Syntheses and characterisation of novel cyclodextrin vinyl derivatives from cyclodextrin-nitrophenol-derivatives

Article abstract of DOI:10.1007/s10847-009-9711-x

In this study novel reactive α-, β- and γ-cyclodextrin-esters (acrylate, pent-4-enoate and undec-10-enoate) have been synthesised and characterised. The syntheses were carried out by using nitrophenol-esters with the ability to form inclusion complexes wi

Full text of DOI:10.1007/s10847-009-9711-x

J Incl Phenom Macrocycl Chem (2010) 67:303–315
DOI 10.1007/s10847-009-9711-x
ORIGINAL ARTICLE
Syntheses and characterisation of novel cyclodextrin vinyl
derivatives from cyclodextrin-nitrophenol-derivatives
•
Anne L. Nielsen Kim Steffensen
•
•
•
Reinhard Wimmer Martin Worm-Leonhard
Kim L. Larsen
Received: 4 August 2009 / Accepted: 17 November 2009 / Published online: 10 December 2009
Ó Springer Science+Business Media B.V. 2009
Abstract In this study novel reactive a-, b- and
c-cyclodextrin-esters (acrylate, pent-4-enoate and undec-
10-enoate) have been synthesised and characterised. The
syntheses were carried out by using nitrophenol-esters with
the ability to form inclusion complexes with cyclodextrins,
thereby aiming at a better control of the substitution degree
and number of positional isomers of the cyclodextrin
derivatives. Derivatives of a-, b- and c-cyclodextrins
modified with three different lengths of carbon-chains were
successfully synthesised and characterised by MALDI-
TOF MS, HPLC and LC-MS/MS, revealing some differ-
ences in LC elution patterns, substitution degrees and
number of produced positional isomers. Differences were
seen as an effect of changing the size of the cyclodextrin as
well as the size of the side-chain being attached. The
inclusion complexes between the nitrophenol esters and the
different cyclodextrins were studied by ITC and selected
ones by 2D ROESY NMR, showing some interesting dif-
ferences in strength and structure of the complexes. These
differences are speculated to be the origin of the different
substitution patterns of the derivatives as observed by
LC-MS/MS.
Keywords Cyclodextrin ꢀ Reactive derivative ꢀ
ROESY NMR ꢀ LC-MS/MS
Introduction
Cyclodextrins (CDs) are cyclic oligomers consisting of six
or more D-glucopyranose units linked with a(1–4) gluco-
sidic bonds and are often shown as truncated, cone-like
structures, where the exterior is polar and the interior less
polar relative to water. The commercially available CDs,
a-, b- and c-CD consists of 6, 7 and 8 glucosyl units,
respectively [1, 2]. Their properties are useful in a variety
of applications due to their strong ability to form inclusion
complexes with a large range of molecules having hydro-
phobic moieties corresponding in size with the cavity of the
CD. The CDs can thus be used to stabilise labile molecules,
to bring poorly water-soluble compounds for example
drugs into solution, masking of unwanted taste and odours,
modify chemical reactivity etc. [3].
A. L. Nielsen
Danish Technological Institute, Kongsvang Alle 29,
˚
8000 Arhus C, Denmark
The properties of the CDs can be altered in a number of
ways, to e.g. increase their solubility, decrease their tox-
icity or alter their inclusion abilities. This can be achieved
by substitution of one or more of the hydroxyl groups in the
glucose units. A wide variety of derivatives of especially b-
CD, but also a- and c-CD, have been made and some are
used extensively. As an example the hydroxypropyl- and
sulphobutylether-derivatives of b-CD have been used
extensively research on drug formulations [3]. In a number
of applications it is of interest to immobilise the CDs on a
surface or build them into a polymer while still maintaining
their inclusion functionality. This way the CDs ability to
K. Steffensen
Cheminova A/S, Thyborønvej 78, 7673 Harboøre, Denmark
R. Wimmer ꢀ K. L. Larsen (&)
Aalborg University, Sohngaardsholmsvej 57, 9000 Aalborg,
Denmark
e-mail: kll@bio.aau.dk
M. Worm-Leonhard
Institute of Forensic Medicine, University of Southern Denmark,
Winsløwparken 17, 5000 Odense C, Denmark
123

304
J Incl Phenom Macrocycl Chem (2010) 67:303–315
form complexes can be used in applications where release
of free CDs is unwanted, for example in hygiene products,
wound care products or removal of cholesterol with CD
containing membranes [3, 4]. When incorporating CDs in
polymers it is especially important to reduce the number of
reactive sites per CD to a minimum. Preferably there
should be only one substitution per CD molecule in order
for the CDs not to act as crosslinkers in the polymer net-
work. Small amounts of derivatives substituted on more
than one hydroxyl group may however be acceptable in
many applications, e.g. Janus and co-workers made water-
soluble CD polymers from unpurified CD vinyl derivatives
[5].
positional isomer [12, 13]. Depending on the type of
derivatives and MS equipment it is possible to separate
the derivatives, get a fingerprint showing how many dif-
ferent derivatives are formed and even specify exactly
which positional isomers have been produced [13]. These
types of analyses have previously been adapted to analyse
derivatives of oligosaccharides or CDs, e.g. both Tu¨ting
and co-workers as well as Lesur and co-workers worked on
fragmentation patterns of regioselectively O-methylated
maltooligosaccharides using ESI-MS/MS [12, 14].
In this study syntheses of a-, b- and c-CD vinyl deriv-
atives are synthesised using 4-nitrophenol-esters (4-NP) as
intermediates in a two-step approach, based on the work of
Harada et al. [11]. The aim is a high yield of monosub-
stituted CD as a result of the inclusion complex formation
during the reaction allowing the produced derivatives to be
used for production of CD polymers via free radical
polymerisation. In the present work 4-nitrophenol is used,
since it forms a stronger complex with b-CD than
3-nitrophenol according to literature (K1:1 *150 and
*560 M-1) [4, 11]. CD monomers with different aliphatic
esters (of 2, 4 and 10 carbon-chains, respectively) between
the CD and the double bond are produced in this way and
characterised by MALDI-TOF MS, NMR and HPLC to get
a full overview of the DS and distribution of the deriva-
tives. A CDScreenÒ HPLC column utilising the ability of
the CDs to form inclusion complexes is applied together
with MS/MS to achieve a detailed fingerprint of the
derivatives. Furthermore, the binding energies of the
inclusion complexes formed between the produced 4-NP-
esters and a-, b- and c-CDs are assessed by isothermal
titration calorimetry (ITC) and the respective complexes
formed between b-CD and NP-acrylate and NP-undec-10-
enoate is studied by 2D ROESY NMR.
When synthesising CD derivatives it is a challenge to
control the degree of substitution (DS) since the CDs have
18 (for a-CD) or more hydroxyl groups, which are chem-
ically more or less equivalent, with the primary ones being
just slightly more reactive than the primary ones [6]. It is
however possible to selectively substitute only one type of
hydroxyl groups to achieve a derivative with specific or
regioselective substitution [7, 8]. Thereby the obtained CD
is substituted on only one of the three types of hydroxyl
groups on the CD surface (2-, 3-, or 6-position in the
glucose ring) and this yields a better control of the DS of
the produced derivative [9]. It is possible to control the
reactions due to the slightly lower reactivity of the C-2 and
C-3 alcohol hydroxyls compared to the C-6 alcohol
hydroxyls [7, 9]. To control the DS several methods can be
employed, many requiring a number of protection and de-
protection steps to prevent more than one hydroxyl group
from reacting and some followed by column chromatog-
raphy for purification of the desired derivative [10].
Another approach to control the DS of the derivatives is to
direct the reaction by using compounds with the ability to
form inclusion complexes with the CDs. This was first
proposed by Harada and co-workers, who used a 3-nitro-
phenol-acrylate ester to control the substitution of a b-CD-
acrylate vinyl derivative [11]. The nitrophenol-ester is
believed to form only one type of complex which will
direct the reaction to the secondary alcohols of the CD (C-2
and C-3) and thereby increase the yield of mono-substi-
tuted product [11]. The yield of mono-substituted deriva-
tives from this reaction is reported to be around 15%, with
emphasis on keeping a low amount of di- and tri-substi-
tuted derivatives.
Materials and methods
Pharmaceutical grade a-, b- and c-CD were purchased from
Wacker Chemie, Burghausen, Germany. 4-nitrophenol
(98%) (4-NP), 3-nitrophenol (98%) (3-NP), acryloyl chlo-
ride (96%, C200 ppm monomethyl ether hydroquinone
(MEHQ) as inhibitor), pentenoyl chloride (98%) and un-
decenoyl chloride ([97%) were all purchased from Sigma–
Aldrich, Steinheim, Germany. a-cyano-4-hydroxycinnamic
acid (pro analysis) (CCA), trifluoroacetic acid (TA), nitro-
cellulose, acetone (99.8%) and acetonitrile (AcN) (99.9%)
all of analytical grade were obtained from Sigma–Aldrich,
St. Louis, Missouri, USA. Potassium hydroxide (KOH) was
obtained from Sigma–Aldrich, Steinheim, Germany.
Hydrochloric acid was purchased from J.T. Baker, Deventer,
Holland and sodium carbonate was obtained from Appli-
Chem, Darmstadt, Germany. Deuterium oxide (D2O),
A way to determine the positional isomers of modified
CDs is LC-MS, which gives a possibility to separate the
different isomers on a suitable column and subsequently
analyse them on MS obtaining the exact mass of each. By
using collision induced dissociation (CID) MS/MS it is
even possible to get a fingerprint of each, thereby revealing
different regio-isomers produced. This is due to differences
in collision stability during CID depending on the type of
123

J Incl Phenom Macrocycl Chem (2010) 67:303–315
305
¨
99.9% was obtained from Larodan Fine Chemicals, Malmo,
Sweden. All chemicals were used as received without fur-
ther purification. Millipore water was obtained from a
Nanopure Diamond system from Barnstead International,
Dubuque, Iowa USA.
plate. All samples were measured in reflector mode
detecting positive ions. FlexAnalysis 2.4 (build 11) soft-
ware (Bruker Daltonik GmbH, Bremen, Germany) was
used for the analysis of the obtained data.
Nuclear magnetic resonance
Syntheses
1H and 2D Rotational nuclear Overhauser Effect Spec-
troscopy (ROESY) NMR analyses were carried out on a
Bruker DRX600 NMR spectrometer (Bruker, Bremen,
Germany) equipped with an XYZ gradient TXI (H/C/N)
probe at 298 K. All samples were prepared in D2O with
presaturation of the water signal. ROESY spectra were
recorded at 300 K. Bruker TOPSPIN software version 2.0
(2006) (Bruker Biospin) was used for analysis of the data.
In the first step the unsaturated acid chlorides (acryloyl
chloride, pentenoyl chloride or undecenoyl chloride) were
reacted with 4-NP adapting the method of Harada [11]. In
brief, 1.1–2.5 g (in each case 12.1 mmoles) of the chosen
acid chloride was added drop wise to a solution of 1.25 g
potassium hydroxide and 1.85 g (13.2 mmoles) of 4-NP in
80 mL of water in an ice bath, keeping the temperature
below 5 °C at all times. In the cases of nitrophenyl-acrylate
(NP-acrylate) and -pent-4-enoate (NP-pent-4-enoate) the
products were obtained as a white/yellow precipitate in 81
and 89% yields, respectively, whereas the nitrophenol-
undec-10-enoate (NP-undec-10-enoate) was obtained as a
Isothermal titration calorimetry
ITC measurements were carried out on a VP-ITC micro
calorimeter (MicroCal Inc. Company, Northampton, MA,
USA) controlled by MicroCal’s VP viewer software (ver.
5.0, Microcal Software Inc., Northampton, MA, USA).
Titration was done with a 10 mM CD solution into 1 mM
solutions of 3-NP, 4-NP and the four NP-derivatives, all in
acetate buffer at pH 3.5. After thermal equilibrium of the
titration cell was reached, an initial delay of 180 s followed
by an initial injection of titrant of 1 lL was applied. After
this injection, volumes from 10 (first 10 datapoints) to 20
(next 10 datapoints) lL and delay times between injections
of 250 s were applied.
1
brown/yellow oil (mp *10 °C) in 87% yield. H NMR,
NP-acr: d = d8.2, d7.05, d6.7, m6.5, d6.3 ppm; NP-und:
d = d8.2, d7.1, s1.4–1.7 ppm. For the next step 1 g of CD
(a-, b-, or c-CD) (0.77–1.03 mmoles) was dissolved in
60 mL of 0.1 M carbonate buffer, pH 10 and either of the
three NP-derivatives was added in a 1:1 molar ratio to CDs.
The NP-acrylate and -pent-4-enoate were dissolved in
1 mL of AcN prior to addition to the aqueous CD solution
to prevent precipitation. The reaction proceeded for 15 min
while stirring and was then adjusted to pH * 3 with
diluted HCl thereby stopping the reaction. The reaction
mixture was cooled to 5 °C and stored overnight before
precipitates of unreacted NP and NP-derivatives were
removed. The CD derivatives were collected as white
powders in yields between 50 and 80% by evaporation of
the solvent and purified from excess NP by recrystallisation
from AcN. The derivatives are named according to the CD
used, the attached chain and in HPLC and MS additionally
after the number of substituents on the specific CD, for
example b-CD-acr1 for b-CD derivative with one acrylate
chain attached.
Data processing was done using Origin 7 SR2 v 7.0383
software (OriginLab Corporation, Northampton, USA)
with the microcal LLC ITC add-On and manual baseline
correction. The data obtained during titration were con-
verted to heat output per injection by integration of the
peaks and correction for the cell volume and sample con-
centration. The data were fitted by using ‘‘one set of sites’’
model in the Origin software based on the Wiseman Iso-
therm [15] as shown in Eq. 1:
dQ
d½NPꢁtot
0
1
ꢀ
ꢁ
Analyses
1
1ꢂ½NPꢁtot
ꢂ
½CDꢁKeq
B
B
@
C
C
A
1
2
0
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ DH V0
þ
ꢀ
ꢀ
ꢁꢁ
MALDI mass spectrometry
Keq 2ꢂ4½NPꢁtot
1
2
1þ½NPꢁtot ꢂ
½CDꢁ
The CD derivatives were analysed using MALDI-TOF MS
(ReflexTMIII from Bruker, Bremen, Germany) using a fast
evaporating nitrocellulose layer and a CCA matrix. Sam-
ples were deposited using a thin layer method, with the
sample being deposited on top of the applied matrix layer
on the target plate. Aqueous solutions of the samples were
prepared, mixed with CCA matrix and applied to the target
ð1Þ
where Q is the heat flow, [NP]tot the total concentration of
NP or NP derivative (ligand), H0 is the standard molar
enthalpy, V0 is the volume of the reaction cell [CD] is the
concentration of CD and Keq is the overall binding con-
stant. The model assumes that only one set of binding sites,
123

306
J Incl Phenom Macrocycl Chem (2010) 67:303–315
all with same binding energy, is present in the sample. In
all cases the number of binding sites (n) was fixed to one in
the software as recommended by Turnbull and Daranas
[16] for systems of relatively low binding constants com-
pared to aqueous solubility (hence the Wiseman ‘‘c’’
parameter which is the product of the receptor concentra-
tion and the binding constant, Ka is less than approximately
10) [16].
CD-acrylate, CD-pent-4-enoate and CD-undec-10-enoate.
In all cases the aim was to achieve a high yield of mono-
substituted CDs for use in synthesis of CD-polymers where
possible crosslinking arising from di-substituted products is
unwanted. Controlling the reaction by formation of CD-NP-
ester complexes should increase the yield of mono-substi-
tuted CDs. In this work 4-NP-esters have been used in the
intermediate product, since the 4-NP is reported to be the NP
having the highest stability constant in complexation with
CDs [17]. The data on the obtained products of derivatives
are first described and is followed by a more detailed study
of the CD–NP inclusion complexes via binding energies and
spatial structure predictions. Finally fingerprints of the
products obtained via LC-MS/MS, giving indications on the
regioselectivity of the reactions, are presented.
HPLC and LC-MS/MS
CD derivatives were measured on a CDScreenÒ column
(ChiroQuest Ltd., Budapest, Hungary) using various gra-
dients of water and AcN (shown in the figure text of the
relevant graphs) on a Dionex HPLC (Dionex Corporation,
Sunnyvale, CA, USA) system, consisting of a Dionex P680
HPLC pump, a Dionex ASI-100 autosampler and an All-
tech 800 (Alltech Associates Inc. Deerfield, IL, USA) ELS
detector along with a UV detector (Dionex UVD 170U).
For data analysis Chromeleon software (Dionex Client 6.60
SP1 Build 1447, Dionex Corporation, Sunnyvale, CA,
USA) was applied. All measurements were done at ambient
temperature (258).
A sketch of the overall reaction is shown in Fig. 1 for
reactions between CDs and NP-acrylate (Fig. 1a) as well as
-pent-4-enoate and -undec-10-enoate (Fig. 1b). Attempts to
produce a- and c-CD derivatives with pent-4-enoate were
not done, instead focus was put on syntheses of a- and c-
CD-und.
Initial characterisation of CD derivatives
For LC-MS/MS analyses the abovementioned
CDScreenÒ column was also applied together with an
Agilent 1200 HPLC-system (consisting of a G1322A
degasser unit, G1311A pump, G1329A autosampler and
G1316A column oven) equipped with Bruker micrOTOF-
Q IITM (Bruker Daltonik, Bremen, Germany) MS detector
running at 4.5 kV for ionisation, 6 L/min of nitrogen as
drying gas at 180 °C and evaporator of 2.5 bars. Formic
acid (0.1%) and AcN were used at a flow of 0.4 mL/min
under isocratic conditions, differing in composition
depending on the CD derivative (a-CD-acr: isocratic con-
ditions, 10% AcN, 90% formic acid, b-CD-acr: isocratic
conditions, 13% AcN, 87% formic acid, b-CD-und: iso-
cratic conditions, 20% AcN, 80% formic acid). The formic
acid ensures a good formation of protonated sugar ions and
good signal in the MS analyses. For Collision Induced
Decomposition (CID) experiments, argon was used as
collision gas and the collision energy was optimised by the
software for each precursor ion ramping from 10 to
100 eV. Cluster ions of sodium formate were used for
internal calibration.
The CD vinyl derivatives were produced using NP-esters as
intermediate carriers of the aliphatic chains, and the
obtained products were analysed by MALDI-TOF MS and
HPLC to reveal the DS and the number of different
derivatives. The MALDI-TOF MS data for the a-, b- and c-
CD-acr derivatives are shown in Fig. 2a and for b-CD-
pent, a- and b-CD-und derivatives in Fig. 2b.
It proved possible to produce the desired acrylate-
derivatives of both a-, b- and c-CD with an acceptable
yield of mono substituted product as observed in Fig. 2a
where a fraction of unreacted CD is also observed. From
Fig. 2b it is evident that derivatives of a- and b-CD-und as
well as b-CD-pent have also been produced, whereas no
product of c-CD-und seem to be formed and only native
c-CD was observed (data not shown). The spectra show a
distribution of unreacted CDs and derivatives with up to
3–4 substitutions on each CD for a- and b-CD-acr whereas
c-CD-acr seems to be of lower DS. The spectrum of b-CD-
pent derivatives show small amounts of di-substituted
compounds whereas the a- and b-CD-und show no trace of
higher DS than mono-substituted, which may indicate
differences in the reaction mechanism for these longer
aliphatic chains compared to the short acrylate-chain. An
explanation to this could be found in differences in the
structure of the inclusion complexes between b-CD and
NP-pent and NP-und respectively, this will be investigated
further in the next section.
Results and discussion
Syntheses of CD vinyl derivatives utilising the CDs ability
to form inclusion complexes with NP-derivatives have
been conducted in this study. The method have been
developed from the work of Harada et al. [11], but
expanded to produce a larger array of derivatives of
The obtained products were analysed by HPLC to obtain
an overview of the relative amounts of the different DS in
123

J Incl Phenom Macrocycl Chem (2010) 67:303–315
307
Fig. 1 Synthesis and
characterization of novel
reactive cyclodextrin
derivatives. Proposed reaction
scheme of CD derivative
O
O
O
O
O
syntheses via inclusion complex
with NP-derivative. a Shows the
reaction with 4-NP-acrylate and
b the reaction with 4-NP-pent-
4-enoate and -undec-10-enoate.
The inclusion of NP into the CD
cavity will probably differ with
the type of CD and this should
be regarded only as a suggestion
+
HO
-CD
-CD
-CD
α
NO2
NO2
β
γ
NO2
NO2
O
(CH2)x
+
O
-CD
-CD
-CD
α
O
O
O
O
β
x= 2 or 8
γ
(CH2)x
(CH2)x
Fig. 2 a MALDI-TOF MS
spectra of a-, b- and c-CD-acr
showing the derivatives found
in the sample. b MALDI-TOF
MS spectra of a-, b- and c-CD-
und and b-CD pent showing the
derivatives present in the
sample. All samples have been
doped with small amounts of
sodium chloride
α-CD/Na+ m/z=995
β-CD/Na+
A
B
m/z=1157
β-CD-pentenoyl1/Na+
α-CD-acryloyl1/Na+ m/z=1049
α-CD-acryloyl2/Na+ m/z=1103
m/z=1103
β-CD/Na+ m/z=1157
α-CD/Na+
m/z=995
β-CD-acryloyl1/Na+ m/z=1211
α-CD-undecenoyl1/Na+
m/z=1161
β-CD-acryloyl2/Na+ m/z=1103
γ-CD/Na+ m/z=1319
β-CD/Na+
m/z=1157
γ-CD-acryloyl1/Na+
m/z=1373
β-CD-undecenoyl1/Na+
m/z=1323
950
1150
1350
1550
950
1150
1350
1550
m/z
m/z
the samples, since this information is not readily accessible
from the MS data. For the HPLC analyses the samples were
analysed on a CDScreenÒ column offering high selectivity
since the separation is based on the ability of the CDs to
form inclusion complexes with nitrophenyl groups of the
stationary phase [18]. The ability of the CD derivatives to
form inclusion complexes will be dependent on the DS and
separation can be achieved. An evaporative light scattering
detector (ELSD) was used, since it offers an excellent
sensitivity towards CDs, and in the next run fractions were
collected and analysed by MALDI-TOF-MS. In Fig. 3 the
chromatogram of the b-CD-acr product is shown, along
with the MALDI-TOF MS data obtained for each peak.
From the HPLC chromatograms a reasonable separation
of the derivatives can be seen, even for mono-substituted
product. The separation of b-CD from mono- and
di-substituted products (b-CD-acr1 and b-CD-acr2) seem
straightforward in Fig. 3a, whereas the higher substitution
degrees doesn’t seem to be well separated from native CD.
The fifth peak contains mono- and di-substituted products,
which where also found in peak three and four, respectively.
The tri-substituted product (b-CD-acr3) is even eluted before
the unmodified b-CD. This is an indication that a higher
number of side-chains can affect the ability of the CD to
form the necessary inclusion complexes with the nitro-
phenyl-compound of the stationary phase in the column and
affect the separation on HPLC. The chromatogram of b-CD-
und is shown in Fig. 3b and this proves the presence of only
unreacted and mono-substituted CD. This confirms the
findings from the initial MALDI-TOF analyses on the crude
mixture and could suggest a better control of the reaction by
the NP-undec-10-enoate-ester derivative compared to the
123

308
J Incl Phenom Macrocycl Chem (2010) 67:303–315
400
350
300
250
200
150
100
50
A
A
3
Peak # m/z value (Na+) Assignment Percent
2
1
1156
1317
β-CD,
β-CD-acr3
β-CD
β-CD-acr1
β-CD-acr2
β-CD-acr1
β-CD-acr2
2.8
2
3
4
5
1156.0
1209.9
1263.9
1209.9
1263.9
77.0
10.3
5.9
4
5
1
4.0
0
14
15
16
17
Ret. time (min)
18
19
20
140
120
100
80
B
B
Peak # m/z value (Na+)
Assignment
β-CD
β-CD-und1
Percent
96.6
3.4
1
1
2
1155.9
1317.8
60
2
40
20
0
10
11
12
13
14
15
16
17
Ret. time (min)
200
180
160
140
120
100
80
C
C
2
Peak # m/z value (Na+) Assignment
Percent
1
1317.9
1479.9
1317.9
1372.0
1372.0
γ-CD
γ-CD-acr3
γ-CD
γ-CD-acr1
γ-CD-acr1
1.1
3
2
3
4
93.7
4.8
0.3
60
1
40
4
20
0
17
18
19
20
21
Ret. time (min)
Fig. 3 HPLC chromatograms of CD-derivatives. Peaks are numbered
and MALDI-TOF MS information for each peak is shown in the
tables. a b-CD-acr; flow: 0.4 mL/min, gradient: 0–2 min: 10% AcN,
2–18 min: 10–30% AcN, 18–24 min: 30% AcN, 24–26 min: 30–10%
AcN, 26–32 min: 10% AcN. b b-CD-und; flow: 0.4 mL/min, isocratic
conditions 15% AcN in millipore water. c c-CD-acr; flow: 0.4 mL/min,
gradient: 0–2 min: 10% AcN, 2–18 min: 10–30% AcN, 18–24 min:
30% AcN, 24–26 min: 30–10% AcN, 26–32 min: 10% AcN
NP-acrylate-ester, but also lower yields. In the case of
c-CD-acr (Fig. 3c) the di-substituted product is not
detected, supporting the findings from MALDI-TOF MS
on the crude mixture. A possible explanation for the
complex separation patterns observed on HPLC could be
the ability of the column to distinguish between positional
isomers of the CD derivatives, i.e. derivatives with equal
number of side-chains, but attached to different hydroxyl
groups on the CD. This will lead to two different posi-
tional isomers of for instance CD-acr1 to leave the col-
umn at different retention times. This possibility of
separation of the regio-isomers of the derivatives will be
further tested by LC-MS/MS.
in Table 1. The DS are all calculated in weight-percentage
relative to each CD.
Some discrepancies are observed between the numbers
from different analysis methods, but overall it is possible to
observe DS between 1.5 and 10 which compares well to the
earlier studies with monomer-yields of around 15% [11]. In
all cases the quantification method should only be con-
sidered as an overall estimate due to the nature of the
methods. The MALDI-TOF method is only considered
‘‘semi-quantitative’’ [19], in some cases enhancing the
formation of ions from one species. Interpretation of HPLC
data can also be quite tricky because of the possibility of
insufficient separation of the free CDs and derivatives. All
in all comparing data from the methods provides some
overview of the DS of the obtained derivatives.
The results obtained using HPLC and MALDI-TOF MS
provided the possibility to calculate the average DS as seen
123

J Incl Phenom Macrocycl Chem (2010) 67:303–315
309
of b-CD-acr1 ions from the two different elution times in
the chromatogram (Fig. 4a, 2–3 on the right hand side)
reveals some difference in the fragmentation patterns
which indicates that these two peaks in fact represents two
different regio-isomers of b-CD-acr1. The b-CD-acr2 ions
are even observed to leave the column as four distinct
peaks (no MS/MS data obtained). The data give strong
indications of two types of regio-isomers of the CD, for
example CDs being substituted on at least two different of
the alcohol hydroxyl (C-2, C-3 and C-6). This can account
for the different fragmentation patterns due to the differ-
ences in the spatial structure. The position of the side-chain
will affect the collision stability of the modified glucose
units and lead to different fragmentation patterns as
observed. The same trends are observed from the data on
the a-CD-acr derivative, and even though a full separation
of the pure CD from the derivatives was not obtained, the
data of the extracted ion chromatogram of a-CD-acr1 show
two semi-overlapping peaks with different fragmentation
patterns as seen in Fig. 4b.
Table 1 Overall substitution degrees obtained from MS and HPLC,
respectively, given in weight percentages
MS (%)
HPLC (%)
a-CD-acr
a-CD-und
b-CD-acr
b-CD-pent
b-CD-und
c-CD-acr
c-CD-und
21.9
1.3
4.8
1.5
10.3
4.4
3.4
4.8
–
34.2
7.2
12.6
11.7
–
Calculated as the average number of substituents on a CD ring
Comparing the data in Table 1 to the MS data in Fig. 2
it is evident that the synthesis of CD-acrylate-derivatives
yields a number of different products appearing as quite
high DS in Table 1 compared to the findings from the other
syntheses. In the case of the b-CD-pent lower overall DS
and yields are observed, but also higher amounts of mono-
substituted CDs as shown by MALDI-TOF MS in Fig. 2.
The b-CD-und synthesis seems to be relatively selective
and yield only one substitution which was detected by
MALDI-TOF MS, thereby leaving the rest of the CDs
unreacted which is reflected by the low yield observed in
Table 1. Whether the yield can be enhanced by changes in
reaction time, temperature, pH or molar ratio between
undec-10-enoate and CD has not been investigated, but this
may be the case.
The extracted ion chromatogram of b-CD-und in Fig. 4c
shows a peak of the b-CD-und1 product just before the
peak of the pure b-CD and another part of b-CD-und1
appearing at the same retention time as the b-CD peak. It
does also seem that there are some impurities being
detected in this chromatogram, possibly due to the rela-
tively low signal. No MS/MS data were obtained for this
derivative, but when comparing to the data obtained for b-
CD-acr it seems likely that the different peaks represent
different positional isomers.
Fingerprinting of CD derivatives
From the obtained LC-MS/MS data it has not been
possible to prove the exact regiochemistry of the deriva-
tives, since the CID did not provide sufficient amounts of
different fragment ions. It is however clear that there are
differences between the patterns of fragment ions arising
from the same ion observed at different retention times in
the chromatogram. This is an indication of different posi-
tional isomers eluting differently from the HPLC column.
The separation of positional isomers has previously been
shown possible with amphiphilic CDs by Kieken and co-
workers, showing the same tendency but with poorer sep-
aration of the isomers due to the use of a column based on
hydrophobicity instead of the CD selective one used in this
study [20].
A fingerprint of the CD-derivatives was obtained by LC-
MS/MS in combination with the CDScreenÒ HPLC col-
umn. From the LC-MS/MS data extracted ion chromato-
grams were obtained showing the elution of an ion with a
specific mass like e.g. b-CD-acr1 of 1211 m/z. This pro-
vided a better picture of the elution pattern of each ion as
the HPLC-ELSD studies showed a seemingly complex
elution pattern. In addition it was possible to perform CID
via argon bombardment to achieve ion fragments for
selected ions. The pattern of ion fragments is specific for
different positional isomers (i.e. substituted on the primary
or secondary hydroxyls) on the CDs due to differences in
collision stability. The obtained data are shown in Fig. 4
with total ion chromatogram, extracted ion chromatograms
and fragment ions in the insert for relevant peaks.
Characterisation of NP-inclusion complexes
For the b-CD-acr derivative the data shown in Fig. 4a
show one peak for the ion of pure b-CD (1135.36 m/z)
in the extracted ion chromatogram but two peaks for the
b-CD-acr1 ion (1189.38 m/z) one of them with almost the
same retention time as the b-CD peak indicating an
incomplete separation of these compounds. Fragmentation
Isothermal titration calorimetry has been applied to deter-
mine binding constants of b-CD and 4-NP, NP-acrylate,
NP-pent-4-enoate and NP-undec-10-enoate, respectively.
This has been done in order to reveal differences in binding
with the different NP-derivatives to compare with the dif-
ferent substitution patterns found by MS/MS, since the
123

310
J Incl Phenom Macrocycl Chem (2010) 67:303–315
A
.
4
1
325.112
β-CD
325.1124
1
587.166
487.165
587.1661
487.1647
649.217
649.2167
145.049
145.0490
289.901
540.070
289.0910
540.0704
200
400
600
.
4
m/z
β-CD-acr
614.6737
1
217.0698
2
3
2
325.1121
379.1236
145.0498
703.2270
541.1741
487.1659
97.0282
3
217.0706
.
β-CD-acr
614.6743
2
325.1127
145.0491
541.1712
487.1658
379.1235
422.4349
703.2244
284.0678
200
400
600
m/z
0
5
10
15
20
25
Time (min)
Fig. 4 Data obtained from LC-MS/MS. a Data from MS/MS data for
b-CD-acr; top left b-CD extracted ion chromatogram, below b-CD-
acr1 ion chromatogram and b-CD-acr2 ion chromatogram. On the
right hand side the CID spectra are shown for each peak as marked
with numbers. b Data from MS/MS data for a-CD-acr; top left a-CD
extracted ion chromatogram, below a-CD-acr1 ion chromatogram and
a-CD-acr2 ion chromatogram. On the right hand side the CID spectra
are shown for each peak as marked. c Showing top left b-CD ion
chromatogram (top) and b-CD-und1 extracted ion chromatogram. No
MS/MS data was obtained for this compound
strength of the complex is speculated to be the main factor
of controlling the reaction. Furthermore, the inclusion
between 4-NP and 3-NP with all three CDs were investi-
gated to enable a comparison with the work of Harada
where a 3-NP derivative was applied [11]. In the titrations
the NP-compound was kept in the cell and CD solution was
loaded in the syringe, all experiments were conducted at
pH 3.5 to avoid any transesterification reaction between
CD and NP-derivatives. The Wiseman Isotherm was used
in fitting of the data to obtain binding constants between
CD and NP-compound and these data are shown in
Table 2.
4-NP were in all cases found to be around twice the size of
CD binding constants obtained with 3-NP. This indicates as
expected that the use of 4-NP-derivatives instead of 3-NP-
derivatives could possibly result in better selectivity in the
reaction with CDs due to the higher strength of the com-
plex between 4-NP and CDs.
The substitution on 4-NP with unsaturated aliphatic
chains may induce a change in the binding energy with
CDs and this was also tested with ITC. It is clear from the
data in Table 2 that this is in fact the case. The 4-NP-
acrylate and 4-NP-pent-4-enoate show a slightly lower
binding constant with b-CD than what was observed for
pure 4-NP. This could possibly be attributed to a change in
the 3D structure of the complex between 4-NP and b-CD
caused by the change of a hydroxyl group on the nitro-
phenol into a small aliphatic chain, which may affect the
For all three types of CDs it is evident that the binding
constant between 3-NP and the CD is weaker than that
between 4-NP and CDs, but also that all of the complexes
are rather weak. The binding constant between CDs and
123

J Incl Phenom Macrocycl Chem (2010) 67:303–315
311
B
α-CD
+MS, 7.74min #457
506.1406
506.141
1
487.167
325.113
649.221
487.1666
325.1133
649.2205
1
145.049
145.0488
163.058
127.038
163.0576
540.076
127.0376
691.179
289.096
185.043
378.024
540.0763
289.0958
185.0427
378.0237
200
400
600
m/z
α-CD-acr
1
8.29min #490
506.140 +MS,
3
2
533.146
2
145.049
145.0490
325.113
325.1134
217.071
217.0706
4874.817.166883
433.099
127.038
379.122
349.064
379.1220
228844.06.807 69
163.060
118855.04.109 42
127.0381
163.0601
349.0636
97.027
433.0991
97.0266
0
4
+MS, 8.75min #517
533.1448
533.146
3
α-CD-acr
2
217.070
217.0703
145.049
325.114
145.0488
325.1135
284.070
284.0697
379.122
127.040
379.1217
127.0397 163.060
487.170
460.104
487.1699
460.1044
349.064
163.0616
349.0688
109.0255
100
200
300
m/z
400
500
0
2
4
6
8
10
12
14
Time (min)
C
β-CD
β-CD-und1
0
5
10
15
20
25
Time (min)
Fig. 4 continued
123

312
J Incl Phenom Macrocycl Chem (2010) 67:303–315
Table 2 Binding constants between 3-NP, 4-NP and 4-NP derivatives and a-, b- and c-CDs given at 30 °C in M-1
3-NP
4-NP
4-NP-acr
4-NP-pent
4-NP-und
a-CD
b-CD
c-CD
23.3 3.01
125 3.61
145 21.4
64.5 3.76
240 14
–
–
–
141 7.05
–
111 5.46
–
782 52.7
–
243 36.7
Errors indicate the error of the fit of the data to the Wiseman isotherm
position of the nitrophenol inside the CD cavity. The
change to a 4-NP carrying a small aliphatic chain may
make the inclusion complex formation less favourable due
to steric hindrance, hence affecting the stability constant
negatively. The 4-NP-undec-10-enoate derivative by far
has the largest binding energy with b-CD. This gives an
indication of a complex formation taking place not only at
the phenyl ring but also at the aliphatic undec-10-enoate
chain. Binding constants of free NP and aliphatic chains of
similar structure to undec-10-enoate from the literature
show that the binding constant with an aliphatic chain is
approximately a factor of 10 higher than that of 4-NP [17,
21], supporting the findings in this study. Another ITC
study of substituted benzoic acids show data in good cor-
relation with the ones found in this study, i.e. with a
stronger binding as an effect of substitution by a longer
aliphatic chain but no effect of relatively short chains [22].
Additional thermodynamic parameters were also
obtained from the ITC studies as shown in Table 3. Theses
thermodynamic parameters are evaluated relative to each
other to give a small insight into the differences in binding
for the different NPs and CDs, even though they are
associated with a relatively high error in systems of low
stability constants as in this case [16].
even less as shown by smaller enthalpy change, but the
entropy contribution is positive, indicating that no rota-
tional freedom is lost.
The complex formation between NP-derivatives and b-
CDs is observed to be driven by quite high enthalpy con-
tributions, especially in case of NP-pent, but they are
accomplished by a great loss of rotational freedom and
negative entropy contribution. In the case of NP-und the
binding is weaker (larger DH) but at the same time no
freedom is lost. The differences in thermodynamics
between the binding of the NP-derivatives indicates that
the binding to NP-und may take place at another part of the
molecule than in case of NP-acr and NP-pent. It could
therefore be assumed that not only the NP-ring of NP-und,
but also the aliphatic undec-10-enoate chain participate in
inclusion complex formation.
The formation of inclusion complexes between b-CD
and the modified NPs was further tested by ROESY 2D
NMR (Rotational Overhauser Effect SpectroscopY), a
technique revealing the proximity in space of components
as those found in an inclusion complex. COSY (Corre-
lation SpectroscopY) spectra were utilised to assign the
signals in the CDs. The COSY spectrum of b-CD and
NP-acrylate (not shown) revealed the following shifts of
the CD protons: H1:5.2, H2: 3.74, H3: 3.98–4.0, H4: 3.7–
3.75, H5: 3.86 and H6: 3.96 ppm. In Fig. 5 a ROESY
spectrum of b-CD and NP-acrylate is shown, crosspeaks
in the spectrum show a spatial interaction between two
molecules as indicated.
Some differences are observed between the different
complex formations. a-CD seems to have the strongest
binding to 3- and 4-NP seen by the large favourable
enthalpy contribution, but this is counteracted by a large
unfavourable entropy contribution, indicating that the fit
between a-CD and NP is strong, but lead to a dramatic loss
in rotational freedom. The interaction between NPs and
b-CD show a weaker binding, but also a lower loss of
rotational freedom observed in the entropy contribution
being around zero. In the case of c-CD, the interaction is
In the ROESY spectrum it is evident that the protons
inside the CD cavity (protons H3, H5 and C6) interact with
the protons on the aromatic NP ring, corresponding to the
crosspeaks at 7–8 ppm on the F2 axis and 3.8–4.2 ppm on
the F1 axis. No interactions between the protons inside the
Table 3 Enthalpy and entropy of complex formation between 3-NP, 4-NP and 4-NP derivatives and a-, b- and c-CDs all given at 30 °C
in kJ/mol
3-NP
4-NP
4-NP-acr
4-NP-pent
4-NP-und
DH
-TDS
DH
-TDS
DH
TDS
DH
-TDS
DH
-TDS
a-CD
b-CD
c-CD
-46.4 5.7
-14.4 0.3
-2.9 0.3
38.4
2.3
-25.0 1.2
-11.3 0.5
-1.7 0.2
14.5
-2.5
–
–
–
–
–
–
-14.6 0.6
2.1
–
-31.5 1.3
19.7
–
-9.7 0.3
-7.1
-9.6
-12.1
–
–
–
–
Errors indicate the error of the fit of the data to the Wiseman isotherm
123

J Incl Phenom Macrocycl Chem (2010) 67:303–315
313
CD-
protons
H-3
H-3
(NP-acr)
H-2
(NP-acr)
-
x
H-5
H-6
xx
xx
xx
xxx
Fig. 5 2D ROESY spectrum of b-CD and NP-acrylate, showing
spatial interactions between the two components. Measured in pure
D2O, pD \ 6 to prevent reaction between NP-acrylate and b-CD.
Crosspeaks (away from the diagonal) show spatial interactions. The
table next to the spectrum shows the strength of the signal as marked
by x for weaker interaction and xxx for stronger interaction
CD-
protons
H-3
H-5
H-6
H-3
(NP-und)
H-2
(NP-und)
(CH2)8
(NP-und)
xx
xxx
xx
x
x
x
xx
x
xx
Fig. 6 ROESY spectrum of b-CD and NP-undec-10-enoate, showing
spatial interactions between the two components. Measured in D2O,
pD \ 6 to prevent reaction between NP-undec-10-enoate and b-CD.
Crosspeaks (away from the diagonal) show spatial interactions. The
table next to the spectrum shows the strength of the signal as marked
by x for weaker interactions and xxx for stronger interactions
cavity and the acrylate-group are observed; these were
expected to be visible at 2–3 ppm. Only the aromatic
protons closest to the nitro group seem to interact with the
protons on H3 of the CD, at the same time the aromatic
protons near the acrylate end seem to have a larger cros-
speak i.e. stronger interaction with the protons on H6 of the
CD. These interactions and their strength is shown in the
table and the observations indicate that the main part of the
complexes have the acrylate end of the NP-ring near the
narrow rim of the CD.
complex, the data may suggest that the formed CD-acrylate
derivatives will be mainly substituted on the H6 (primary)
of the CDs. Since more than one position of the acrylate
chain on the CD was observed from LC-MS/MS experi-
ments, it seems the conformation found from ROESY
NMR is not giving all the details about the reactive con-
formation in solution.
The COSY spectrum of b-CD and NP-undec-10-enoate
(not shown) revealed the following shifts of CD protons:
H1: 5.2, H2: 3.74, H3: 3.9–4.0, H4: 3.68–3,7, H5: 3.86 and
H6: 3.96 ppm. The ROESY spectrum of b-CD and NP-
undec-10-enoate is shown in Fig. 6.
If the ROESY data are assumed to represent the most
frequent as well as most reactive configuration of the
123

314
J Incl Phenom Macrocycl Chem (2010) 67:303–315
In the ROESY spectrum it is again evident that the
binding constant with 4-NP and indications of one pre-
dominant 3D conformation shown by ROESY. In addition
the complexes of NPs with b- and c-CD are around four
times that of NP complexes with a-CD without any
observed tendency to differences in substitution pattern. If
the inclusion complex plays a key role in the reaction
mechanism a higher difference in the obtained derivative
distributions was expected. Since this is not the case it
could be assumed that the effect of the NP-acrylate inclu-
sion is insignificant compared to other factors affecting the
DS. This is supported by the finding of relatively low
stability constants between CDs and NPs. The observations
of more mono-substituted product in the derivatives with
longer aliphatic chains (pent-4-enoate and undec-10-eno-
ate) indicate that the presence of an aliphatic chain on the
NP-derivatives provides a better control of the DS than a
small acrylate-chain, possibly by self-inclusion leading to
occupation of the cavity.
protons inside the CD cavity (H3, H5 and H6) interact with
the protons on the NP ring, corresponding to the crosspeaks
at 7.0–8.0 ppm on the F2 axis and 3.7–4.0 ppm on the F1
axis, as it was also observed in the case of the NP-acrylate/
b-CD system. The crosspeaks at approximately 1.5 ppm on
the F2 axis and 3.7–4.0 ppm on the F1 axis shows that
inclusion of the aliphatic undec-10-enoate chain into the
CD cavity also occurs. The intensity of the crosspeaks
arising from the undec-10-enoate interaction is observed to
be stronger than that of the NP interaction. The intensity of
the peaks is dependent on both the amount of a given
complex as well as the distance between the interacting
protons in space [23], hence it is difficult to predict the
ratio between CDs having included the NP and CDs having
included the undec-10-enoate. It seems very unlikely that
the same CD is able to include both at the same time. This
is supported by the findings in ITC experiments where a
reasonable fit to a 1:1 model was found. Binding constants
from ITC show a stronger binding with the 4-NP-undec-10-
enoate than with 4-NP, also supporting the findings that the
aliphatic undec-10-enoate chain plays a role in the complex
formation. The interactions between the aromatic ring and
the CD cavity are observed on protons of H3, H5 and H6
and it is not possible to determine which end of the ring is
closest to the narrow rim of the CD. In the aliphatic region
it is not possible to determine the position either, indicating
that a mixture of different configurations is present in the
solution. The presence of two different configurations of
the NP-undec-10-enoate/b-CD complex may affect the
regioselectivity of the reaction and e.g. the overall DS and
distribution of derivatives. It could from these data be
expected that the reaction between NP-undec-10-enoate
and b-CD could follow various routes due to the many
conformations found. On the contrary the analyses of the
product seem to reveal a higher selectivity in this case,
yielding high amounts of mono-substituted product as
shown by MALDI-TOF as well as LC-MS/MS, even
though more than one positional isomer was formed. One
possible speculation could be a self-inclusion mechanism
where the undec-10-enoate chain is included in the CD it is
attached to and prevents more than one substitution to take
place on each CD, but this has not been investigated
further.
Harada and co-workers reported the synthesis of b-CD-
acr to be regioselective for the C-2 secondary alcohol of the
CDs using a 3-NP-acrylate as intermediate [11], but in this
work LC-MS/MS analyses have indicated several types of
positional isomers being present in the samples.
Conclusion
Cyclodextrin monomers carrying vinyl groups have been
synthesised using a method of inclusion with nitrophenyl-
derivatives for controlling the substitution degrees. It
proved possible to achieve CD monomers of a-, b-and
c-CDs with a chain length of 2 carbons between the CD and
the double bond. Changing the NP-derivative to carry a
longer carbon chain also proved possible and products of
b-CD-pent-4-enoate plus a- and b-CD-undec-10-enoate
were obtained. In the case of b-CD-undec-10-enoate only
small amounts of substitutions higher than one was observed,
whereas more different substitutions were observed with
a-, b- and c-CD-acr.
The produced products were studied in more detail by
separation of the CD derivatives on a CDScreenÒ HPLC
column and use of MS/MS. From these studies at least two
different substitution patterns were distinguished in the
mono-substituted CD-acr1 product indicating low regiose-
lectivity of the reaction. This also seemed to be the case
with the b-CD-undec-10-enoate even though less of the
higher substitutions were found in the analyses of the
products. Studies on the complexes between NP-deriva-
tives and CDs by ITC and 2D ROESY NMR were con-
ducted to reveal differences between inclusion of 4-NP-
acrylate and 4-NP-undec-10-enoate. From ITC it was
proved that the undec-10-enoate-chain adds to the stability
of the complex between b-CD and 4-NP-undec-10-enoate,
In this work a-, b- and c-CD-acr derivatives have been
studied along with b-CD-pent and a- and b-CD-und and the
effect on the reaction of the chain length observed. From
MS and HPLC data no significant differences in distribu-
tion and number of derivatives were observed from reac-
tions with a-, b- and c-CD, respectively, even though the
ITC data reveals a rather large difference in the stability
constants. The b-CD derivatives even seem to give more of
the di- and tri- substituted products, in spite of the higher
123

J Incl Phenom Macrocycl Chem (2010) 67:303–315
315
10. Lee, M.H., Yoon, K.J., Ko, S.W.J.: Synthesis of a vinyl monomer
containing beta-cyclodextrin and crafting onto cotton fiber. Appl.
Polym. Sci. 80, 438–446 (2001)
11. Harada, A., Furue, M., Nozakura, S.: Cyclodextrin-containing
polymers. 1. Preparation of polymers. Macromolecules 9, 701–
704 (1976)
12. Lesur, D., Gassama, A., Moreau, V., Djedaini-Pilard, F., Brique,
A., Pilard, S.: Electrospray ionization mass spectrometry: a key
analytical tool for the characterization of regioselectively deriva-
tized maltooligosaccharides obtained starting from natural beta-
cyclodextrin. Rapid Commun. Mass Spectrom. 20, 747–754 (2006)
13. Zaia, J.: Mass spectrometry of oligosaccharides. Mass Spectrom.
Rev. 23, 161–227 (2004)
14. Tuting, W., Adden, R., Mischnick, P.: Fragmentation pattern of
regioselectively O-methylated maltooligosaccharides in electro-
spray ionisation-mass spectrometry/collision induced dissocia-
tion. Int. J. Mass Spectrom. 232, 107–115 (2004)
whereas the 4-NP-pent-4-enoate and 4-NP-acrylate has
more or less the same stability with b-CD as the pure 4-NP.
2D ROESY NMR confirmed the presence of an inclusion
complex of NP-derivatives and CDs in aqueous solution,
and gave indications of the 3D structure of the complex. To
sum up the results show a limited control of the substitution
reaction even though the complex formation between 4-NP
and CD was proved and studied by ITC. Only in the case of
CD-undec-10-enoate a control of the synthesis seems to be
observed; maybe due to self-inclusion leading to occupa-
tion of the cavity.
Acknowledgements The authors would like to acknowledge Lars
Duelund, University of Southern Denmark for his technical support
with ITC analyses.
15. Wiseman, T., Williston, S., Brandts, J.F., Lin, L.N.: Rapid
measurement of binding constants and heats of binding using a
new titration calorimeter. Anal. Biochem. 179, 131–137 (1989)
16. Turnbull, W.B., Daranas, A.H.: On the value of c: can low
affinity systems be studied by isothermal titration calorimetry.
J. Am. Chem. Soc. 125, 14859–14866 (2003)
References
17. Connors, K.A.J.: Population characteristics of cyclodextrin complex
stabilities in aqueous-solution. Pharm. Sci. 84, 843–848 (1995)
18. Szeman, J., Csabai, K., Kekesi, K., Szente, L., Varga, G.: Novel
stationary phases for high-performance liquid chromatography
analysis of cyclodextrin derivatives. J. Chromatogr. A 1116, 76–
82 (2006)
1. Stella, V.J., Rajewski, R.A.: Cyclodextrins: their future in drug
formulation and delivery. Pharm. Res. 14, 556–567 (1997)
2. Larsen, K.L.J.: Large cyclodextrins. Incl. Phenomena 43, 1–13
(2002)
3. Del Valle, E.M.M.: Cyclodextrins and their uses: a review. Pro-
cess Biochem. 39, 1033–1046 (2004)
19. Bartsch, H., Konig, W.A., Strassner, M., Hintze, U.: Quantitative
determination of native and methylated cyclodextrins by matrix-
assisted laser desorption/ionization time-of-flight mass spec-
trometry. Carbohydr. Res. 286, 41–53 (1996)
20. Kieken, F., West, C., Keddadouche, K., Elfakir, C., Choisnard, L.,
Geze, A., Wouessidjewe, D.J.: Characterisation of complex
amphiphilic cyclodextrin mixtures by high-performance liquid
chromatography and mass spectrometry. Chromatogr. A 1189,
385–391 (2008)
21. Lo Meo, P., D’Anna, F., Riela, S., Gruttadauria, M., Noto, R.:
Spectrophotometric study on the thermodynamics of binding of
alpha- and beta-cyclodextrin towards some p-nitrobenzene
derivatives. Org. Biomol. Chem. 1, 1584–1590 (2003)
22. Hofler, T., Wenz, G.J.: Determination of binding energies
between cyclodextrins and aromatic guest molecules by micro-
calorimetry. Incl. Phenomena 25, 81–84 (1996)
23. Hull, W.E.: Two-dimensional and related NMR methods in
structural analysis of oligo- and polysaccharides. In: Croasmun,
W.R., Carlson, R.M.K. (eds.) The Two-Dimensional NMR
Spectroscopy—Applications for Chemists and Biochemists, pp.
334–343. Wiley, New York (1994)
4. Nielsen, A.L., Madsen, F., Larsen, K.L.: Controlled release of
ibuprofen from cyclodextrin modified PVP/PEG hydrogels. Drug
Del. 16(2), 92–101 (2009)
5. Janus, L., Crini, G., El-Rezzi, V., Morcellet, M., Cambiaghi, A.,
Torri, G., Naggi, A., Vecchi, C.: New sorbents containing beta-
cyclodextrin. Synthesis, characterization, and sorption properties.
React. Funct. Polym. 42, 173–180 (1999)
6. Khan, A.R., Forgo, P., Stine, K.J., D’Souza, V.T.: Methods for
selective modifications of cyclodextrins. Chem. Rev. 98, 1977–
1996 (1998)
7. Martina, K., Trotta, F., Robaldo, B., Belliardi, N., Jicsinszky, L.,
Cravotto, G.: Efficient regioselective functionalizations of
cyclodextrins carried out under microwaves or power ultrasound.
Tetrahedron Lett. 48, 9185–9189 (2007)
8. Atsumi, M., Izumida, M., Yuan, D.Q., Fujita, K.: Selective syn-
thesis and structure determination of 6(A), 6(C), 6(E)-tri(O-sul-
fonyl)-beta-cyclodextrins. Tetrahedron Lett. 41, 8117–8120
(2000)
9. Yu, D.H., Steffensen, K., Tranholm, J., Nielsen, A.L., Wimmer,
R., Larsen, K.L.J.: Regioselective alkanoylation of cyclodextrins.
Incl. Phenomena 57, 333–338 (2007)
123