules
that wereacid
labelled
with differently
coloured
dyes.
Thus acoloured
quasi-racemate
included
blue L-α-amino
derivative
and the red
coloured dyes. Thus a quasi-racemate included the blue L-α-amino acid derivative and the red D-αD -α-amino acid derivative. Proline was selected as the test compound. The idea was to treat an
amino acid derivative. Proline was selected as the test compound. The idea was to treat an equimolar
equimolar mixture of these coloured probe molecules with a library of chiral selectors on synthetic
mixture of these coloured probe molecules with a library of chiral selectors on synthetic beads in
beads in which each bead carried a different chiral selector (one-bead-one-selector, OBOS). It was
which each bead carried a different chiral selector (one-bead-one-selector, OBOS). It was reasoned
reasoned that a highly enantioselective binding would yield beads that were either red or blue, whereas
that a highly enantioselective binding would yield beads that were either red or blue, whereas nonnon-enantioselective
binding would yield beads that were brown. Thus the best selector was directly
enantioselective binding would yield beads that were brown. Thus the best selector was directly
visualized
through
a
screening.The
Thereddest
reddestand
and
bluest
beads
found
visualized through atwo-colour
two-colourdifferential
differential binding
binding screening.
bluest
beads
found
with
each
probe
to determine
determinethe
thestructures
structuresofof
their
associated
chiral
with
each
probewere
werethen
thenpicked
pickedand
and decoded
decoded to
their
associated
chiral
selectors.
Finally
the
leading
resolving
resins
were
individually
resynthesized
on
a
gram
scale
selectors. Finally the leading resolving resins were individually resynthesized on a gram scale andand
again
treated
enantiomers.Subsequent
Subsequenttreatment
treatment
with
NaOMe
again
treatedwith
withexcess
excessred
redand
andblue
blue proline
proline enantiomers.
with
NaOMe
andand
HPLC
quantification
determinationofofthe
theenantiomeric
enantiomeric
excess
HPLC
quantificationofofthe
thereleased
released dyes
dyes provided
provided aadetermination
excess
(ee)(ee)
of of
thethe
proline
The identified
identifiedenantioselective
enantioselective
resolving
resin
was
then
prolinederivatives
derivativesbound
boundto
to the
the beads.
beads. The
resolving
resin
was
then
tested
via
abilitytotokinetically
kineticallyresolve
resolve racemic
racemic proline
tested
via
itsitsability
proline ester
esterin
insolution
solution[51].
[51].
Figure
Structureofofthe
themultimodal
multimodal receptor
ofof
differently
dye-tagged
Figure
13.13.Structure
receptor array
arrayand
andkinetic
kineticresolution
resolution
differently
dye-tagged
solid-supported
amino
acids
from
a
60-member
library.
Module
A
contains
15
different
and
solid-supported amino acids from a 60-member library. Module A contains 15 different D- andD-L-amino
L-amino acids, module B contains RR or SS diamine enantiomers, and the receptor module C contains
acids, module B contains RR or SS diamine enantiomers, and the receptor module C contains RRRR or
RRRR
or SSSS enantiomers
× 2different
× 2 = 60library
different
library [51].
members
[51].
13
SSSS
enantiomers
resulting inresulting
15 × 2 ×in2 15
= 60
members
Figure
13 Figure
reproduced
reproduced with permission from [52].
with permission from [52].
Molecules2016,
2016,21,
21,1535
1535
Molecules
1313ofof3535
2.6. On-Bead Combinatorial Approach of Individual Library Members
2.6. On-Bead Combinatorial Approach of Individual Library Members
Fréchet et al. used an on-bead combinatorial approach to the rational design of optimized CSPs
FréchetAt
et first
al. used
an on-bead
combinatorial
approach
to the rational
design
optimized
CSPs
for HPLC.
a library
of 3 × 12
= 36 L-α-amino
acid anilides
(consisting
of 3ofα-amino
acids
and
for
HPLC.
At
first
a
library
of
3
×
12
=
36
L
-α-amino
acid
anilides
(consisting
of
3
α-amino
acids
12 aromatic amines) was synthesized in solution, then attached to functionalized macroporous
and
12 aromatic
amines) was synthesizedbeads
in solution,
then attached
to functionalized
macroporous
polymeric
methacrylate/dimethacrylate
which were
packed into
a ‘library column’
and tested
polymeric
methacrylate/dimethacrylate
beads
which
were
packed
into
a
‘library
column’
and
tested for
for the enantioseparation of the target analyte. The lead selector of the library, i.e., L-proline-1the
enantioseparation
of
the
target
analyte.
The
lead
selector
of
the
library,
i.e.,
L
-proline-1-indananilide,
indananilide, employed for the HPLC enantioseparation of the target racemate N-(3,5-dinitrobenzoyl)employed
for the HPLC
of the target
racemate
leucine diallylamide,
was enantioseparation
identified by a deconvolution
process
usingN-(3,5-dinitrobenzoyl)-leucine
11 ‘sublibrary columns’ of lower
diallylamide,
was
identified
by
a
deconvolution
process
using
11
‘sublibrary
of [53,54].
lower
diversity, each of which containing a CSP with a reduced number of library columns’
components
diversity,
each
of
which
containing
a
CSP
with
a
reduced
number
of
library
components
[53,54].
Fréchet et al. also combined the library approach and the reciprocal principle for the development of
Fréchet
et al. alsoCSPs
combined
the library
and the
for the development
enantioselective
for HPLC
[55,56]. approach
Thus according
to reciprocal
Figure 14,principle
a single enantiomer
(+)-T of the
oftarget
enantioselective
CSPs
for
HPLC
[55,56].
Thus
according
to
Figure
14,
a
single
enantiomer
racemate was at first immobilized onto a polymeric support and the resulting CSP was(+)-T
used
offor
the
racemate
was
first immobilized
onto
a polymeric
andprepared
the resulting
CSP
thetarget
HPLC
screening
of at
sublibraries
of racemic
compounds
thatsupport
have been
by parallel
was
used for the
HPLC screening
of sublibraries of member
racemic compounds
have
beenprepared
preparedas
combinatorial
synthesis.
The best enantioseparated
of this librarythat
(±)-S
is then
by
parallel
combinatorial
synthesis.
The
best
enantioseparated
member
of
this
library
)-S is
the single enantiomer (−)-S, coupled to a solid support and used in the second step for the(±
required
then
prepared as the
single
enantiomer
(−)-S,This
coupled
to amethodology
solid support
used
in the second
enantioseparation
of the
target
racemate (±)-T.
reciprocal
hasand
been
exemplified
for the
step
for
the
required
enantioseparation
of
the
target
racemate
(
±
)-T.
This
reciprocal
methodology
target racemate N-(3,5-dinitrobenzoyl)-leucine and the parallel library components of racemic 4-aryl-1,4has
been exemplified for
the target
and the
parallel library
dihydropyrimidines
obtained
in aracemate
one-pot N-(3,5-dinitrobenzoyl)-leucine
three-component reaction (Biginelli
dihydropyrimidine
components
of racemic
4-aryl-1,4-dihydropyrimidines
obtained in a one-pot
three-component
reactiona
three-component
condensation
reaction) [56]. Hydrogen-bonding
and π-π
interactions played
(Biginelli
dihydropyrimidine
three-component
condensation
reaction)
[56].
Hydrogen-bonding
and
dominant role in chiral recognition.
π-π interactions played a dominant role in chiral recognition.
Figure
Reciprocal
combinatorial
approach
to the to
optimization
of a CSP. of
Reproduced
with permission
Figure14.14.
Reciprocal
combinatorial
approach
the optimization
a CSP. Reproduced
with
ofpermission
the Royal Society
of Chemistry
from
[55]. from [55].
of the Royal
Society of
Chemistry
Librariesofofpotential
potentialchiral
chiralselectors
selectorshave
havealso
alsobeen
beenprepared
preparedby
bythe
theUgi
Ugithree-component
three-componentreaction
reaction
Libraries
β-lactams and
and screened
screened for
using
thethe
reciprocal
approach
involving
a CSP
ofofβ-lactams
for their
theirenantioselectivity
enantioselectivitybyby
using
reciprocal
approach
involving
a
comprised
of
the
immobilized
model
target
compound
N-(3,5-dinitrobenzoyl)L
-leucine
[57].
The
CSP comprised of the immobilized model target compound N-(3,5-dinitrobenzoyl)-L-leucine [57].
lead
candidate
identified
from
of 2-oxo-azetidine
2-oxo-azetidine acetic
aceticacid
acid
The
lead
candidate
identified
fromthe
thelibrary
libraryofofracemic
racemic phenylamides
phenylamides of
derivativeswas
wassubsequently
subsequentlysynthesized
synthesizedininbulk.
bulk.The
Theup-scaled
up-scaledracemate
racematewas
wasthen
thenresolved
resolvedby
bychiral
chiral
derivatives
preparative
HPLC
and
one
enantiomer
was
immobilized
on
a
solid
support.
The
obtained
CSP
showed
preparative HPLC and one enantiomer was immobilized on a solid support. The obtained CSP showed
enantioselectivities of α = 3 for various amino acid derivatives. Interestingly, the enantioselectivities
Molecules 2016, 21, 1535
14 of 35
enantioselectivities of α = 3 for various amino acid derivatives. Interestingly, the enantioselectivities
observed were even higher than those obtained during the reciprocal screening on the immobilized
target compound N-(3,5-dinitrobenzoyl)-L-leucine [57].
2.7. Batch-Screening of Peptide Libraries and the Reciprocal Principle
Previous studies by Pirkle, Hyun et al. had shown that the enantiomers of racemic
N-(3,5-dinitrobenzoyl)-peptide selectands are enantioseparated on a number of CSPs [58]. It was
therefore reasoned that in a reciprocal fashion N-(3,5-dinitrobenzoyl)-peptide CSPs should be capable
of enantioseparating various racemic compounds [59]. In an off-column batch-screening approach,
Welch et al. probed the selector-selectand interaction directly, rather than resorting to a bonded
analogue of the selectand used as CSP and packed into a column [60,61]. Thus the individual
components of a parallel library of 50 silica-supported N-(3,5-dinitrobenzoyl)-dipeptide CSPs were
prepared by solid phase synthesis on aminopropylsilica (APS) particles [59]. Whereas the ‘aa1’
position (Figure 15, top) consisted of the five D-α-amino acids Phg, Val, Pro, Gln and Phe, the
‘aa2’-position (Figure 15, top) was comprised of the same five α-amino acids in the L- and D-form,
respectively, giving rise to 50 different dipeptides. The members of the library were subsequently
screened for the enantioselective recognition of the test racemate N-(2-naphthyl)alanine diethylamide
by an efficient batch-screening process. Thus individual vials containing approximately 50 mg of the
candidate CSPs were placed into an autosampler and a dilute solution of the target racemate was
added to each vial. The concentration (1 mL; 10−5 M in 2-propanol/n-hexane (20/80, v/v)) of the
analyte solution was low enough to avoid saturation of the available interaction sites on the CSP.
During equilibration using a rotary platform shaker, the individual enantiomers were partitioned
between the solid and liquid phases of the vial. After equilibration, enantioselective HPLC of 50 µL
injections of the supernatant solution in each vial afforded two peaks for the individual enantiomers.
A 1:1 ratio revealed no enantioselectivity of the adsorbent, whereas ratios up to 1:2.72 indicated
the presence of an enantioselective adsorbent. It was found that the homochiral combinations (DD)
of the library components exhibited a higher enantioselectivity than the heterochiral combinations
(LD) [60]. The presence of steric bulk in the ‘aa 2’ position and of hydrogen-bonding groups in
the ‘aa1’ position proved to be important for enantiomeric discrimination of the test racemate.
APS-bonded D-Gln-D-Val-DNB, D-Pro-D-Val-DNP and D-Gln-D-Phe-DNP were identified as the
leading enantioselective selectors for racemic N-(2-naphthyl)alanine diethylamide. Based on these
findings, an improved library of individual dipeptides was designed. It included amino acids with
side chain functional groups such as Glu, His, and Arg, Whereas the ‘aa1’ position (Figure 15, top)
consisted of the seven L-α-amino acids Gln, Asn, Ser, His, Asp, Arg and Glu, the ‘aa2’-position
(Figure 15, top) was comprised of the L-α-amino acids Val, Leu, Ileu, Phe, Tyr, tert-Leu and Trp and the
D -α-amino acids Val, Leu, Ileu, Phe, Tyr and Trp, respectively, giving rise to 91 different dipeptides.
N-(2-naphthyl)alanine diethylamide was enantioseparated on APS-bonded L-Glu-L-Leu-DNP with the
high enantioselectivity of α = 20.74 (Figure 15, bottom) [61]. Up to 100 mg of the compound could be
enantioseparated using an analytical column (4.6 mm × 250 mm I.D.) in a single injection. The versatile
approach of Welch et al. permitted the identification of the necessary structural requirements for chiral
recognition by comparing the relative performance of various CSPs in the library. The enantioselective
library approach, commercialized jointly by Regis Technology and MediChem Research [reported in:
Chem. Eng. News 1999, 11, p. 101], may also prove useful for the discovery of improved selectors for
other types of supramolecular chromatographic interactions.
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
15 of 35
15 of 35
Figure
Top: DNB-dipeptide
libraries
usedasasCSP
CSP for
for the
of N-(2-naphthyl)alanine
Figure 15.
Top:15.DNB-dipeptide
libraries
used
theenantioseparation
enantioseparation
of N-(2-naphthyl)alanine
diethylamide. Bottom: The optimized aminopropylsilica-bonded L-Glu-L-Leu-DNB selector enantioseparates
diethylamide. Bottom: The optimized aminopropylsilica-bonded L-Glu-L-Leu-DNB selector
the test racemate with high enantioselectivity. Column: 4.6 mm × 250 mm I.D., eluent: 2-propanol/nenantioseparates the test racemate with high enantioselectivity. Column: 4.6 mm × 250 mm I.D.,
hexane (20/80 v/v). Reprinted with permission from [61]. Copyright (1999) American Chemical Society.
eluent: 2-propanol/n-hexane (20/80 v/v). Reprinted with permission from [61]. Copyright (1999)
American
AnChemical
improvedSociety.
strategy for the evaluation of common CSPs by the straightforward single-vial-
equilibration-approach of Welch et al. was based on rapid LC/MS screening of isotopically labelled
pseudoenantiomers present in the supernatant instead of the HPLC analysis of the enantiomers [62].
An improved strategy for the evaluation of common CSPs by the straightforward single-vialConcurrently, another batch-wise screening method was described by Wang and Li [63]. The
equilibration-approach
of Welch
et to
al.that
wasofbased
rapid
screening
of incubation
isotopically
approach was almost
identical
Welchon
et al.
withLC/MS
the exception
that the
waslabelled
pseudoenantiomers
present
in
the
supernatant
instead
of
the
HPLC
analysis
of
the
enantiomers
performed with the selector attached to a polymeric resin, instead of a chromatographic support. The [62].
reliability of the
established
stepwise solid-phase
synthesis
considered
Concurrently,
another
batch-wise
screeningMerrifield-type
method waspeptide
described
bywas
Wang
and Li [63].
as an advantage.
However,
it wastoalso
recognized
would not
The approach
was almost
identical
that
of Welchthat
et the
al. screening
with theresult
exception
thatnecessarily
the incubation
correlate with the chromatographic experiment as no chromatographic support was used for the batch
was performed
with the selector attached to a polymeric resin, instead of a chromatographic support.
screening step [38,39]. Crucial elements of the approach were the reciprocity of chromatographic
The reliability
of
the established stepwise solid-phase Merrifield-type peptide synthesis was considered
separation and the use of enantiomeric libraries. Thus, a potential chiral selector from the parallel library
as an advantage.
However,
it wasresin.
alsoThe
recognized
that the
screening
result
would
not necessarily
was linked onto
a solid-phase
racemic analyte
in the
proper solvent
was
then allowed
to
correlateequilibrate
with thewith
chromatographic
experiment
as noThe
chromatographic
was used
this potential selector
on the resin.
enantiomeric ratiosupport
of the analyte
in the for the
supernatant
was
analysed
after the
equilibration
period
by circular
dichroism.
A selective
adsorption
batch screening
step
[38,39].
Crucial
elements
of the
approach
were
the reciprocity
of chromatographic
of one
of the
the two
to the resin
was indicative
an efficient
chiralselector
selector. from
It wasthe
thenparallel
separation
and
useenantiomers
of enantiomeric
libraries.
Thus, aofpotential
chiral
linked onto a chromatographic support, and its enantioselectivity toward the target analyte was
library was linked onto a solid-phase resin. The racemic analyte in the proper solvent was then allowed
evaluated. The feasibility of the parallel library screening procedure was compared with the model
to equilibrate
this
potential
selector
onenantioseparation
the resin. The enantiomeric
ratio of ester
the analyte
system with
studied
earlier
of the chiral
HPLC
of N-(1-naphthyl)leucine
using a in the
supernatant
was analysed
after
the equilibration period by circular dichroism. A selective adsorption
16-member
small library
[50].
al. simplified the
reciprocal
using onlyof
one
(all-L)chiral
which was
equilibrated
of one of theBluhm
two et
enantiomers
to the
resinsystem
was by
indicative
anlibrary
efficient
selector.
It was then
on two
columns containing
the immobilized
target racemate (1-naphthyl)leucine
ester either
in
linked onto
a different
chromatographic
support,
and its enantioselectivity
toward the target
analyte
was
the
L- or enantiomeric D-form, followed by incubation of each CSP with a mixture library [64]. The
evaluated. The feasibility of the parallel library screening procedure was compared with the model
adsorbed members of the library on the two columns were washed off and analysed by a reversed
system studied earlier of the chiral HPLC enantioseparation of N-(1-naphthyl)leucine ester using a
phase HPLC assay. Components with the same retention time but different intensities in both
16-member
small library
[50].
chromatograms
represented
an optimized selector. Its chemical structure was then determined by
Bluhm
et oral.LC-MS-MS.
simplified
thereciprocal
reciprocal
system
by using candidate,
only onei.e.,
library
(all-L)N-3,5which was
LC-MS
In the
fashion,
the optimized
immobilized
dinitrobenzoylL-leucinecolumns
ester, wascontaining
used as CSP
to immobilized
enantioseparatetarget
racemic
N-(1-naphthyl)leucine
equilibrated
on two different
the
racemate
(1-naphthyl)leucine
ester in
with
theLhigh
α =followed
12 [64]. In an
study,
200-member
parallel
ester either
the
- or enantioseparation
enantiomeric Dfactor
-form,
byoptimized
incubation
of aeach
CSP with
a mixture
library [64]. The adsorbed members of the library on the two columns were washed off and
analysed by a reversed phase HPLC assay. Components with the same retention time but different
intensities in both chromatograms represented an optimized selector. Its chemical structure was
then determined by LC-MS or LC-MS-MS. In the reciprocal fashion, the optimized candidate, i.e.,
immobilized N-3,5-dinitrobenzoyl-L-leucine ester, was used as CSP to enantioseparate racemic
N-(1-naphthyl)leucine ester with the high enantioseparation factor α = 12 [64]. In an optimized
Molecules 2016, 21, 1535
16 of 35
study, a 200-member parallel dipeptide library was created and screened with 10-undecenyl-N(1-naphthyl)leucinate [65]. The library was prepared using Fmoc-protected α-amino acids with
amino-methylated polystyrene (AMPS) resin as solid support. The library contained three modules.
Modules 2 and 3 consisted of ten different L-α-amino acids, whereas the N-terminating module 1 was
an acetyl group or a 3,5-dinitrobenzoyl (DNB) group. Each single library member was incubated with
a solution of racemic 10-undecenyl-N-(1-naphthyl)leucinate. After equilibration, the supernatant was
collected and assessed for enantiomeric excess ee by circular dichroism (CD).
In another study, an 81-member dipeptide library was evaluated for the target compound
containing a stereogenic phosphorus atom, i.e., tert-butyl-1-(2-methylnaphthyl)phosphane oxide [66].
The member with the highest enantioselectivity was DNB-D-Asn-L-Thr-Abu-AMPS (Abu = 4-aminobutanoic acid). It was attached to aminopropylsilica and used as CSP for the target racemate in HPLC.
However, the enantioselcetivities between the batch-experiment and the chromatographic experiment
did not match due to the different solid supports. Therefore a longer linker had to be employed to
immobilize the chiral selector onto silica to observe an enantioseparation factor of α = 3.2 for the
racemic phosphine oxide. A 121-member peptide library was also tested for the enantioseparation of
atropisomeric 1,10 -bi-2,20 -naphthol. A trityl-protected di-asparagine selector (Fmoc-Asn(Trt)-Asn(Trt))
yielded a separation factor of α = 7.2 [67]. By the reciprocal principle, it can be assumed that the target
racemate 1,10 -bi-2,20 -naphthol, for which a highly enantioselective peptide has been identified, could
also be used as a CSP for the enantioseparation of the racemic peptides.
Another high throughput screening protocol was proposed for chiral selector discovery. It was
modeled after the protocol for biological screening of candidate drugs from chemical libraries.
The procedure was based on target distribution between an aqueous phase and an organic phase [68].
3. Single-Walled Carbon Nanotubes, SWCNTs
In single-walled carbon nanotubes (SWCNTs), a single layer graphene sheet is rolled-up into
a tubular structure. The use of SWCNTs as selective stationary phases for the chromatographic
separation of various classes of achiral compounds has been reviewed [1]. SWCNTs exist in three forms,
i.e., two achiral structures (armchair and zig-zag) and a chiral structure. Chiral single-walled carbon
nanotubes (n,m)-SWCNTs are characterized by n,m- indices, where n and m are coordination numbers
of the carbon atoms in the hexagonal network. By convention, when n is greater than m, the species is
right-handed or P (for positive), and when m is greater than n, the species is left-handed or M (for negative).
(n,m)-SWCNTs represent interesting supramolecular entities for the reciprocal chromatographic
selector/selectand relationship. The enantiomers of nine pre-separated single chirality (n,m)-SWCNTs
of varying enantiomeric purity were separated on Sephacryl gel containing allyl-dextran as the
chiral selector by gel permeation chromatography (Figure 16) [69]. The successful enantioseparation
was proved by their opposite CD-spectra. The enantiomeric purities are not known and they
varied for different (n,m)-single chirality SWCNTs with the second eluted enantiomer being more
enriched [69]. The fractionation of (n,m)-SWCNTs by countercurrent chromatography also led to chiral
recognition [70]. The method was based on the partition of sodium deoxycholate (SDC) dispersed
(n,m)-SWCNTs in a polyethylene glycol/dextran aqueous two-phase system. The (n,m)-dependent
bimodal elution peak width of the chiral nanotubes were interpreted as enantiomeric recognition
in the presence of chiral SDC and polymeric dextran [70]. Dextran represents a complex branched
glucan (a polysaccharide made of D-glucose building blocks). By the reciprocal approach, nonracemic
(n,m)-SWCNTs may be considered in the future as novel CSPs for the enantiomeric separation of
lower linear dextrins (‘acyclodextrins’, maltooligosaccharides) which are readily available in both
enantiomeric forms obtained from D- and L-glucose, respectively. Incidentally, acetylated/silylated
dextrins with a different degree of oligomerization and devoid of a molecular cavity have been used
as CSPs for the enantioseparation of derivatized amino acids and various underivatized racemic
compounds by GC, thus highlighting the role of the polar external surface of the selector for chiral
recognition in contrast to the well-established inclusion mechanism exerted by cyclodextrins [71].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
17 of 35
17 of 35
Figure 16. Illustration of the enantioseparation of P- and M-SWCNTs by gel permeation liquid
Figure 16. Illustration of the enantioseparation of P- and M-SWCNTs by gel permeation liquid
chromatography on an allyl-dextran stationary phase. Reprinted with permission from [69]. Copyright
chromatography on an allyl-dextran stationary phase. Reprinted with permission from [69]. Copyright
(2014) American Chemical Society.
(2014) American Chemical Society.
Using molecular dynamics simulations, Raffaini and Ganazzoli investigated in a theoretical
Using
simulations,
Raffaini andformed
Ganazzoli
in aacids
theoretical
study
themolecular
adsorptiondynamics
and denaturation
of an oligopeptide
by 16investigated
chiral L-α-amino
and
study
the aadsorption
and denaturation
an oligopeptide
formed
16 outer
chiralsurface
L -α-amino
acids
and
having
helical structure
in the nativeofstate
on both the inner
andbythe
of the
chiral
having
a helical
structure in thewith
native
state on handedness,
both the inner
the armchair
outer surface
of the chiral
(10,20)and (20,10)-SWCNTs
an opposite
andand
of the
(16,16)-SWCNT
nanotube
a similar diameter
[72]. The molecular
indicated
that the
(10,20)and with
(20,10)-SWCNTs
withfor
ancomparison
opposite handedness,
and ofsimulations
the armchair
(16,16)-SWCNT
enantioseparation
of the
chiral carbon
nanotubes[72].
by chromatographic
methods are
feasible that
usingthe
nanotube
with a similar
diameter
for comparison
The molecular simulations
indicated
oligopeptides of aofsufficient
length.
It was
also suggested
that natural oligopeptides
could
not only
enantioseparation
the chiral
carbon
nanotubes
by chromatographic
methods are
feasible
using
be
used
to
separate
enantiomeric
nanotubes,
but
in
the
reciprocal
fashion
also
chiral
nanotubes
of be
oligopeptides of a sufficient length. It was also suggested that natural oligopeptides could not only
single
handedness
could
be
considered
as
chiral
selectors
for
racemic
oligopeptides
(Giuseppina
used to separate enantiomeric nanotubes, but in the reciprocal fashion also chiral nanotubes of single
Raffiani, personal
communication,
handedness
could be
considered as2016).
chiral selectors for racemic oligopeptides (Giuseppina Raffiani,
Unspecified
chiral
(n,m)-SWCNTs
were tested as CSPs for the enantioseparation of twelve classes
personal communication, 2016).
of racemic pharmaceuticals by nano-LC [73,74]. The finding would require that the commercially
Unspecified chiral (n,m)-SWCNTs were tested as CSPs for the enantioseparation of twelve
available chiral (6,7)-SWCNT (carbon >90%, 0.7–0.9 nm diameter by fluorescence) obtained from Sigmaclasses of racemic pharmaceuticals by nano-LC [73,74]. The finding would require that the
Aldrich was nonracemic or even enantiomerically pure. Neither the mode of the required enantiomeric
commercially available chiral (6,7)-SWCNT (carbon >90%, 0.7–0.9 nm diameter by fluorescence)
enrichment of the synthetic racemic (6,7)-SWCNT nor its chiroptical data was reported. Some racemic
obtained from Sigma-Aldrich was nonracemic or even enantiomerically pure. Neither the mode
compounds showed a deviation of the expected 1:1 peak ratio and no enantioseparation was noticed for
of the
the chiral
required
enantiomeric
of theofsynthetic
racemic (6,7)-SWCNT
nor its
chiroptical
(7,8)-SWCNTs.
Alsoenrichment
in another report
the continuous-flow
enantioseparation
of carvedilol
data
was
reported.
Some
racemic
compounds
showed
a
deviation
of
the
expected
1:1
peak ratio
with fluorescent detection on chiral carbon nanotubes, chiroptic data and enantiomeric compositions
and
no
enantioseparation
was
noticed
for
the
chiral
(7,8)-SWCNTs.
Also
in
another
report
of the
of the chiral selector were not mentioned [75]. As chiral SWCNTs contain equal amounts of left- and
continuous-flow
enantioseparation
of carvedilol
fluorescent
detection
chiral carbon nanotubes,
right-handed helical
structures, their
use as CSPwith
requires
the separation
of theseon
non-superimposable
mirror
chiroptic
data
and
enantiomeric
compositions
of
the
chiral
selector
were
not
mentioned
[75].
As
image forms into single enantiomers of defined enantiomeric purity as outlined by Peng et al. [76]. chiral
SWCNTs
equal
amountswhether
of left- and
helical
their use as CSP requires
In contain
order to
investigate
the right-handed
use of SWCNTs
canstructures,
improve enantioseparations
on anthe
separation
thesecomprised
non-superimposable
mirror
image forms
intoliquid
single(R)-N,N,N-trimethyl-2-aminobutanolenantiomers of defined enantiomeric
existingofCSP,
of the chiral
nonracemic
ionic
bis(trifluoromethane-sulfon)imidate,
purity
as outlined by Peng et al. [76]. two capillary columns, i.e., one containing the chiral ionic liquid
alone
and the
containing
the chiral
liquid
together
with
the SWCNTs
were tested by on
GCan
In order
toother
investigate
whether
the ionic
use of
SWCNTs
can
improve
enantioseparations
[77]. ItCSP,
was shown
that the
column
containing
SWCNTs
improved the enantioselectivity of the
existing
comprised
of capillary
the chiral
nonracemic
ionic
liquid (R)-N,N,N-trimethyl-2-aminobutanolchiral ionic liquid. The ‘nonchiral’ role
of capillary
the SWCNT
was due
to one
an increase
of the
surface
ofliquid
the
bis(trifluoromethane-sulfon)imidate,
two
columns,
i.e.,
containing
the
chiralarea
ionic
inner
capillary
wall
by
formation
of
a
layer
with
a
skeletal
network,
leading
the
enhanced
retention
alone and the other containing the chiral ionic liquid together with the SWCNTs were tested by GC [77].
times,
whichthat
improved
the resolution
Rs for the
enantiomers
[77]. the enantioselectivity of the
It was
shown
the capillary
columnfactor
containing
SWCNTs
improved
chiral ionic liquid. The ‘nonchiral’ role of the SWCNT was due to an increase of the surface area of the
4. Chromatography of Fullerenes and the Reciprocal Principle
inner capillary wall by formation of a layer with a skeletal network, leading the enhanced retention
The molecular
recognition
of fullerene
various selectors
times, which
improved
the resolution
factorcongeners
Rs for the on
enantiomers
[77]. by liquid chromatography
has been reviewed [1,78,79], including enantioseparation of chiral native fullerenes or its derivatives
4. [1].
Chromatography
of Fullerenes
and the Reciprocal
For more selective
molecular recognition
leading toPrinciple
large separation factors between C60 and C70
(α > 2), a supramolecular approach was first employed by Hawkins et al. [80]. As a strong π-acceptor
The molecular recognition of fullerene congeners on various selectors by liquid chromatography
has been reviewed [1,78,79], including enantioseparation of chiral native fullerenes or its derivatives [1].
Molecules 2016, 21, 1535
18 of 35
Molecules2016,
2016,21,
21,1535
1535
Molecules
18of
of35
35
18
For more selective molecular recognition leading to large separation factors between C60 and C70
(α
> 2), for
afor
supramolecular
approach
was first aemployed
by Hawkins
et al. [80]. As a strong π-acceptor
selector
theπ-donor
π-donorfullerene
fullerene
selectands,
aCSP
CSPcontaining
containing
N-(3,5-dinitrobenzoyl)-phenylglycine
selector
the
selectands,
N-(3,5-dinitrobenzoyl)-phenylglycine
selector for
the π-donor
fullerene
CSP
containing N-(3,5-dinitrobenzoyl)-phenylglycine
(DNBPG),
originally
developed
byselectands,
Pirkleet
etal.
al.afor
for
enantioseparations
[23],was
wasemployed
employed(Figure
(Figure17).
17).
(DNBPG),
originally
developed
by
Pirkle
enantioseparations
[23],
(DNBPG), originally developed by Pirkle et al. for enantioseparations [23], was employed (Figure 17).
Figure
17.
HPLC
separation
of
(12.2
min)
and
(23.5
min)
on
Pirkle-type
ionically
DNBPG
Figure17.
17.HPLC
HPLCseparation
separationof
ofCC6060
(12.2min)
min)and
andCC
C707070(23.5
(23.5min)
min)on
onaaaPirkle-type
Pirkle-typeionically
ionicallyDNBPG
DNBPG
Figure
60(12.2
containing
column
(250
×
10
mm
I.D.)
eluted
with
n-hexane
at
5.0
mL/min
and
detected
at
280
nm,
containingcolumn
column(250
(250× ×
mm
I.D.)
eluted
with
n-hexane
mL/min
detected
at 280
nm,
containing
1010
mm
I.D.)
eluted
with
n-hexane
at at
5.05.0
mL/min
andand
detected
at 280
nm,
αα
2.25
(r.t.).
Reprinted
with
permission
from
[80].
Copyright
(1990)
American
Chemical
Society.
= 2.25
(r.t.).
Reprinted
with
permission
from
[80].
Copyright
(1990)
American
Chemical
Society.
==α2.25
(r.t.).
Reprinted
with
permission
from
[80].
Copyright
(1990)
American
Chemical
Society.
An
unprecedented
increase
in
retention of
of C
C6060 and
and CC7070with
with increasing
increasing column
column temperature
temperature was
was
An
Anunprecedented
unprecedentedincrease
increasein
in retention
retention
of
C
60 and C70 with increasing column temperature was
observed
by
Pirkle
and
Welch
on
the
covalently
bonded
π-acidic
N-(3,5-dinitrobenzoyl)-phenylglycine
observed
observedby
byPirkle
Pirkleand
andWelch
Welchon
onthe
thecovalently
covalentlybonded
bondedπ-acidic
π-acidicN-(3,5-dinitrobenzoyl)-phenylglycine
N-(3,5-dinitrobenzoyl)-phenylglycine
(DNBPG)
containing
column
(250
×
4.6
mm
I.D.)
(Figure
18)
[81].
Van’t
Hoff
plots
revealed
therare
rarecase
case
(DNBPG)
containing
column
(250
×
4.6
mm
I.D.)
(Figure
18)
[81].
Van’t
plots
revealed
(DNBPG) containing column (250 × 4.6 mm I.D.) (Figure
18)
[81].Hoff
Van’t
Hoff
plots the
revealed
the
of
positive
enthalpy
and
entropy
values,
i.e.,
the
gain
in
entropy
is
accompanied
by
an
endothermic
ofrare
positive
enthalpy
entropyand
values,
i.e., the
gain i.e.,
in entropy
accompanied
by an endothermic
case of
positiveand
enthalpy
entropy
values,
the gainis in
entropy is accompanied
by an
adsorption.
adsorption.
endothermic adsorption.
Figure
18.N-(3,5-dinitrobenzoyl)-phenylglycine
N-(3,5-dinitrobenzoyl)-phenylglycine
(DNBPG)
bonded
toReprinted
silica. Reprinted
Reprinted
with
Figure 18.
(DNBPG)
bonded
to silica.
with permission
Figure
18.
N-(3,5-dinitrobenzoyl)-phenylglycine
(DNBPG)
bonded
to
silica.
with
from [81]. Copyright
(1990)
American
permission
from[81].
[81].Copyright
Copyright
(1990)Chemical
AmericanSociety.
ChemicalSociety.
Society.
permission
from
(1990)
American
Chemical
Tripodal
3,5-dinitrobenzoate
ester
and
2,4-dinitrophenyl
ether
stationary
phases
containing
Tripodal
Tripodal 3,5-dinitrobenzoate
3,5-dinitrobenzoate ester
ester and
and 2,4-dinitrophenyl
2,4-dinitrophenyl ether
ether stationary
stationary phases
phases containing
containing
three
π-acidic
functionalities
for
simultaneous
multipoint
supramolecular
interaction
with
fullerenes
three
threeπ-acidic
π-acidicfunctionalities
functionalitiesfor
forsimultaneous
simultaneousmultipoint
multipointsupramolecular
supramolecularinteraction
interactionwith
withfullerenes
fullerenes
(‘Buckyclutcher’)
were
developed
and
they
provided
the
highest
retention
and
the
greatest
separation
(‘Buckyclutcher’)
were
developed
and
they
provided
the
highest
retention
and
the
greatest
separation
(‘Buckyclutcher’) were developed and they provided the highest retention and the greatest separation
factor
for
the
/C
70 mixture
mixture [82].
[82].
A remarkable
remarkable
separation
of 15
15
fullerene
congeners
in the
the
range
of
factor
/C
70
A
separation
of
congeners
in
range
of
factorfor
forthe
theCCC606060
/C
[82].
A
remarkable
separation
of fullerene
15
fullerene
congeners
in the
range
70 mixture
C
60
–C
96
,
including
enantiomerically
enriched
specimens,
by
HPLC
on
a
syndiotactic
poly(methyl
Cof
60–C
, including
enantiomerically
enriched
specimens,
by by
HPLC
onona asyndiotactic
C6096–C
enantiomerically
enriched
specimens,
HPLC
syndiotacticpoly(methyl
poly(methyl
96 , including
methacrylate)
selector
in
45
min
with
chlorobenzene
as
eluent
has
been
described
[83].
methacrylate)
selector
in
45
min
with
chlorobenzene
as
eluent
has
been
described
[83].
methacrylate) selector in 45 min with chlorobenzene as eluent has been described [83].
(R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneamino-oxy)propionic
acid(TAPA)
(TAPA)
(Figure
1b,
and
(R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneamino-oxy)propionic
acid
(TAPA)
(Figure
(R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneamino-oxy)propionic acid
(Figures
1b1b,
andand
19),
Figure
19),
a
strong
π-electron
acceptor
due
to
the
tetranitrofluorenylidene
moiety,
forms
supramolecular
Figure
19),π-electron
a strong π-electron
acceptor
due
to the tetranitrofluorenylidene
forms supramolecular
a strong
acceptor due
to the
tetranitrofluorenylidene
moiety,moiety,
forms supramolecular
charge
charge
transfer
complexes
withappropriate
appropriate
π-donor
molecules.
TAPA
wasto
used
toseparate
separate
andwith
C70
charge
transfer
complexes
with
π-donor
molecules.
was
used
to
CC6060C
and
transfer
complexes
with appropriate
π-donor
molecules.
TAPATAPA
was used
separate
C60 and
70 C70
with
a
separation
factor
α
up
to
2.6
and
the
same
peculiar
increase
of
retention
with
increasing
with
a separation
α 2.6
upand
to 2.6
same increase
peculiarofincrease
retention
withtemperature
increasing
a separation
factorfactor
α up to
the and
samethe
peculiar
retentionofwith
increasing
temperature
previously
observed
with
π-acidic
N-(3,5-dinitrobenzoyl)-phenylglycine
(DNBPG)
was
temperature
previously
observed
with
π-acidic
N-(3,5-dinitrobenzoyl)-phenylglycine
(DNBPG)
previously observed with π-acidic N-(3,5-dinitrobenzoyl)-phenylglycine (DNBPG) was foundwas
and
found
and
interpreted
as
arising
from
solute
solvation
at
low
temperatures
[84].
The
same
authors
found
and interpreted
as arising
solute at
solvation
at low temperatures
[84]. authors
The same
authors
interpreted
as arising from
solutefrom
solvation
low temperatures
[84]. The same
mentioned
mentioned
that
60and
and
can
behave
as both
both π-electron
π-electron
donor
and π-electron
π-electron
acceptor
(theaffinity
electron
mentioned
that
CC60can
CC7070 can
as
and
acceptor
(the
electron
that C60 and
C70
behave
asbehave
both π-electron
donor anddonor
π-electron
acceptor (the
electron
of
affinity
of
C
60amounts
amounts
to
2.6
eV).
affinity
of
C
60
to
2.6
eV).
C60 amounts to 2.6 eV).
The
enantioseparation
of the
the tris-adduct
tris-adductC
60[C(COOEt)
[C(COOEt)22]]]33 and
and hexakis-adduct
hexakis-adduct CC6060[C(COOEt)
[C(COOEt)22]]66
The
tris-adduct
CC60
60
Theenantioseparation
enantioseparation of
of
the
[C(COOEt)
2 3 and hexakis-adduct C60 [C(COOEt)2 ]6
entailed
with
an
inherent
chiral
substitution
pattern
(Figure
20)
has
also
been
described
on
the
CSP
entailed
entailedwith
withan
aninherent
inherentchiral
chiralsubstitution
substitutionpattern
pattern(Figure
(Figure20)
20)has
hasalso
alsobeen
beendescribed
describedon
onthe
theCSP
CSP
(R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneaminooxy)propionic
acid
(TAPA)
bonded
to
silica
gel
(R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneaminooxy)propionic
acid
(TAPA)
(R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneaminooxy)propionic
acid
(TAPA)bonded
bondedto
tosilica
silicagel
gel
(Figure
19)
by
micro
HPLC
[85].
(Figure
(Figure19)
19)by
bymicro
microHPLC
HPLC[85].
[85].
Molecules 2016, 21, 1535
Molecules 2016,
2016, 21,
21, 1535
1535
Molecules
19 of 35
19 of
of 35
35
19
Structure of
of (R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneamino-oxy)propionic
(R)-(−)-2-(2,4,5,7-tetranitro-9-fluorenylideneamino-oxy)propionicacid
acid(TAPA)
(TAPA)
Figure 19. Structure
bonded
to
silica
gel
[4,84].
bonded to silica gel [4,84].
60
The
ofof
C60
20,20,
top)
hashas
been
achieved
by
The enantioseparation
enantioseparationof
ofbis-,
bis-,tristris-and
andhexakis-adducts
hexakis-adducts
C(Figure
top)
been
achieved
60 (Figure
HPLC
on
the
Whelk-O1-CSP
[86].
As
the
Whelk-O1-CSP
(Figure
3c)
was
originally
optimized
for
by HPLC on the Whelk-O1-CSP [86]. As the Whelk-O1-CSP (Figure 3c) was originally optimized
60 may
racemic
naproxen
[23], [23],
it may
be speculated
that some
bis-, trisof C60
be
for racemic
naproxen
it may
be speculated
that some
bis-,and
tris-hexakis-adducts
and hexakis-adducts
of Calso
60 may
enantioseparated
by
the
silica-bonded
naproxen
selector
(Figure
3a)
via
a
twofold
reciprocal
recognition
also be enantioseparated by the silica-bonded naproxen selector (Figure 3a) via a twofold reciprocal
scenario.
recognition scenario.
Figure
of the
tris-adduct
C60
60[C(COOEt)22]33. Bottom: Enantioseparation of a hexakisFigure 20.
20. Top:
Top:Structure
Structure
of the
tris-adduct
C60 [C(COOEt)2 ]3 . Bottom: Enantioseparation of a
60
[C(COOEt)
2
]
6
on
the
CSP
TAPA
bonded
to
aminopropyl
silica gel silica
by micro
HPLC
(α HPLC
= 1.1).
adduct
C
60
2 6
hexakis-adduct
C60 [C(COOEt)
]
on
the
CSP
TAPA
bonded
to aminopropyl
gel by
micro
2 6
Packed
capillarycapillary
(20 cm ×(20
0.25
I.D.),
acetonitrile/water
(75/25, (75/25,
v/v), room
(α = 1.1).fused-silica
Packed fused-silica
cmmm
× 0.25
mmμL/min
I.D.), µL/min
acetonitrile/water
v/v),
temperature.
Reprinted
with
permission
from
[85].
room temperature. Reprinted with permission from [85].
In order to optimize supramolecular recognition between fullerenes and polyaromatic
In order to optimize supramolecular recognition between fullerenes and polyaromatic compounds
compounds (PAHs), the reciprocal principle of changing the chromatographic roles of selector and
(PAHs), the reciprocal principle of changing the chromatographic roles of selector and selectand
selectand has been suggested by Saito et al. [87]. In liquid chromatography, fullerene selectors can be
has been suggested by Saito et al. [87]. In liquid chromatography, fullerene selectors can be used
used as supramolecular stationary phases either as native crushed specimen packed into capillaries or
as supramolecular stationary phases either as native crushed specimen packed into capillaries
60 phase (Figure 21)
covalently linked to a functionalized silica gel [78,88,89]. The chemically bonded C60
or covalently linked to a functionalized silica gel [78,88,89]. The chemically bonded C60 phase
[87] showed preferential interaction with triphenylene and perylene which possess partial structures
(Figure 21) [87] showed preferential interaction with triphenylene and perylene which possess partial
similar to that of C60
60. Interestingly, larger retention factors are observed for nonplanar vs. planar PAH
structures similar to that of C60 . Interestingly, larger retention factors are observed for nonplanar
60
molecules of comparable molecular size. Thus effective molecular recognition occurs between the C60
vs. planar PAH molecules of comparable molecular size. Thus effective molecular recognition
selector and nonplanar selectands with a similar molecular curvature of the fullerene [87]. Bonded C60
60
occurs between the C60 selector and nonplanar selectands with a similar molecular curvature of
60 and C70
70 [87,88]. As
phases were also used for self-recognition, i.e., for the congener separation of C60
the fullerene [87]. Bonded C60 phases were also used for self-recognition, i.e., for the congener
60 phase entailed
compared to an octadecylsilica (ODS) reversed phase devoid of aromatic moieties, a C60
separation of C60 and C70 [87,88]. As compared to an octadecylsilica (ODS) reversed phase devoid
high retention factors
for C60
60 and C70
70 and an enhanced separation factor (α = 2.9) [87].
of aromatic moieties, a C60 phase entailed high retention factors for C60 and C70 and an enhanced
separation factor (α = 2.9) [87].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
20 of 35
20 of 35
Figure 21.
21. Synthetic
Synthetic scheme
scheme for
for aaCC6060 bonded
bonded silica
silica stationary
stationary phase.
phase. Reproduced
Reproduced with
with permission
permission
Figure
from [87].
[87]. Copyright
Copyright Wiley-VCH,
Wiley-VCH, Weinheim,
Weinheim,Germany.
Germany.
from
Fullerenes have
have also
also been
been used
used as
as stationary
stationary phases
phases in
in gas
gas chromatography
chromatography (GC).
(GC). A
A glass
glass capillary
capillary
Fullerenes
column
(15
m
×
0.29
mm
I.D.)
was
coated
with
pure
C60
(film
thickness
0.1
μm)
using
a
high-pressure
column (15 m × 0.29 mm I.D.) was coated with pure C60 (film thickness 0.1 µm) using a high-pressure
static method
method[90].
[90].The
Thethermostable
thermostablesolid
solid
C60
phase
behaved
liquid
phase
squalane
its
static
C60
phase
behaved
likelike
thethe
liquid
phase
squalane
in itsinvan
vanWaals
der Waals
interaction
n-alkanes.
der
interaction
with with
n-alkanes.
C
60 linked to poly(dimethylsiloxane) (Figure 22, top) was coated on a 10 m × 0.25 mm I.D. fused
C60 linked to poly(dimethylsiloxane) (Figure 22, top) was coated on a 10 m × 0.25 mm I.D.
silica
and used
thefor
selective
separation
of isomeric
hexachlorobiphenyls
differing
in orthofused column
silica column
andfor
used
the selective
separation
of isomeric
hexachlorobiphenyls
differing
in
chlorine
substitution
pattern.
Thermodynamic
parameters
were
obtained
by
the
retention-increment
ortho-chlorine substitution pattern. Thermodynamic parameters were obtained by the retention-increment
method RR′0 [91,92].
[91,92]. It
It separates
separates supramolecular
method
supramolecular interactions
interactions from
from non-specific
non-specific interactions
interactions of
of the
the linker.
linker.
It
was
shown
that
the
retention
behaviour
of
PCBs
depends
strongly
on
the
degree
of
ortho-chlorine
It was shown that the retention behaviour of PCBs depends strongly on the degree of ortho-chlorine
substitution which
which in
in turn
turn is
is linked
linked with
with non-planarity
non-planarity (Figure
(Figure 22,
22, bottom).
bottom). The
The deviation
deviation from
from
substitution
planarity
of
PCBs
strongly
decreases
the
supramolecular
interaction
with
C
60 [93]. Since planar PCBs
planarity of PCBs strongly decreases the supramolecular interaction with C60 [93]. Since planar PCBs
are more
more toxic
toxic than
than non-planar
non-planar PCBs,
PCBs, aa tentative
tentative link
link of
ofretention
retentionbehaviour
behaviouron
onCC60 and
and toxicity
toxicity has
has
are
60
been
advanced
[93].
been advanced [93].
In Table
Table44 the
the selective
selective interaction
interaction of
of aromatic
aromatic compounds
compoundswith
withC
C60 linked
to poly(dimethylsiloxane)
In
60 linked to poly(dimethylsiloxane)
(Figure 22,
byby
thethe
retention-increment
R’, i.e.,
value
two indicates
that the retention
(Figure
22,top)
top)isisexpressed
expressed
retention-increment
R0 ,a i.e.,
a of
value
of two indicates
that the
of
the
PCB
congener
on
C
60 is twice as compared to a reference column devoid of the fullerene
retention of the PCB congener on C60 is twice as compared to a reference column devoid of the
selector [94].
fullerene
selector [94].
Table 4.4. Retention-increments
Retention-increments R’
for the
the selective
selective interaction
interaction of
of aromatic
aromatic compounds
compounds with
with C
C60..
Table
R0 for
60
60 phase (0.25 μm).
Complexation
column
(k):
10
m
x
0.25
mm
I.D.
fused
silica
column
coated
with
the
C
Complexation column (k): 10 m x 0.25 mm I.D. fused silica column coated with the C60 phase
without (k
C060, ):without
10 m × C
0.25
mm I.D. fused silica column coated with dimethylReference
(k0, column
(0.25
µm). column
Reference
60 ): 10 m × 0.25 mm I.D. fused silica column coated
acetylaminopropyl(methyl)polysiloxane
(0.25 μm) [94].(0.25 µm) [94].
with
dimethyl-acetylaminopropyl(methyl)polysiloxane
Analyte
Analyte
1.2-Dichlorobenzene
1.2-Dichlorobenzene
2.6-Dichlorotoluene
2.6-Dichlorotoluene
Nitrobenzene
Nitrobenzene
2-Nitrotoluene
2-Nitrotoluene
Aniline
Aniline
Phenol
Phenol
Naphthalene
Naphthalene
1-Methylnaphthalene
2-Methylnaphthalene
1-Methylnaphthalene
Indole
2-Methylnaphthalene
Quinoline
Indole
Isoquinoline
Quinoline
4-Chlorophenol
Isoquinoline
2.4-Dichlorophenol
2.4.6-Trichlorophenol
4-Chlorophenol
2.4-Dichlorophenol
2.4.6-Trichlorophenol
T (°C)
R’ = 0(k − k0)/k0
T (◦ C)
R = (k − k0 )/k0
80
0.61
80
80
0.690.61
80
0.69
80
0.950.95
80
80
0.910.91
80
80
80
1.581.58
90
90
0.500.50
90
0.73
90
0.730.83
120
120
120
0.830.70
120
120
0.701.28
120
120
1.281.87
120
2.94
120
1.870.34
120
120
2.940.77
120
120
120
0.341.29
120
0.77
120
1.29
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
21 of 35
21 of 35
Figure
C60
60 linked
linked to
to aminopropyl
aminopropyl poly(dimethylsiloxane).
poly(dimethylsiloxane). Bottom:
Figure 22.
22. Top:
Top: Synthesis
Synthesis of
of C
Bottom: Gas
Gas
chromatographic
elution
order
of
hexachlorobiphenyl
congeners
with
different
degree
of
ortho-chloro
chromatographic elution order of hexachlorobiphenyl congeners with different degree of ortho-chloro
substitution.
substitution.The
Theincreased
increasedretention
retentiontime
timeof
ofthe
thesecond
secondeluted
elutedenantiomer
enantiomerisismarked
markedwith
withan
anarrow.
arrow.
◦
10
10 m
m ××0.25
0.25mm
mmI.D.
I.D.fused
fusedsilica
silicacolumn
column coated
coated with
with the
the C
C6060phase
phase(0.25
(0.25μm),
µm),190
190 °C,
C,carrier:
carrier: 0.6
0.6bar
bar
(at
(at gauge)
gauge) helium,
helium, electron
electron capture
capture detection.
detection. Reprinted
Reprinted with
with permission
permission from [93].
5.
5. The
TheReciprocal
ReciprocalSelectand/Selector
Selectand/SelectorSystem
SystemofofFullerenes/Cyclodextrins
Fullerenes/Cyclodextrins
n =n ,
A
system
parpar
excellence
is represented
by fullerenes
(Cn, (C
A supramolecular
supramolecularselectand/selector
selectand/selector
system
excellence
is represented
by fullerenes
number
of
carbon
atoms)
and
cyclodextrins
(CDn,
n
=
number
of
D
-glucose
building
blocks).
It
also
n = number of carbon atoms) and cyclodextrins (CDn, n = number of D-glucose building blocks).
lends
the principle
of reciprocal
recognition.
According
to Figure
23, the23,role
selector
and
It alsoitself
lendstoitself
to the principle
of reciprocal
recognition.
According
to Figure
theof
role
of selector
selectand
can
be
reversed.
and selectand can be reversed.
The
The non-chromatographic
non-chromatographic supramolecular
supramolecular recognition
recognition phenomenon
phenomenon between
between fullerenes
fullerenes and
and
cyclodextrins
is
well
established.
γ-Cyclodextrin
(CD8)
forms
water-soluble
bicapped
2:1
association
cyclodextrins is well established. γ-Cyclodextrin (CD8) forms water-soluble bicapped 2:1 association
complexes
complexes with
with [60]fullerene
[60]fullerene (C
(C6060))[95–101].
[95–101]. Also
Also β-cyclodextrin
β-cyclodextrin(CD7)
(CD7)may
may undergo
undergo complexation
complexation
with
C
60 [102,103].
with C60 [102,103].
Shape
of C
C60 and
C70 on native γ-cyclodextrin chemically
Shape selectivity
selectivity governs
governs the
the HPLC
HPLC separation
separation of
60 and C70 on native γ-cyclodextrin chemically
bonded
to
silica
[104].
C
70 was much more strongly retained than C60 on this stationary phase whereas
bonded to silica [104]. C70 was much more strongly retained than C60 on this stationary phase whereas
no
of the
the two
two fullerenes
fullerenesoccurred
occurredon
onthe
thecorresponding
corresponding
unmodified
silica
indicating
no separation
separation of
unmodified
silica
indicating
thatthat
the
the
separation
is
due
to
the
selective
supramolecular
interaction
with
the
cyclodextrin
selector
[104].
separation is due to the selective supramolecular interaction with the cyclodextrin selector [104]. InIna
areciprocal
reciprocalstrategy,
strategy,Bianco
Biancoetetal.al.used
usedaaCC60-bonded
-bonded phase
phase (Figure
(Figure 24)
24) in
in an
an aqueous/organic
aqueous/organic medium
medium
60
for
the
high-efficiency
HPLC
separation
of
α-,
βand
γ-cyclodextrins
(CD6-CD8)
[105].
As
expected,
for the high-efficiency HPLC separation of α-, β- and γ-cyclodextrins (CD6-CD8) [105]. As expected,
the
the trace
trace shows
shows increased
increased retention
retention with
with increased
increased molecular
molecular weight
weight of
of CDn
CDn (Figure
(Figure 25).
25).
With
the
advent
of
the
enzymatic
access
to
large-ring
cyclodextrins
[106,107],
separation
With the advent of the enzymatic access to large-ring cyclodextrins [106,107], separationof
of the
the
congeners
CDn
represented
a
considerable
analytical
challenge.
Koizumi
et
al.
separated
CD6-CD85,
congeners CDn represented a considerable analytical challenge. Koizumi et al. separated CD6-CD85,
obtained by the action of cyclodextrin glycosyltransferase from Bacillus macerans, on synthetic
amylose by high-performance anion-exchange chromatography (HPAEC) with pulsed amperometric
Molecules 2016, 21, 1535
22 of 35
obtained by the action of cyclodextrin glycosyltransferase from Bacillus macerans, on synthetic amylose
Molecules
22
Molecules 2016,
2016, 21,
21, 1535
1535
22 of
of 35
35
by high-performance anion-exchange chromatography (HPAEC) with pulsed amperometric detection
using
a 25 cm × 4 amm
I.D. ×Dionex
CarboPac
PA-100 column
(Figure 26) [108],
whereas
Bogdanski et
detection
detection using
using a 25
25 cm
cm × 44 mm
mm I.D.
I.D. Dionex
Dionex CarboPac
CarboPac PA-100
PA-100 column
column (Figure
(Figure 26)
26) [108],
[108], whereas
whereas
al. Bogdanski
separated
CD6-CD21
by
liquid-chromatography
with
electrospray
ionization
mass
spectrometric
Bogdanski et
et al.
al. separated
separated CD6-CD21
CD6-CD21 by
by liquid-chromatography
liquid-chromatography with
with electrospray
electrospray ionization
ionization mass
mass
detection
(LC/ESI-MS)
using
a 25 cm ×using
4 mmaa 25
I.D.
LiChrospher
[109].
With one
exception
spectrometric
detection
(LC/ESI-MS)
×× 44 mm
LiChrospher
22 column
[109].
With
2 column NH
spectrometric
detection
(LC/ESI-MS)
using
25 cm
cm
mm I.D.
I.D. NH
LiChrospher
NH
column
[109].
With
(CD9
on
HPAEC)
the
congeners
CDn
were
eluted
from
the
stationary
phases
according
to
the
degree
one
exception
(CD9
on
HPAEC)
the
congeners
CDn
were
eluted
from
the
stationary
phases
according
one exception (CD9 on HPAEC) the congeners CDn were eluted from the stationary phases according
of polymerization
(Figure
26).
to
the
degree
of
polymerization
(Figure
26).
to the degree of polymerization (Figure 26).
Figure
Schematic
representation
of
the
reciprocal
principle
cyclodextrins.
Figure
23.23.
Schematic
betweenfullerenes
fullerenesand
and
cyclodextrins.
Figure
23.
Schematicrepresentation
representationof
ofthe
thereciprocal
reciprocal principle
principle between
between
fullerenes
and
cyclodextrins.
Reproduced
from
Bogdanski,
Doctoral
Thesis,
University
Reproduced
from
A.A.
Bogdanski,
Tübingen,Tübingen,
Tübingen,Germany,
Germany,2007.
2007.
Reproduced
from
A.
Bogdanski,Doctoral
DoctoralThesis,
Thesis,University
University of Tübingen,
Tübingen,
Tübingen,
Germany,
2007.
Figure
Synthetic
pathway
6,6-[60]fulleropyrrolidine
via
Figure
24.
Synthetic
pathwaytoto
toaaa6,6-[60]fulleropyrrolidine
6,6-[60]fulleropyrrolidine derivative
derivative
by
aziridine
ring
opening
viavia
Figure
24.24.
Synthetic
pathway
derivativeby
byaziridine
aziridinering
ringopening
opening
1,3-dipolar
addition
to
C
60
and
grafting
the
fullerene-containing
triethoxysilane
to
silica
microparticles
1,3-dipolar
addition
C60and
andgrafting
graftingthe
thefullerene-containing
fullerene-containing triethoxysilane
1,3-dipolar
addition
to to
C60
triethoxysilanetotosilica
silicamicroparticles
microparticles
leading
to aa chemically
homogeneous
material
with
defined
and high
chromatographic
leading
chemicallyhomogeneous
homogeneousmaterial
material with
with defined
defined structure
structure
leading
to to
a chemically
structure and
andhigh
highchromatographic
chromatographic
efficiency.
Reprinted
with
permission
from
[105].
Copyright
(1997)
American
Chemical
Society.
efficiency.
Reprinted
with
permission
from
[105].
Copyright
(1997)
American
Chemical
efficiency. Reprinted with permission from [105]. Copyright (1997) American ChemicalSociety.
Society.
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
23 of 35
23 of 35
23 of 35
Figure 25.
25. Separation
Separation of
of α-,
α-, ββ- and
and γ-cyclodextrins
γ-cyclodextrins (CD6,
(CD6, CD7,
CD7, CD8)
CD8) on
on immobilized
immobilized C
C60
60 by HPLC.
Figure
Figure 25. Separation of α-, β- and γ-cyclodextrins (CD6, CD7, CD8) on immobilized C60 by HPLC.
250 ××1.8
stainless
steel
column
packed
withwith
[60]fulleropyrrolidine-based
silica (Hypersil,
5 μm),
250
1.8mm
mmI.D.
I.D.
stainless
steel
column
packed
[60]fulleropyrrolidine-based
silica (Hypersil,
250 × 1.8 mm I.D. stainless steel column packed with [60]fulleropyrrolidine-based silica (Hypersil, 5 μm),
0.25 mL/min
water/methanol/tetrahydrofuran
(80/10/10 for
3 min and
linear
gradient
to
5eluent:
µm), eluent:
0.25 mL/min
water/methanol/tetrahydrofuran
(80/10/10
for then
3 min
and then
linear
eluent: 0.25 mL/min water/methanol/tetrahydrofuran (80/10/10 for 3 min and then linear gradient to
◦
40/30/30
in 40/30/30
25
atinat
25
permission
from[105].
[105].
Copyright
(1997)
American
gradient
to
2525°C.
min,
at 25 C. with
Reprinted
with from
permission
from
[105].
Copyright
(1997)
40/30/30
in min,
25 min,
°C.Reprinted
Reprinted
with
permission
Copyright
(1997)
American
American
Chemical
Chemical
Society.
Chemical
Society.Society.
Figure
26. Separation
cyclodextrincongeners
congeners (CD6–CD85)
obtained
by cyclodextrin
Figure
26. Separation
of of
cyclodextrin
(CD6–CD85)enzymatically
enzymatically
obtained
by cyclodextrin
glycosyltransferase from Bacillus macerans on synthetic amylose by high-performance anion-exchange
glycosyltransferase from Bacillus macerans on synthetic amylose by high-performance anion-exchange
Figure
26. Separation(HPAEC)
of cyclodextrin
congeners
(CD6–CD85)
enzymatically
by cyclodextrin
chromatography
with pulsed
amperometric
detection
using a 25 cmobtained
× 4 mm I.D.
Dionex
chromatography (HPAEC) with pulsed amperometric detection using a 25 cm × 4 mm I.D. Dionex
glycosyltransferase
Bacillus
macerans
syntheticfrom
amylose
CarboPac PA-100from
column.
Reprinted
withon
permission
[108].by high-performance anion-exchange
CarboPac PA-100 column. Reprinted with permission from [108].
chromatography (HPAEC) with pulsed amperometric detection using a 25 cm × 4 mm I.D. Dionex
In striking
contrast,
unusual with
supramolecular
selectivity
CarboPac
PA-100
column.an
Reprinted
permission from
[108].pattern has been observed for the
reciprocal
separation
of an
CD6–CD15
on a fullerene
C60 selector
anchored
to silica
particles
In
striking
contrast,
unusualcongeners
supramolecular
selectivity
pattern
has been
observed
for the
(Figure
27),
which
strongly
deviates
from
the
usually
observed
correlation
of
retention
and
molecular
reciprocal
separation
of CD6–CD15
congeners
on a fullerene
C60 selector
silica particles
In striking
contrast,
an unusual
supramolecular
selectivity
pattern anchored
has been to
observed
for the
weight
[110].
Thus,
whereas
CD6 and
CD7the
exhibit
a lowobserved
retention correlation
and are eluted
~5 min, CD8–CD10
(Figure
27),
which
strongly
deviates
from
usually
ofin
retention
reciprocal
separation
of CD6–CD15
congeners
on a fullerene
C60 selector anchored
to and
silicamolecular
particles
are eluted only after a long retention gap of about 15 min with the elution order CD8 < CD10 < CD9 in
weight
Thus,strongly
whereas deviates
CD6 andfrom
CD7 the
exhibit
a low
retentioncorrelation
and are eluted
in ~5 min,
(Figure[110].
27), which
usually
observed
of retention
andCD8–CD10
molecular
~20 min. The higher CD11-CD15 congeners are eluted together with CD6 in ~5 min (Figure 27). The
are
eluted
only
afterwhereas
a long retention
gap
of
abouta15
min
with the
elution
orderinCD8
< CD10
< CD9
weight
[110].
Thus,
CD6
and
CD7
exhibit
low
retention
and
are
eluted
~5
min,
CD8–CD10
drop of retention between CD10 and CD11 of 15 min is indeed remarkable. The unusual elution order
in
~20
min.
The
higher
CD11-CD15
congeners
are
eluted
together
with
CD6
in
~5
min
(Figure
are eluted
onlydetected
after a long
gap of about
15 min with
the elution(ESI-MS)
order CD8
CD10 <ion
CD927).
in
has been
withretention
LC-electrospray
ionization-mass
spectrometry
of <sodium
The
drop
of retention
between
CD10
and confirmed
CD11
15
is indeed
The
unusual
elution
~20 min.
The
CD11-CD15
congeners
are of
eluted
together
with
CD6 in ~5anmin
(Figure 27).
The
adducts
ofhigher
CD6-CD15.
The results
were
formin
CD6-CD25
by remarkable.
employing
LC-evaporative
order
has
been detected
with
LC-electrospray
ionization-mass
spectrometry
(ESI-MS)
sodium
ion
droplight
of
retention
between
CD10
and
CD11[110].
of 15
min is indeed remarkable.
The
unusualofelution
order
scattering
detection
(ELSD)
system
adducts
CD6-CD15.
TheLC-electrospray
results were confirmed
for CD6-CD25
by employing
an LC-evaporative
has beenof detected
with
ionization-mass
spectrometry
(ESI-MS)
of sodium ion
light
scattering
detection
(ELSD)
system
adducts
of CD6-CD15.
The
results
were [110].
confirmed for CD6-CD25 by employing an LC-evaporative
light scattering detection (ELSD) system [110].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
24 of 35
24 of 35
Figure 27.
traces
of of
thethe
selective
separation
of CD6-CD15
sodium
ion adducts
on C60
Figure
27. Overlaid
OverlaidLC-ESI-MS
LC-ESI-MS
traces
selective
separation
of CD6-CD15
sodium
ion adducts
60
stationary
phase,
0.5
mL/min
on
a
250
×
4
mm
I.D.
stainless
steel
column
packed
with
silica-bonded
C
on C60 on a 250 × 4 mm I.D. stainless steel column packed with silica-bonded C60 stationary phase,
water
to acetonitrile/tetrahydrofuran
gradient elution,
25 °C
[110]. Reproduced
from A. Bogdanski,
0.5
mL/min
water to acetonitrile/tetrahydrofuran
gradient
elution,
25 ◦ C [110]. Reproduced
from A.
Doctoral Thesis,
University
Tübingen,
Bogdanski,
Doctoral
Thesis, of
University
ofTübingen,
Tübingen,Germany,
Tübingen,2007.
Germany, 2007.
The retention-increment method [91,92] has been employed to quantify the unusual selectivity.
The retention-increment method [91,92] has been employed to quantify the unusual
It separates supramolecular interactions from non-specific interactions of the linker. Thus two
selectivity. It separates supramolecular interactions from non-specific interactions of the linker.
stationary phases, i.e., one containing the spacer alone and one containing the C60 selector bonded via
Thus two stationary phases, i.e., one containing the spacer alone and one containing the C60 selector
the spacer were prepared (Figure 28). Nucleosil (5 μm, 100 Å) was reacted in toluene with
bonded via the spacer were prepared (Figure 28). Nucleosil (5 µm, 100 Å) was reacted in toluene with
3-glycidoxypropyltrimethoxysilane to yield 3-glycidoxypropyl-silica and (i) the epoxide was reacted
3-glycidoxypropyltrimethoxysilane to yield 3-glycidoxypropyl-silica and (i) the epoxide was reacted
with bis(1-aminohexyl)malonate to the silica-bonded spacer without selector and (ii) the same epoxide
with bis(1-aminohexyl)malonate to the silica-bonded spacer without selector and (ii) the same epoxide
was reacted with bis(1-aminohexyloxycarbonyl)malonate-dihydro[60]fullerene [111]. The striking
was reacted with bis(1-aminohexyloxycarbonyl)malonate-dihydro[60]fullerene [111]. The striking
selectivity change of CDn occurred only on the C60-selector but not on the silica-bonded spacer devoid
selectivity change of CDn occurred only on the C60 -selector but not on the silica-bonded spacer
of the C60 selector. Due to the necessary use of an LC elution gradient only apparent relative
devoid of the C60 selector. Due to the necessary use of an LC elution gradient only apparent relative
complexation constants Krel (related to CD6) were obtained (Table 5). They nevertheless highlight the
complexation constants Krel (related to CD6) were obtained (Table 5). They nevertheless highlight
remarkable stability differences for the CD6-CD12 congeners in their supramolecular interaction with
the remarkable stability differences for the CD6-CD12 congeners in their supramolecular interaction
C60 [110]. The striking selectivity pattern (Table 5) may be subject to molecular modelling studies in
with C60 [110]. The striking selectivity pattern (Table 5) may be subject to molecular modelling studies
the future. The well-known deviation from the ideal ‘doughnut’ structure of small CDn (n < 10) due
in the future. The well-known deviation from the ideal ‘doughnut’ structure of small CDn (n < 10)
to conformationally strain-induced band flips and kinks in CDn (n > 10) [112] may reveal clues of
due to conformationally strain-induced band flips and kinks in CDn (n > 10) [112] may reveal clues of
molecular recognition pattern.
molecular recognition pattern.
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
25 of 35
25 of 35
25 of 35
Figure 28. Structure of the silica-bonded spacer stationary phase (top) and the silica-bonded C60 stationary
Figure
of thethe
silica-bonded
spacer
stationary
phase (top)
the and
silica-bonded
C60 stationary
Figure 28.
28.Structure
Structure
silica-bonded
spacer
stationary
phaseand
(top)
the silica-bonded
C60
phase (bottom)
[110].ofReproduced
from A.
Bogdanski,
Doctoral
Thesis,
University
of Tübingen,
phase
(bottom)
[110].
Reproduced
from
A.
Bogdanski,
Doctoral
Thesis,
University
of
Tübingen,
stationary phase (bottom) [110]. Reproduced from A. Bogdanski, Doctoral Thesis, University
of
Tübingen, Germany, 2007.
Tübingen, Tübingen,
Germany, 2007.
Germany, 2007.
Table 5. Apparent complexation constants Krel of cyclodextrins and silica-bonded [60]fullerene (C60)
Table
5. Apparent
Apparent complexation
complexation constants
constants K
Krel
rel of cyclodextrins and silica-bonded [60]fullerene (C60
60))
Table 5.
related to CD6 (α-cyclodextrin) [110].
related to CD6 (α-cyclodextrin) [110].
Krel
Krel
CD6
CD6
1
1
Krel
CD7
CD8
CD7
CD8
CD7
CD8
39
4.5
39
CD6
4.5
1
4.5
39
CD9
CD9
46
CD9 46 CD10
46
42
CD10
CD10
CD11
42
42
1
CD11
CD11
1
CD12 1
2
CD12
CD12
2
2
It has been speculated [110] that by using a single [78]fullerene selector (C78), the most strongly
It has been speculated [110] that by using a single [78]fullerene selector (C78), the most strongly
interacting
CD-congener
can
be identified
and,
according
to the reciprocal(Cprinciple,
could
then
It has been
speculatedcan
[110]
by using
a single
[78]fullerene
strongly
78 ), the most
interacting
CD-congener
be that
identified
and,
according
to the selector
reciprocal principle,
could
then
applied
as
a
silica-bonded
resolving
agent
for
the
separation
of
all
five
isomeric
[78]fullerenes
interacting
can be
identified
and,for
according
to the reciprocal
could[78]fullerenes
then applied
applied
as CD-congener
a silica-bonded
resolving
agent
the separation
of all principle,
five isomeric
including
the
direct
enantioseparation
of
chiral
D
3-[78]-fullerene [113].
as a silica-bonded
agent for the
separation
of all five isomeric
including
the directresolving
enantioseparation
of chiral
D3-[78]-fullerene
[113]. [78]fullerenes including the
direct enantioseparation of chiral D3 -[78]-fullerene [113].
6. Chromatography of Calixarenes and the Reciprocal Principle
6. Chromatography of Calixarenes and the Reciprocal Principle
6. Chromatography
of the
Calixarenes
and
Reciprocal
Principle
Gutsche reviewed
application
ofthe
calixarenes
(Figure
29) as reciprocal selectors and selectands
Gutsche reviewed the application of calixarenes (Figure 29) as reciprocal selectors and selectands
[114].Gutsche reviewed the application of calixarenes (Figure 29) as reciprocal selectors and selectands [114].
[114].
Figure 29.
29. The
structure
of calix[4]arene.
Reproduced
with permission
of the
Royal Chemistry
Society of
Figure
of calix[4]arene.
Reproduced
with permission
of the Royal
Society
Figure
29.The
Thestructure
structure
of calix[4]arene.
Reproduced
with permission
of the
Royalof Society
of
Chemistry
from [114].
from
[114]. from
Chemistry
[114].
Following the purification of C60- and C70-fullerenes with calix[8]arene [115,116], a reciprocal
Following the purification of C
-fullerenes with
with calix[8]arene
calix[8]arene [115,116],
[115,116], a reciprocal
C60
60-- and C70
70-fullerenes
separation of calix[4]arene, calix[6]arene and calix[8]arene (Figure 30, top) by microcolumn liquid
calix[6]arene and calix[8]arene
calix[8]arene (Figure 30, top) by microcolumn liquid
separation of calix[4]arene, calix[6]arene
chromatography (μ-LC) on a chemically-bonded C60 silica stationary phase (Figure 21) has been
chromatography (μ-LC) on a chemically-bonded C60 silica stationary phase (Figure 21) has been
reported by Saito et al. [117]. An improved sseparation of the same calixarene congeners by μ-LC on
reported by Saito et al. [117]. An improved sseparation of the same calixarene congeners by μ-LC on
Molecules 2016, 21, 1535
26 of 35
chromatography
Molecules
2016, 21, 1535(µ-LC)
on a chemically-bonded C60 silica stationary phase (Figure 21) has26been
of 35
reported by Saito et al. [117]. An improved sseparation of the same calixarene congeners by µ-LC
a 6,6-[60]fulleropyrrolidine
silica
stationary
phase
(Figure
waslater
laterreported
reportedby
byBianco
Biancoet
et al.
al.
aon6,6-[60]fulleropyrrolidine
silica
stationary
phase
(Figure
24)24)
was
(Figure 30, bottom) [105].
Figure 30. Top: Structures of tert-butyl-calix[n]arenes. Bottom: HPLC trace for the separation of
Figure 30. Top: Structures of tert-butyl-calix[n]arenes. Bottom: HPLC trace for the separation of
tert-butylcalix[n]arenes on a 6,6-[60]fulleropyrrolidine stationary phase (Figure 24). 250 × 1.8 mm I.D.
tert-butylcalix[n]arenes on a 6,6-[60]fulleropyrrolidine stationary phase (Figure 24). 250 × 1.8 mm I.D.
stainless steel column. Eluent: CH2Cl2/2-propanol (99.5/0.5), flow rate 0.3 mL/min; T: 25 °C; UV detection
stainless steel column. Eluent: CH2 Cl2 /2-propanol (99.5/0.5), flow rate 0.3 mL/min; T: 25 ◦ C;
at 280 nm. Reprinted with permission from [105]. Copyright (1997) American Chemical Society.
UV detection at 280 nm. Reprinted with permission from [105]. Copyright (1997) American
Chemical Society.
Achiral chromatographic selectivity has been achieved for various amino acid esters on
silica-bonded calix[4]arene tetraester [118,119]. The enantioseparation of binaphthyl atropisomers on
Achiral chromatographic selectivity has been achieved for various amino acid esters on
the chiral acylcalix[4]arene amino acid derivatives (N-l-alaninoacyl)calix[4]arene and (N-l-valinoacyl)
silica-bonded calix[4]arene tetraester [118,119]. The enantioseparation of binaphthyl atropisomers on
calix[4]arene) by electrokinetic chromatography (EKC) has been reported [120]. By enantioselective
the chiral acylcalix[4]arene amino acid derivatives (N-l-alaninoacyl)calix[4]arene and (N-l-valinoacyl)
liquid chromatography a chiral calixarene functionalized with (−)-ephedrine at the lower rim and
calix[4]arene) by electrokinetic chromatography (EKC) has been reported [120]. By enantioselective
chemically bonded to silica has been used to separate racemic 1-phenyl-2,2,2-trifluoroethanol [121].
liquid chromatography a chiral calixarene functionalized with (−)-ephedrine at the lower rim and
A chiral resorc[4]arene selector [122] of the basket-type containing four ω-unsaturated alkyl chains
chemically bonded to silica has been used to separate racemic 1-phenyl-2,2,2-trifluoroethanol [121].
was synthesized from resorcinol and 1-undecenal [123]. Afterwards chiral L-valine-tert-butylamide
A chiral resorc[4]arene selector [122] of the basket-type containing four ω-unsaturated alkyl chains
moieties were attached to the eight hydroxyl groups. The chiral resorc[4]arene was then linked via
was synthesized from resorcinol and 1-undecenal [123]. Afterwards chiral L-valine-tert-butylamide
the ω-unsaturated alkyl chains to poly(dimethyl-methylhydro)siloxane via platinum-catalyzed
moieties were attached to the eight hydroxyl groups. The chiral resorc[4]arene was then linked
hydrosilylation to yield chemically bonded Chirasil-Calix-Val (Figure 31) [123]. On a 20 m × 0.25 mm
via the ω-unsaturated alkyl chains to poly(dimethyl-methylhydro)siloxane via platinum-catalyzed
I.D. fused silica capillary column coated with Chirasil-Calix (0.25 μm) the N(O,S)-trifluoroacetyl-D,Lhydrosilylation to yield chemically bonded Chirasil-Calix-Val (Figure 31) [123]. On a 20 m × 0.25 mm
α-amino acids Ala, Abu, Val, Ile, Leu, Pro, Phe, Glu, Tyr, Orn and Lys were enantioseparated by GC
I.D. fused silica capillary column coated with Chirasil-Calix (0.25 µm) the N(O,S)-trifluoroacetyl-D,Lwith Rs > 2 [122].
α-amino acids Ala, Abu, Val, Ile, Leu, Pro, Phe, Glu, Tyr, Orn and Lys were enantioseparated by GC
with Rs > 2 [122].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
27 of 35
27 of 35
Synthesis of the CSP Chirasil-Calix. Reprinted
Reprinted with permission from [123].
Figure 31. Synthesis
In
dual chiral
chiral recognition
recognition system
system Chirasil-Calix-Val-Dex,
Chirasil-Calix-Val-Dex, aa resorc[4]arene
resorc[4]arene containing
containing
In the
the dual
L
permethylated β-cyclodextrin
L-valine
-valine diamide
diamide (Calix-Val)
(Calix-Val) and
and permethylated
β-cyclodextrin (CD)
(CD) were
were chemically
chemically linked
linked to
to
poly(dimethylsiloxane)
[124].
The
selector
system
belongs
to
the
category
of
mixed
chiral
stationary
poly(dimethylsiloxane) [124]. The selector system belongs to the category of mixed chiral stationary
phases
CSP
both
N-TFA-valine
ethyl
ester
(due(due
the presence
of Calix-Val)
and
phases [125].
[125]. On
Onthis
thisbinary
binary
CSP
both
N-TFA-valine
ethyl
ester
the presence
of Calix-Val)
chiral
unfunctionalized
1,2-trans-dialkylcyclohexanes
(due(due
to to
thethe
presence
and chiral
unfunctionalized
1,2-trans-dialkylcyclohexanes
presenceofofthe
the CD)
CD) were
were
simultaneously
enantioseparated
(Figure
32)
[124].
In
order
to
account
for
matched
and
mismatched
simultaneously enantioseparated (Figure 32) [124]. In order to account for matched and mismatched
enantioselectivities
imparted by
by the
thetwo
twoselectors,
selectors,the
thechirality
chiralityofofCalix-Val
Calix-Val
could
changed
from
enantioselectivities imparted
could
bebe
changed
from
to
to
L
to
D
.
When
racemic
Calix-LD-Val
were
used,
any
observed
enantioselectivity
could
only
arise
L to D . When racemic Calix-LD-Val were used, any observed enantioselectivity could only arise from
from
the presence
CD [124].
the presence
of CDof[124].
In
example above,
above, the
the chiral
chiral resorc[4]arene
resorc[4]arene is
is used
used as
as an
an enantioselective
enantioselective selector.
selector. In
In the
the example
In aa
reversed
fashion
the
enantiomers
of
a
chiral
resorc[4]arene
compound
can
also
be
studied
as
selectands.
reversed fashion the enantiomers of a chiral resorc[4]arene compound can also be studied as selectands.
Thus,
tetrabenzoxacine resorc[4]arene
CSP
Thus, tetrabenzoxacine
resorc[4]arene (Figure
(Figure 33,
33, top)
top) was
was enantioseparated
enantioseparated by
by HPLC
HPLC on
on the
the CSP
Chiralpak
Okamoto et
et al.
al. [126].
[126]. The
The chromatograms
chromatograms show
show temperature-dependent
temperature-dependent
Chiralpak AD
AD developed
developed by
by Okamoto
interconversion
profiles
due
to
enantiomerization
during
enantioseparation
(Figure
interconversion profiles due to enantiomerization during enantioseparation (Figure 33,
33, bottom)
bottom)[127].
[127].
By
peak-form
analysis
featuring
plateau
formation
[128],
the
inversion
barrier
was
determined
By peak-form analysis featuring plateau formation [128], the inversion barrier was determined to
to
ΔG
=
92
±
2
kJ/mol
(298K)
[127].
∆G = 92 ± 2 kJ/mol (298K) [127].
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
Molecules 2016, 21, 1535
28 of 35
28 of 35
28 of 35
Figure 32.
32. The
The mixed
mixed CSP
CSP Chirasil-Calix-Val-Dex.
Chirasil-Calix-Val-Dex.
Figure
Figure 32. The mixed CSP Chirasil-Calix-Val-Dex.
Figure 33. Top: Interconverting enantiomers of tetrabenzoxacine resorc[4]arene by HPLC on the CSP
Figure
Top:
Interconverting
enantiomers
tetrabenzoxacine
resorc[4]arene
HPLC
CSP
Figure
33.33.
Top:
Interconverting
enantiomers
of of
tetrabenzoxacine
resorc[4]arene
byby
HPLC
onon
thethe
CSP
Chiralpak
AD.
Bottom:
Distorted
peak profiles
caused
by enantiomerization
(dotted
line:
expected
Chiralpak
AD.
Bottom:
Distorted
peak
profiles
caused
by
enantiomerization
(dotted
line:
expected
Chiralpak
AD.
Bottom:
Distorted
peak
profiles
caused
enantiomerization
(dotted
line:
expected
peak
profiles
devoid
of interconversion).
Reprinted
withbypermission
from [127].
peak
profiles
devoid
of interconversion).
Reprinted
with
permission
from
[127].
peak
profiles
devoid
of interconversion).
Reprinted
with
permission
from
[127].
This finding is recalled here, because one may speculate to employ interconverting calixarene
This finding
is recalled
here, abecause
one may
speculate
employofinterconverting
calixarene
enantiomers
in a reciprocal
fashion
dynamic
the
spirit to
of ato
scenario
column deracemization
This finding
is recalled
here,as
because
oneCSP
mayinspeculate
employ
interconverting
calixarene
enantiomers
in a reciprocal
fashion
as aan
dynamic
CSP in the spirit
of a scenario
of column
deracemization
described
byinWelch
[129].fashion
Thus
when
enantiomerically
pure
which
interacts
with high
enantiomers
a reciprocal
as a dynamic
CSP in the spirit
of selectand,
a scenario of
column
deracemization
described
by
Welch
[129].
Thus
when
an
enantiomerically
pure
selectand,
which
interacts
with
high
affinity
and
enantioselectivity
with
an
interconverting
selector
present
in
the
stationary
phase,
is
added
described by Welch [129]. Thus when an enantiomerically pure selectand, which interacts with
high
affinity
and
enantioselectivity
with
an
interconverting
selector
present
in
the
stationary
phase,
is
added
to the mobile
phase, it may cause
of the
selector,
thereby
creating
a nonracemic
for
affinity
and enantioselectivity
withderacemization
an interconverting
selector
present
in the
stationary
phase, isCSP
added
to the mobile
phase, it may
cause deracemization
of the
selector,
thereby This
creating
a nonracemic
CSP
for
subsequent
enantiomeric
separations
of the selectand
in the
its
racemic
intriguing
was
to
the mobile
phase, it may
cause deracemization
of
selector,form.
thereby creating
a scenario
nonracemic
subsequent
enantiomeric
separations
of
the
selectand
in
its
racemic
form.
This
intriguing
scenario
was
demonstrated
by HPLC
with a conformationally
naphthamide
atropisomer
it was
CSP
for subsequent
enantiomeric
separations of labile
the selectand
in its racemic
form.[129].
ThisFirst
intriguing
demonstrated
by
HPLC
with
a
conformationally
labile
naphthamide
atropisomer
[129].
First
it was
shown that
naphthamide
can bewith
enantioseparated
on the
Whelk-O1
CSP (Figure
3c). Under
scenario
was the
demonstrated
by HPLC
a conformationally
labile
naphthamide
atropisomer
[129].
shown that conditions
the naphthamide
can be enantioseparated
CSPWith
(Figure
Under
stopped-flow
[130], deracemization
took placeon
in the
the Whelk-O1
mobile phase.
the 3c).
reciprocal
stopped-flow
conditions
[130],
deracemization
took
place
in
the
mobile
phase.
With
the
reciprocal
concept the racemic naphthamide was then linked to silica whereas the soluble analog of the
concept the racemic naphthamide was then linked to silica whereas the soluble analog of the
Molecules 2016, 21, 1535
29 of 35
First it was shown that the naphthamide can be enantioseparated on the Whelk-O1 CSP (Figure 3c).
Under stopped-flow conditions [130], deracemization took place in the mobile phase. With the
reciprocal concept the racemic naphthamide was then linked to silica whereas the soluble analog of the
enantiomerically pure Whelk-O1 selector was added to the mobile phase. After deracemization of the
CSP, the racemic analog of the Whelk-O1 selector was enantioseparated, however, with the expected
subsequent decrease of the separation factor α due to gradual racemization of the CSP [129].
The reciprocal concept of interconverting molecules in separation methods could also be
extended to the liquid chromatographic investigation of conformational cis/trans-diastereomers of
L-peptidyl- L-proline. Studies of proline-containing peptides separated on calixarene-bonded silica gels
have been investigated by Gebauer et al. [131] and interconversion barriers of L-proline containing
dipeptides by dynamic CE were determined [132].
7. Conclusions
Since the first reports by Mikeš and Pirkle on the use of a reciprocal approach in chromatography
through changing the role of selectands and selectors, many examples have been reported in different
areas of selective supramolecular interactions including combinatorial advances. Advantages and
limitations of the reciprocal concept are outlined. The reciprocal strategy clearly aids the development of
optimized selectors for difficult separations of selectands including enantiomers. Various scenarios of
the interchanging role of selectors and selectands in separation science are discussed in the hope that
this will stir further endeavors devoted to the principle of reciprocity in supramolecular recognition.
The present account should be seen as part of the emerging field of supramolecular separation
methods [1,2] with further emphasis also to molecular sensor approaches and surface related
recognition phenomena. The optimization of selector/selectand systems is the precondition for
meaningful mechanistic studies by various molecular modelling approaches.
Acknowledgments: This work was partially funded by the German Research Council (DFG Schu 395/18-2) in 2005
on ‘Combinatorial Chiral Selectors for the Enantiomeric Separation by Chromatography and Electromigration’.
Conflicts of Interest: The author declares no conflict of interest.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Schurig, V. Supramolecular Chromatography. In Applications of Supramolecular Chemistry; Schneider, H.-J.,
Ed.; CRC-Press: Boca Raton, FL, USA, 2012; pp. 129–157.
Ciogli, A.; Kotoni, D.; Gasparrini, F.; Pierini, M.; Villani, C. Chiral Supramolecular Selectors for Enantiomer
Differentiation in Liquid Chromatography. In Differentiation of Enantiomers I; Schurig, V., Ed.; Springer:
Heidelberg, Germany, 2013; Volume 340, pp. 73–106.
Allenmark, S.; Schurig, V. Chromatography on chiral stationary phases. J. Mater. Chem. 1997, 7, 1955–1963.
[CrossRef]
Mikeš, F. The Resolution of Chiral Compounds by Modern Liquid Chromatography. Ph.D. Thesis,
The Weizmann Institute of Science, Rehovot, Israel, 1975.
Ashby, W.R. An Introduction to Cybernetics; Chapman & Hall Ltd.: London, UK, 1956; p. 10.
Allenmark, S. Separation of Enantiomers by Protein Based Chiral Phases. In A Practical Approach to Chiral
Separations by Liquid Chromatography; Subramanian, G., Ed.; Wiley-VCH: Weinheim, Germany, 1994; p. 183.
Okamoto, Y.; Yashima, E. Polysaccharide derivatives for chromatographic separation of enantiomers.
Angew. Chem. Int. Ed. 1998, 37, 1020–1043. [CrossRef]
Zeleke, Y.M.; Smith, G.B.; Hofstetter, H.; Hofstetter, O. Enantiomer separation of Amino Acids in
immunoaffinity Micro LC-MS. Chirality 2006, 18, 544–550. [CrossRef] [PubMed]
Mikeš, F.; Boshart, G.; Gil-Av, E. Helicenes, Resolution on chiral charge-transfer complexing agents using
high performance liquid chromatography. J. Chem. Soc. Chem. Commun. 1976, 99–100. [CrossRef]
Mikeš, F.; Boshart, G.; Gil-Av, E. Resolution of optical isomers by high-performance liquid chromatography,
using coated and bonded chiral charge-transfer complexing agents as stationary phases. J. Chromatogr. 1976,
122, 205–221. [CrossRef]
Molecules 2016, 21, 1535
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
30 of 35
Mikeš, F.; Boshart, G. Resolution of optical isomers by high-performance liquid chromatography—A
comparison of two selector systems. J. Chromatogr. 1978, 149, 455–464. [CrossRef]
Kim, Y.H.; Balan, A.; Tishbee, A.; Gil-Av, E. Chiral differentiation by the P-(+)-Hexahelicene-7,70 -dicarboxylic
acid disodium salt. Resolution of N-2,4-Dinitrophenyl-α-amino-acid esters by high performance liquid
chromatography. J. Chem. Soc. Chem. Commun. 1982, 1336–1337. [CrossRef]
Pirkle, W.H. The nonequivalence of physical properties of enantiomers in optically active solvents.
Differences in nuclear magnetic resonance spectra. I. J. Am. Chem. Soc. 1966, 88, 1837. [CrossRef]
Pirkle, W.H.; Burlingame, T.G.; Beare, S.D. Optically Active NMR Solvents VI. The determination of optical
purity and absolute configuration of amines. Tetrahedron Lett. 1968, 9, 5849–5852. [CrossRef]
Pirkle, W.H.; Hoover, D.J. NMR chiral solvating agents. Top. Stereochem. 1982, 13, 263–331.
Pirkle, W.H.; House, D.W.; Finn, J.M. Broad spectrum resolution of optical isomers using chiral
high-performance liquid chromatographic bonded phases. J. Chromatogr. 1980, 192, 143–158. [CrossRef]
Pirkle, W.H.; Däppen, R. Reciprocity in chiral recognition: Comparison of several chiral stationary phases.
J. Chromatogr. A 1987, 404, 107–115. [CrossRef]
Pirkle, W.H.; Hyun, M.H.; Bank, B. A rational approach to the design of highly-effective chiral stationary
phases. J. Chromatogr. 1984, 316, 585–604. [CrossRef]
Pirkle, W.H.; Pochapsky, T.C.; Burke, J.A., III; Deming, K.C. Systematic Studies of Chiral Recognition
Mechanisms. In Chiral Separations; Stevenson, D., Wilson, I.D., Eds.; Plenum Press: New York, NY, USA,
1988; pp. 23–35.
Pirkle, W.H.; Pochapsky, T.C. Considerations of chiral recognition relevant to the liquid chromatographic
separation of enantiomers. Chem. Rev. 1989, 89, 347–362. [CrossRef]
Pirkle, W.H.; Readnour, R.S. Chromatographic approach to the measurement of the interstrand distance for
some chiral bonded phases. Anal. Chem. 1991, 63, 16–20. [CrossRef]
Pirkle, W.H.; Welch, C.J.; Lamm, B. Design, synthesis, and evaluation of an improved enantioselective
naproxen selector. J. Org. Chem. 1992, 57, 3854–3860. [CrossRef]
Welch, C.J. Evolution of chiral stationary phase design in the Pirkle laboratories. J. Chromatogr. A 1994, 666,
3–26. [CrossRef]
Taylor, D.R.; Clift, E. The Use of Reciprocal Interaction in the Design of Pyrethroid Specific Chiral Stationary
Phases. In Recent Advances in Chiral Separations; Stevenson, D., Wilson, I.D., Eds.; Plenum Press: New York,
NY, USA, 1990; pp. 39–41.
Terabe, S. Electrokinetic chromatography: An interface between electrophoresis and chromatography.
Trends Anal. Chem. 1984, 8, 129–134. [CrossRef]
Chankvetadze, B. Principles of Enantiomer Separation in Electrokinetic Chromatography. In Electrokinetic
Chromatography—Theory, Instrumentation and Applications; Pyell, U., Ed.; John Wiley & Sons: Chichester, UK,
2006; pp. 179–206.
Jakubetz, H.; Juza, M.; Schurig, V. Electrokinetic Chromatography employing an anionic and a cationic
β-cyclodextrin derivative. Electrophoresis 1997, 18, 897–904. [CrossRef] [PubMed]
Terabe, S.; Ozaki, H.; Otsuka, K.; Ando, T. Electrokinetic chromatography with 2-O-Carboxymethyl-βcyclodextrin as a moving “Stationary” Phase. J. Chromatogr. 1985, 332, 211–217. [CrossRef]
Mayer, S.; Schleimer, M.; Schurig, V. Dual chiral recognition system involving cyclodextrin derivatives in
capillary electrophoresis. J. Microcolumn Sep. 1994, 6, 43–48. [CrossRef]
Tait, R.J.; Thompson, D.O.; Stella, V.J.; Stobaugh, J.F. Sulfobutyl ether β-cyclodextrin as a chiral discriminator
for use with capillary electrophoresis. Anal. Chem. 1994, 66, 4013–4018. [CrossRef]
Chankvetadze, B.; Endresz, G.; Blaschke, G. About some aspects of the use of charged cyclodextrins for
capillary electrophoresis enantioseparation. Electrophoresis 1994, 15, 804–807. [CrossRef] [PubMed]
Chankvetadze, B.; Blaschke, G. Enantioseparations in capillary electromigration techniques: Recent
developments and future trends. J. Chromatogr. A 2001, 906, 309–363. [CrossRef]
Guttman, A.; Paulus, A.; Cohen, A.S.; Grinberg, N.; Karger, B.L. Use of complexing agents for selective
separation in high-performance capillary electrophoresis: Chiral resolution via cyclodextrins incorporated
within polyacrylamide gel columns. J. Chromatogr. 1988, 448, 41–53. [CrossRef]
Wistuba, D.; Bogdanski, A.; Larsen, K.L.; Schurig, V. δ-Cyclodextrin as novel chiral probe for enantiomeric
separation by electromigration methods. Electrophoresis 2006, 27, 4359–4363. [CrossRef] [PubMed]
Molecules 2016, 21, 1535
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
31 of 35
Mayer, S.; Schurig, V. Enantiomer separation by electrochromatography on capillaries coated with
Chirasil-Dex. J. Sep. Sci. 1992, 15, 129–130.
Mayer, S.; Schurig, V. Enantiomer separation by electrochromatography on open tubular columns coated
with Chirasil-Dex. J. Liq. Chromatogr. 1993, 16, 915–931. [CrossRef]
Trapp, O.; Trapp, G.; Schurig, V. University of Tübingen, Tübingen, Germany. Unpublished work. 2003.
Li, T. Peptide and peptidomimetic chiral selectors in liquid chromatography. J. Sep. Sci. 2005, 28, 1927–1931.
[CrossRef] [PubMed]
Li, T. Development of efficient chiral stationary phases from peptide libraries. LabPlus Int. 2005, 19, 15–18.
Svec, F.; Wulff, D.; Fréchet, J.M. Combinatorial Approaches to Recognition of Chirality: Preparation and Use
of Materials for the Separation of Enantiomers. In Chiral Separation Techniques, 2nd ed.; Subramanian, G., Ed.;
Wiley-VCH: Weinheim, Germany, 2000; pp. 57–94.
Huang, P.Y.; Carbonell, R.G. Affinity purification of proteins using ligands derived from peptide libraries.
Biotechnol. Bioeng. 1995, 47, 288–297. [CrossRef] [PubMed]
Kaufman, D.B.; Hentsch, M.E.; Baumbach, G.A.; Buettner, J.A.; Dadd, C.A.; Huang, P.Y.; Hammond, D.J.;
Carbonell, R.G. Affinity purification of fibrinogen using a ligand from a peptide library. Biotechnol. Bioeng.
2002, 77, 278–289. [CrossRef] [PubMed]
Huang, P.Y.; Carbonell, R.G. Affinity chromatographic screening of soluble combinatorial peptide libraries.
Biotechnol. Bioeng. 1999, 63, 633–641. [CrossRef]
Jung, G.; Hofstetter, H.; Feiertag, S.; Stoll, D.; Hofstetter, O.; Wiesmüller, K.-H.; Schurig, V. Cyclopeptide
libraries as new chiral selectors in capillary electrophoresis. Angew. Chem. Int. Ed. 1996, 35, 2148–2150.
[CrossRef]
Chiari, M.; Desperati, V.; Manera, E.; Longhi, R. Combinatorial synthesis of highly selective
cyclohexapeptides for separation of amino acid enantiomers by capillary electrophoresis. Anal. Chem.
1998, 70, 4967–4973. [CrossRef] [PubMed]
Bluhm, L.; Huang, J.M.; Li, T.Y. Recent advances in peptide chiral selectors for electrophoresis and liquid
chromatography. Anal. Bioanal. Chem. 2005, 382, 592–598. [CrossRef] [PubMed]
De Lorenzi, E.; Massolini, G.; Molinari, P.; Galbusera, C.; Longhi, R.; Marinzi, C.; Consonni, R.;
Chiari, M. Chiral capillary electrophoresis and nuclear magnetic resonance investigation on the
structure-enantioselectivity relationship in synthetic cyclopeptides as chiral selectors. Electrophoresis 2001, 22,
1373–1384. [CrossRef]
Marinzi, C.; Longhi, R.; Chiari, M.; Consonni, R. Capillary electrophoresis investigation on the
structure-enantioselectivity relationship in synthetic cyclopeptides as chiral selectors. Electrophoresis 2001, 22,
3257–3262. [CrossRef]
Schurig, V.; Wistuba, D.; Jung, G.; Wiesmüller, K.-H. University of Tübingen, Tübingen, Germany.
Unpublished work. 2010.
Wu, Y.; Wang, Y.; Yang, A.; Li, T. Screening of mixture combinatorial libraries for chiral selectors: A reciprocal
chromatographic approach using enantiomeric libraries. Anal. Chem. 1999, 71, 1688–1691. [CrossRef]
[PubMed]
Weingarten, M.D.; Sekanina, K.; Still, W.C. Enantioselective resolving resins from a combinatorial library.
Kinetic resolution of cyclic amino acid derivatives. J. Am. Chem. Soc. 1998, 120, 9112–9113. [CrossRef]
Maier, N.; Franco, P.; Lindner, W. Separation of enantiomer: Needs, challenges, perspectives. J. Chromatogr. A
2001, 906, 3–33. [CrossRef]
Murer, P.; Lewandowski, K.; Svec, F.; Fréchet, J.M.J. Combinatorial “Library on Bead” approach to polymeric
materials with vastly enhanced chiral recognition. J. Chem. Soc. Chem. Commun. 1998, 2559–2560. [CrossRef]
Murer, P.; Lewandowski, K.; Svec, F.; Fréchet, J.M.J. “On Bead” combinatorial approach to the design of
chiral stationary phases. Anal. Chem. 1999, 71, 1278–1284. [CrossRef] [PubMed]
Lewandowski, K.; Murer, P.; Svec, F.; Fréchet, J.M.J. Highly selective chiral recognition on polymer supports:
Preparation of a combinatorial library of dihydropyrimidines and its screening for novel Chiral HPLC
ligands. J. Chem. Soc. Chem. Commun. 1998, 2237–2238. [CrossRef]
Lewandowski, K.; Murer, P.; Svec, F.; Fréchet, J.M.J. A combinatorial approach to recognition of chirality:
The preparation of highly enantioselective aryldihydropyrimidine selectors for chiral HPLC. J. Comb. Chem.
1999, 1, 105–112. [CrossRef] [PubMed]
Molecules 2016, 21, 1535
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
32 of 35
Brahmachary, E.; Ling, F.H.; Svec, F.; Fréchet, J.M.J. Chiral recognition: Design and preparation of
chiral stationary phases using selectors derived from Ugi multicomponent condensation reactions and
a combinatorial approach. J. Comb. Chem. 2003, 5, 441–450. [CrossRef] [PubMed]
Pirkle, W.H.; Alessi, D.M.; Hyun, M.H.; Pochapsky, T.C. Separation of some enantiomeric dipeptides and
tripeptides on chiral stationary phases. J. Chromatogr. 1987, 398, 203–209. [CrossRef]
Welch, C.J.; Bhat, G.; Protopopova, M.N. Silica-based solid phase synthesis of chiral stationary phases.
Enantiomer 1988, 3, 463–469.
Welch, C.J.; Protopopova, M.N.; Bhat, G.A. Microscale synthesis and screening of chiral stationary phases.
Enantiomer 1988, 3, 471–476.
Welch, C.J.; Bhat, G.; Protopopova, M.N. Selection of an optimized adsorbent for preparative
chromatographic enantioseparation by microscale screening of a second generation chiral phase library.
J. Comb. Chem. 1999, 1, 364–367. [CrossRef] [PubMed]
Welch, C.J.; Pollard, S.D.; Mathre, D.J.; Reider, P.J. Improved method for rapid evaluation of chiral stationary
phase libraries. Org. Lett. 2001, 3, 95–98. [CrossRef] [PubMed]
Wang, Y.; Li, T. Screening of a parallel combinatorial library for selectors for chiral chromatography.
Anal. Chem. 1999, 71, 4178–4182. [CrossRef] [PubMed]
Bluhm, L.H.; Wang, Y.; Li, T. An alternative procedure to screen mixture combinatorial libraries for selectors
for chiral chromatography. Anal. Chem. 2000, 72, 5201–5205. [CrossRef] [PubMed]
Wang, Y.; Bluhm, L.H.; Li, T. Identification of chiral selectors from a 200-member parallel combinatorial
library. Anal. Chem. 2000, 72, 5459–5465. [CrossRef] [PubMed]
Blodgett, J.; Wang, Y.; Li, T.; Polavarapu, P.L.; Drabowicz, J.; Pietrusiewicz, K.M.; Zygo, K. Resolution of
tert-Butyl-1-(2-methylnaphthyl)phosphine oxide using selectors identified from a chemical combinatorial
library. Anal. Chem. 2002, 74, 5212–5216. [CrossRef] [PubMed]
Huang, J.; Li, T. Efficient resolution of racemic 1,10 -Bi-2-naphthol with chiral selectors identified from a small
library. J. Chromatogr. A 2005, 1062, 87–93. [CrossRef] [PubMed]
Chen, Z.; Yang, Y.; Werner, S.; Wipf, P.; Weber, S.G. A screening method for chiral selectors that does not
require covalent attachment. J. Am. Chem. Soc. 2006, 22, 2208–2209. [CrossRef] [PubMed]
Liu, H.; Tanaka, T.; Kataura, H. Optical isomer separation of single-chirality carbon nanotubes using gel
column chromatography. Nano Lett. 2014, 14, 6237–6243. [CrossRef] [PubMed]
Zhang, M.; Khripin, C.Y.; Fagan, J.A.; McPhie, P.; Ito, Y.; Zheng, M. Single-step total fractionation of
single-wall carbon nanotubes by countercurrent chromatography. Anal. Chem. 2014, 86, 3980–3984.
[CrossRef] [PubMed]
Sicoli, G.; Pertici, F.; Jiang, Z.; Jicsinszky, L.; Schurig, V. Gas-chromatographic approach to probe the
absence of molecular inclusion in enantioseparations by carbohydrates. Investigation of linear dextrins
(“acyclodextrins”) as novel chiral stationary phases. Chirality 2007, 19, 391–400. [CrossRef] [PubMed]
Raffaini, G.; Ganazzoli, F. Separation of chiral nanotubes with an opposite handedness by chiral oligopeptide
adsorption: A molecular dynamics study. J. Chromatogr. A 2015, 1425, 221–230. [CrossRef] [PubMed]
Ahmed, M.; Yajadda, M.M.A.; Han, Z.J.; Su, D.; Wang, G.; Ostrikov, K.K.; Ghanem, A. Single-walled carbon
nanotube-based polymer monoliths for the enantioselective Nano-liquid chromatographic separation of
racemic pharmaceuticals. J. Chromatogr. A 2014, 1360, 100–109. [CrossRef] [PubMed]
Alhassan, H.; Antony, V.; Ghanem, A.; Yajadda, M.M.A.; Han, Z.J.; Ostrikov, K.K. Organic/Hybrid
nanoparticles and single-walled carbon nanotubes: Preparation methods and chiral applications. Chirality
2014, 26, 683–691. [CrossRef] [PubMed]
Silva, R.A.; Talio, M.C.; Luconi, M.O.; Fernandez, L.P. Evaluation of carbon nanotubes as chiral selectors for
continuous-flow enantiomeric separation of carvedilol with fluorescent detection. J. Pharm. Biomed. Anal.
2012, 70, 631–635. [CrossRef] [PubMed]
Peng, X.; Komatsu, N.; Bhattacharya, S.; Shimawaki, T.; Aonuma, S.; Kimura, T. Optically active single-walled
carbon nanotubes. Nat. Nanotechnol. 2007, 2, 361–365. [CrossRef] [PubMed]
Zhao, L.; Ai, P.; Duan, A.-H.; Yuan, L.-M. Single-Walled carbon nanotubes for improved enantioseparations
on a chiral ionic liquid stationary phase in GC. Anal. Bioanal. Chem. 2011, 399, 143–147. [CrossRef] [PubMed]
Jinno, K. Molecular Recognition for Fullerenes in Liquid Chromatography. In Chromatographic Separations
Based on Molecular Recognition; Jinno, K., Ed.; Wiley-VCH: New York, NY, USA, 1997; pp. 147–237.
Molecules 2016, 21, 1535
33 of 35
Baena, J.R.; Gallego, M.; Valcárcel, M. Fullerenes in the analytical sciences. Trends Anal. Chem. TRAC 2002, 21,
187–198. [CrossRef]
80. Hawkins, J.M.; Lewis, T.A.; Loren, S.D.; Meyer, A.; Heath, J.R.; Shibato, Y.; Saykally, R.J. Organic chemistry of
C60 (Buckminsterfullerene): Chromatography and osmylation. J. Org. Chem. 1990, 55, 6250–6252. [CrossRef]
[PubMed]
81. Pirkle, W.H.; Welch, C.J. An unusual effect of temperature on the chromatographic behavior of
buckminsterfullerene. J. Org. Chem. 1991, 56, 6973–6974. [CrossRef]
82. Welch, C.J.; Pirkle, W.H. Progress in the design of selectors for buckminsterfullerene. J. Chromatogr. 1992, 609,
89–101. [CrossRef]
83. Kawauchi, T.; Kitaura, A.; Kawauchi, M.; Takeichi, T.; Kumaki, J.; Lida, H.; Yashima, E. Separation of C-70
over C-60 and selective extraction and resolution of higher fullerenes by syndiotactic helical poly(methyl
methacrylate). J. Amer. Chem. Soc. 2010, 132, 12191–12193. [CrossRef] [PubMed]
84. Diack, M.; Compton, R.N.; Guiochon, G. Evaluation of stationary phases for the separation of
buckminsterfullerenes by high-performance liquid chromatography. J. Chromatogr. 1993, 639, 129–140.
[CrossRef]
85. Gross, B.; Schurig, V.; Lamparth, I.; Hirsch, A. Enantiomer separation of [60]Fullerene derivatives by
Micro-column-HPLC using (R)-(−)-TAPA as chiral stationary phase. J. Chromatogr. A 1997, 791, 65–69.
[CrossRef]
86. Gross, B.; Schurig, V.; Lamparth, I.; Herzog, A.; Djojo, F.; Hirsch, A. Enantiomeric separation of [60]Fullerene
derivatives with an inherent chiral addition pattern. J. Chem. Soc. Chem. Commun. 1997, 1117–1118. [CrossRef]
87. Saito, Y.; Ohta, H.; Terasaki, H.; Katoh, Y.; Nagashima, H.; Jinno, K. Separation of polycyclic aromatic
hydrocarbons with a C60 bonded silica phase in microcolumn liquid chromatography. J. Sep. Sci. 1995, 18,
569–572. [CrossRef]
88. Stalling, D.L.; Guo, C.; Kuo, K.C.; Saim, S. Fullerene-linked particles as LC chromatographic media and
modification of their donor/acceptor properties by secondary chemical reactions. J. Microcolumn Sep. 1993, 5,
223–235. [CrossRef]
89. Stalling, D.L.; Guo, C.Y.; Saim, S. Surface-linked C60/70 -Polystyrene Divinylbenzene beads as a new
chromatographic material for enrichment of Co-planar PCBs. J. Chromatogr. Sci. 1993, 31, 265–278. [CrossRef]
90. Golovnya, R.V.; Terenina, M.B.; Ruchkina, E.L.; Karnatsevich, V.L. Fullerene c60 as a new stationary phase in
capillary gas chromatography. Mendeleev Commun. 1993, 3, 231–233. [CrossRef]
91. Schurig, V. Molecular Association in Complexation Gas Chromatography. In Chromatographic Separations
Based on Molecular Recognition; Jinno, K., Ed.; Wiley-VCH: New York, NY, USA, 1997; pp. 371–418.
92. Schurig, V. Elaborate treatment of retention in chemoselective chromatography—The retention increment
approach and non-linear effects. J. Chromatogr. A 2009, 1216, 1723–1736. [CrossRef] [PubMed]
93. Glausch, A.; Hirsch, A.; Lamparth, I.; Schurig, V. Retention Behaviour of polychlorinated biphenyls on
Polysiloxane-Anchored C60 in gas chromatography. J. Chromatogr. A 1998, 809, 252–257. [CrossRef]
94. Gross, B. Enantiomerentrennung chiraler Fullerenaddukte mittels LC-Bestimmung von Retentionsinkrementen
an Cyclodextrin-, Calixaren- und Fulleren-Stationärphasen. Ph.D. Thesis, University of Tübingen, Tübingen,
Germany, 1998.
95. Andersson, T.; Nilsson, K.; Sundahl, M.; Westman, G.; Wennerström, O. C-60 embedded in γ-cyclodextrin:
A water-soluble fullerene. J. Chem. Soc. Chem. Commun. 1992, 604–606. [CrossRef]
96. Sundahl, M.; Andersson, T.; Nilsson, K.; Wennerström, O.; Westman, G. Clusters of C60 -fullerene in a water
solution containing γ-cyclodextrin—A photophysical study. Synth. Metals 1993, 55–57, 3252–3257. [CrossRef]
97. Priyadarsini, K.I.; Mohan, H.; Tyagi, A.K.; Mittal, J.P. Inclusion complex of γ-cyclodextrin-C-60-formation,
characterization, and photophysical properties in aqueous solutions. J. Phys. Chem. 1994, 98, 4756–4759.
[CrossRef]
98. Kanazawa, K.; Nakanishi, H.; Ishizuka, Y.; Nakamura, T.; Matsumoto, M. An NMR Study of the Buckminster
Fullerene complex with Cyclodextrin in aqueous solution. Fuller. Sci. Technol. 1994, 2, 189–194. [CrossRef]
99. Yoshida, Z.; Takekuma, H.; Takekuma, S.-I.; Matsubara, Y. Molecular recognition of C60 with γ-Cyclodextrin.
Angew. Chem. Int. Ed. Engl. 1994, 33, 1597–1599. [CrossRef]
100. Priyadarsini, K.I.; Mohan, H.; Mittal, J.P. Characterization and properties of γ-Cyclodextrin-C60 Complex in
aqueous solution. Fuller. Sci. Technol. 1995, 479–493. [CrossRef]
79.
Molecules 2016, 21, 1535
34 of 35
101. Wang, H.M.; Wenz, G. Molecular solubilization of Fullerene C60 in water by γ-Cyclodextrin thioethers.
Beilstein. J. Org. Chem. 2012, 8, 1644–1651. [CrossRef] [PubMed]
102. Samal, S.; Geckeler, K.E. Cyclodextrin-fullerenes: A new class of water-soluble fullerenes. J. Chem. Soc., Chem.
Commun. 2000, 1101–1102. [CrossRef]
103. Yu, Y.; Shi, Z.; Zhao, Y.; Ma, Y.; Xue, M.; Ge, J. Highly Water-Soluble [60]Fullerene-ethylenediamino-betacyclodextrin inclusion complex: The synthesis and self-assembly with poly (acrylic acid). J. Supramol. Chem.
2008, 20, 295–299. [CrossRef]
104. Cabrera, K.; Wieland, G.; Schäfer, M. High-performance liquid chromatographic separation of fullerenes
(C-60 and C-70) using chemically bonded gamma-cyclodextrin as stationary phase. J. Chromatogr. A 1993,
644, 396–399. [CrossRef]
105. Bianco, A.; Gasparrini, F.; Maggini, M.; Misiti, D.; Polese, A.; Prato, M.; Scorrano, G.; Toniolo, C.; Villani, C.
Molecular recognition by a Silica-bound Fullerene derivative. J. Amer. Chem. Soc. 1997, 119, 7550–7554.
[CrossRef]
106. Larsen, K.L. Large Cyclodextrins. J. Inclusion Phenom. Macrocycl. Chem. 2002, 43, 1–13. [CrossRef]
107. Ueda, H.; Endo, T. Large-Ring Cyclodextrins. In Cyclodextrin and Their Complexes: Chemistry, Analytical
Methods, Applications; Dodziuk, H., Ed.; Wiley-VCH: Weinheim, Germany, 2008; pp. 370–380.
108. Koizumi, K.; Sanbe, K.; Kubota, Y.; Terada, Y.; Takaha, T. Isolation and Characterization of cyclic alpha-(1
->4)-Glucans having degrees of polymerization 9–31 and their quantitative analysis by high-performance
anion-exchange chromatography with pulsed amperometric detection. J. Chromatogr. A 1999, 852, 407–416.
[CrossRef]
109. Bogdanski, A.; Larsen, K.L.; Bischoff, D.; Ruderisch, A.; Jung, G.; Süßmuth, R.; Schurig, V. Proceed. In
Proceedings of the 12th International Cyclodextrin. Symposium, Montpellier, France, 16–19 May 2004; p. 171.
110. Bogdanski, A.; Wistuba, D.; Larsen, K.L.; Hartnagel, U.; Hirsch, A.; Schurig, V. Reciprocal principle of
molecular recognition in supramolecular chromatography—Highly selective analytical separations of
cyclodextrin congeners on a Silica-bonded [60]Fullerene stationary phase. New J. Chem. 2010, 34, 693–698.
[CrossRef]
111. Braun, M.; Hartnagel, U.; Ravanelli, E.; Schade, B.; Böttcher, C.; Vostrowsky, O.; Hirsch, A. Amphiphilic [5:1]and [3:3]-Hexakisadducts of C-60. Eur. J. Org. Chem. 2004, 9, 1983–2001. [CrossRef]
112. Saenger, W.; Jacob, J.; Gessler, K.; Steiner, T.; Hoffmann, D.; Sanbe, H.; Koizumi, K.; Smith, S.M.; Takaha, T.
Structures of the common cyclodextrins and their larger analogues-beyond the doughnut. Chem. Rev. 1998,
98, 1787–1802. [CrossRef] [PubMed]
113. Diederich, F.; Whetten, R.L.; Thilgen, C.; Ettl, R.; Chao, I.; Alvarez, M.M. Fullerene isomerism: Isolation of
c2v,-c78 and d3-c78. Science 1991, 254, 1768–1770. [CrossRef] [PubMed]
114. Gutsche, C.D. Calixarenes Revisited. In Monographs in Supramolecular Chemistry; Stoddart, J.F., Ed.; The Royal
Society of Chemistry: Cambridge, UK, 1998; p. 190.
115. Suzuki, T.; Nakashima, K.; Shinkai, S. Very convenient and efficient purification method for fullerene (C60)
with 5, 11, 17, 23, 29, 35, 41, 47-Octa-tert-butylcalix (8) arene-49, 50, 51, 52, 53, 54, 55, 56-octol. Chem. Lett.
1994, 23, 699–702. [CrossRef]
116. Atwood, J.L.; Koutsantonis, G.A.; Raston, C.L. Purification of C60 and C70 by selective complexation with
calixarenes. Nature 1994, 368, 229–231. [CrossRef]
117. Saito, Y.; Ohta, H.; Terasaki, H.; Katoh, Y.; Nagashima, H.; Jinno, H.; Itoh, K.; Trengove, R.D.;
Harrowfield, J.M.; Li, S.F.Y. Separation of calixarenes with a chemically bonded C-60 silica stationary
phase in microcolumn liquid chromatography. J. Sep. Sci. 1996, 19, 475–477.
118. Brindl, R.; Albert, K.; Harris, S.J.; Tröltzsch, C.; Horne, E.; Glennon, J.D. Silica-bonded Calixarenes in
chromatography I. Synthesis and characterization by solid-state NMR spectroscopy. J. Chromatogr. A 1996,
731, 41–46.
119. Glennon, J.D.; Horne, E.; Hall, K.; Cocker, D.; Kuhn, A.; Harris, S.J.; McKervey, M.A. Silica-bonded calixarenes
in chromatography. 2. Chromatographic retention of metal ions and amino acid ester hydrochlorides.
J. Chromatogr. A 1996, 731, 47–55. [CrossRef]
120. Sanchez-Peña, M.; Zhang, Y.; Warner, I.M. Enantiomeric separations by use of calixarene electrokinetic
chromatography. Anal. Chem. 1997, 69, 3239–3242. [CrossRef]
Molecules 2016, 21, 1535
35 of 35
121. Healy, L.O.; McEnery, M.M.; McCarthy, D.G.; Harris, S.J.; Glennon, J.D. Silica-Bonded Calixarenes in
chromatography: Enantioseparations on molecular basket phases for rapid Chiral LC. Anal. Lett. 1998, 31,
1543–1551. [CrossRef]
122. Pfeiffer, J.; Schurig, V. Enantiomer separation of amino acid derivatives on a new polymeric chiral
Calix[4]arene stationary phase by capillary gas chromatography. J. Chromatogr. A 1999, 840, 145–150.
[CrossRef]
123. Ruderisch, A.; Pfeiffer, J.; Schurig, V. Synthesis of an enantiomerically pure resorcinarene with pendant
L -Valine Residues and its attachment to a Polysiloxane (Chirasil-Calix). Tetrahedron Asymmetry 2001, 12,
2025–2030. [CrossRef]
124. Ruderisch, A.; Pfeiffer, J.; Schurig, V. Mixed chiral stationary phase containing modified resorcinarene and
β-cyclodextrin selectors bonded to a polysiloxane for enantioselective gas chromatography. J. Chromatogr. A
2003, 994, 127–135. [CrossRef]
125. Allenmark, S. A Note concerning chromatography on mixed stationary phases. Enantiomer 1999, 4, 67–69.
126. Okamoto, Y.; Kawasaki, M.; Kawashima, M.; Hatada, K. Chromatographic resolution. 7. Useful chiral packing
materials for high-performance liquid chromatographic resolution of enantiomers: Phenylcarbamates of
polysaccharides coated on silica gel. J. Am. Chem. Soc. 1984, 106, 5357–5359. [CrossRef]
127. Trapp, O.; Caccamese, S.; Schmidt, C.; Böhmer, V.; Schurig, V. Enantiomerization of an inherently chiral
resorcarene derivative: Determination of the interconversion barrier by computer simulation of the dynamic
HPLC experiment. Tetrahedron Asymmetry 2001, 12, 1395–1398. [CrossRef]
128. Bürkle, W.; Karfunkel, H.; Schurig, V. Dynamic phenomena during enantiomer resolution by complexation
gas chromatography. A kinetic study of enantiomerization. J. Chromatogr. 1984, 288, 1–14. [CrossRef]
129. Welch, C. Imprintible brush-type chiral stationary phase. J. Chromatogr. 1995, 689, 189–193. [CrossRef]
130. Reich, S.;
Trapp, O.;
Schurig, V. Enantioselective stopped-flow multidimensional gas
chromatography—Determination of the inversion barrier of 1-Chloro-2,2-dimethylaziridine. J. Chromatogr. A
1999, 892, 487–498. [CrossRef]
131. Gebauer, S.; Friebe, S.; Scherer, G.; Gübitz, G.; Krauss, G.J. High-performance liquid chromatography
on Calixarene-bonded Silica Gels—Separation of cis/trans isomers of Proline-containing peptides.
J. Chromatogr. Sci. 1998, 36, 388–394. [CrossRef]
132. Schoetz, G.; Trapp, O.; Schurig, V. Determination of the cis-trans isomerization barrier of several
L -Peptidyl- L -Proline dipeptides by dynamic capillary electrophoresis and computer simulation.
Electrophoresis 2001, 22, 2409–2415. [CrossRef]
© 2016 by the author; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).