International Journal of
Molecular Sciences
Article
Determination of Melting Parameters of Cyclodextrins Using
Fast Scanning Calorimetry
Askar K. Gatiatulin * , Ivan A. Grishin, Aleksey V. Buzyurov
Marat A. Ziganshin and Valery V. Gorbatchuk
Citation: Gatiatulin, A.K.; Grishin,
I.A.; Buzyurov, A.V.;
Mukhametzyanov, T.A.; Ziganshin,
M.A.; Gorbatchuk, V.V.
Determination of Melting Parameters
of Cyclodextrins Using Fast Scanning
Calorimetry. Int. J. Mol. Sci. 2022, 23,
13120. https://doi.org/10.3390/
ijms232113120
Academic Editors: Antonino
Mazzaglia, Lajos Szente and
Francesco Trotta
Received: 30 September 2022
Accepted: 21 October 2022
Published: 28 October 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
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 (https://
creativecommons.org/licenses/by/
4.0/).
Int. J. Mol. Sci. 2022, 23, 13120. https://doi.org/10.3390/ijms232113120
, Timur A. Mukhametzyanov ,
A.M. Butlerov Institute of Chemistry, Kazan Federal University, 18 Kremlevskaya, 420008 Kazan, Russia
* Correspondence: agatiatu@kpfu.ru
Abstract: The first evidence of native cyclodextrins fusion was registered using fast scanning calorime-
try (FSC) with heating rates up to 40,000 K s−1 . The endothermal effects, detected at low heating
rates, correspond to the decomposition processes. Upon the increase of the heating rate the onset
of these effects shifts to higher temperatures, reaching a limiting value at high heating rates. The
limiting temperatures were identified as the melting points of α-, β- and γ-cyclodextrins, as the
decomposition processes are suppressed at high heating rates. For γ-cyclodextrin the fusion enthalpy
was measured. The activation energies of thermal decomposition of cyclodextrins were determined
by dependence of the observed thermal effects on heating rates from 4 K min−1 in conventional
differential scanning calorimetry to 40,000 K s−1 in FSC. The lower thermal stability and activation
energy of decomposition of β-cyclodextrin than for the other two cyclodextrins were found, which
may be explained by preliminary phase transition and chemical reaction without mass loss. The
obtained values of fusion parameters of cyclodextrins are needed in theoretical models widely used
for prediction of solubility and solution rates and in preparation of cyclodextrin inclusion compounds
involving heating.
Keywords: cyclodextrin; fusion; melting; thermal analysis; fast scanning calorimetry; chip calorimetry
1. Introduction
Native cyclodextrins (CDs) are pharmaceutically important biomolecules, which ap-
plications require a knowledge of their melting parameters. Melting points and fusion
enthalpies are essential parameters used in theoretical models for prediction of solubil-
ity [1–3] and in correlations with dissolution rates [4,5], which are key physical values for
pharmaceuticals [6,7]. CDs are used as excipients enhancing the solubility and dissolution
rate of active pharmaceutical ingredients [8,9]. So, the melting parameters of CDs may be
used to expand the existing predictive approaches also to the solubility and dissolution
rates of CDs and their inclusion compounds. The problem is in a limited thermal stability of
CDs, which does not allow to study their fusion because CDs decompose long before melt-
ing point when studied by conventional methods, e.g., by differential scanning calorimetry
(DSC) [10]. So, a special experimental technique is necessary to overcome this problem.
A state-of-the-art method for determination of melting parameters for compounds
of low thermal stability is fast scanning calorimetry (FSC) [6,11–13]. FSC was used to
determine melting properties of thermally labile biomolecules: proteins [14], peptides [15],
amino acids [5], pharmaceuticals [16,17], mono and disaccharides [1], and nucleobases [18].
This method uses special chip sensors that are both sample holders and electrical circuits
for sample heating and determination of its temperature and heat flow at scanning rates of
several thousand K per second due to the extremely low addenda heat capacity and sample
size [19]. Due to the rapid heating, the fusion may complete before the decomposition
processes of studied compounds [15,18]. Even fastest conventional DSC (Hyper-DSC by
Perkin Elmer) used to detect fusion of pharmaceuticals within separate pans [20] with
https://www.mdpi.com/journal/ijms
Int. J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 2 of 8

rates of several thousand K per second due to the extremely low addenda heat capacity
Int. J. Mol. Sci. 2022, 23, 13120 and sample size [19]. Due to the rapid heating, the fusion may complete before the2de- of 8
composition processes of studied compounds [15,18]. Even fastest conventional DSC (Hy-
per-DSC by Perkin Elmer) used to detect fusion of pharmaceuticals within separate pans
[20] with heating rates of a few hundred K min−1 is not suitable to study the melting of
heating rates of a few hundred K min−1 is not suitable to study the melting of cyclodextrins
cyclodextrins because they have too big and thermally labile molecules. So, in the present
because they have too big and thermally labile molecules. So, in the present work, FSC
work, FSC method was first used to determine melting parameters of native cyclodextrins.
method was first used to determine melting parameters of native cyclodextrins.
Our approach to the study of CDs melting properties is based on the analysis of their
Our approach to the study of CDs melting properties is based on the analysis of
thermal effects dependence on heating rate. In relatively slow thermal analysis, decompo-
their thermal effects dependence on heating rate. In relatively slow thermal analysis,
sition of cyclodextrins occurs [10]. This process for β-cyclodextrin and other carbohy-
decomposition of cyclodextrins occurs [10]. This process for β-cyclodextrin and other
drates has a significant dependence of decomposition temperature on heating rate [21].
carbohydrates has a significant dependence of decomposition temperature on heating
So,
rateconventional DSC method
[21]. So, conventional DSCwas usedwas
method in the present
used in thework
presentto determine kinetic param-
work to determine kinetic
eters of thermal
parameters decomposition
of thermal decompositionfor three native
for three CDs:
native CDs:α-,α-,
β-,β-,and
andγ-cyclodextrin.
γ-cyclodextrin. From
From
these
these data,
data,the
thecorrelation
correlationofofdecomposition
decompositiontemperature
temperaturewithwithheating
heatingrate
ratewas
wasestimated.
estimated.
FSC
FSC method
method was was applied
applied to
to find
find the
the scanning
scanning raterate above
above which
which this
this correlation
correlation breaks
breaks
resulting
resulting in relatively small further shift of the observed endothermic effect by by
in relatively small further shift of the observed endothermic effect tempera-
temperature.
ture.
SuchSuch marginal
marginal dependence
dependence on heating
on heating rate
rate is is intrinsic
intrinsic to fusion
to fusion of biomolecules
of biomolecules in
in FSC
FSC studies
studies [1,15,18].
[1,15,18]. FromFrom the onset
the onset temperature
temperature of mainof thermal
main thermal effect
effect in in thisofrange
this range of
heating
heating rates the melting points of native CDs
rates the melting points of native CDs were determined. were determined.

2.
2. Results
Resultsand
andDiscussion
Discussion
To
To study
study the
the kinetics
kinetics of
of thermal
thermal decomposition
decomposition and and fusion
fusion of
of native
native CDs,
CDs, the
thesamples
samples
of
of αCD,
αCD, βCD,
βCD, and
and γCD
γCD were
were studied
studied by methods of simultaneous TG/DSC TG/DSC with
with heating
heating
rates of
rates of 4, 10 and 30
30 K min−−11 and
K min and FSC
FSC with the rates
rates of 30–40,000KKss−−11. The TG/DSC
of 30–40,000 TG/DSC andand
FSC curves obtained
FSC obtained are
areshown
shownininFigure
Figure1 1forfor αCD
αCD andandin in
Supplementary
Supplementary Information (SI)
Information
for βCD
(SI) andand
for βCD γCD. From
γCD. Fromthese curves
these thethe
curves onset temperatures
onset temperatures To T
ofo the main
of the endothermic
main endother-
peaks
mic werewere
peaks determined for each
determined sample,
for each Table
sample, 1. 1.
Table

Figure 1. The influence of heating rates on the curves of thermal analysis of dry α-cyclodextrin: (a)
Figure 1. The influence of heating rates on the curves of thermal analysis of dry α-cyclodextrin:
simultaneous TG (solid lines) and DSC (dashed lines), (b) FSC experiment, where the cross-points
(a) simultaneous TG (solid lines) and DSC (dashed lines), (b) FSC experiment, where the cross-points
of dotted lines indicate the onset points To of the main endothermic effect of decomposition or fu-
of dotted lines indicate the onset points To of the main endothermic effect of decomposition or fusion.
sion.

The analysis of data obtained indicates that T values significantly increase with
Table 1. Onset temperatures (To) of main endothermic peako for anhydrous natural cyclodextrins at
the increase in heating rate up to 2000 K s −1 , Figure 2a. At higher heating rates β of
different heating rates (β) *.
2000–40,000 K s−1 , To values do not change significantly, Figure 2a. Such behavior can be
α-Cyclodextrin
explained by two different processes β-Cyclodextrin γ-Cyclodextrin
corresponding to To . At low heating rates, the values
β/K s−1
To correspond to theToonset
/°C points β/K s−1
of thermal To/°C
decomposition. β/Kthis
Since s−1 decomposition
To/°C is a
chemical
0.0667 process, T increases
o291 significantly
0.0667 with increasing
251 heating rate.
0.0667 At higher heating
290
rates,
0.1667 302 To values
the nearly constant indicate that267
0.1667 a fusion occurs first in cyclodextrins,
0.1667 305
preceding
0.5 their further
321 decomposition.
0.5 Thus, for α-,
293 β-, and γ-cyclodextrins,
0.5 the melting
320
and 474 ◦ C, respectively.
Tm is 507, 501,470
point500 30 These
415 values are calculated
30 as the average
405
ones for heating rates 2000–40,000 K s−1 .
 4000 505 2000 501 2000 482
8000 506 8000 491 10,000 473
16,000 506 40,000 512 40,000 466
40,000 508
Int. J. Mol. Sci. 2022, 23, 13120 * Heating rates of 0.0667, 0.1667 and 0.5 K s−1 (4, 10 and 30 K min−1, respectively) were used in3DSC
of 8
experiments, 30–40,000 K s in FSC experiments.
−1

Table The analysis

1. Onset of data (T
temperatures obtained indicates that To values significantly increase with the
o ) of main endothermic peak for anhydrous natural cyclodextrins at
increase in heating rate up
different heating rates (β) *. to 2000 K s−1, Figure 2a. At higher heating rates β of 2000–40,000
K s−1, To values do not change significantly, Figure 2a. Such behavior can be explained by
α-Cyclodextrin
two different processes corresponding β-Cyclodextrin
to To. At low heating rates, γ-Cyclodextrin
the values To corre-
spond
β/K to
s the onset T points◦ ◦ T oa/◦ chemi-
o / C of thermal β/K decomposition. Since this β/Kdecomposition is
− 1 s − 1 To/ C s − 1 C
cal process,
0.0667 T o increases significantly with increasing heating rate. At higher heating rates,
291 0.0667 251 0.0667 290
the nearly
0.1667 constant T302 o values indicate that a fusion267
0.1667 occurs first in0.1667
cyclodextrins, preceding
305
their further
0.5 decomposition.
321 Thus, for0.5 α-, β-, and 293
γ-cyclodextrins,0.5the melting point 320 Tm is
500 and 474 °C,
507, 501, 470respectively. 30 These values 415 are calculated as 30 the average405 ones for
2000 509
heating rates 2000–40,000 K s . −1 500 477 500 448
4000 505 2000 501 2000 482
Melting of αCD was revealed also by polarized light microscopy. Upon heating
8000 506 8000 491 10,000 473
above its melting point
16,000 506 of 520 °C 40,000derived from T512 o vs. β correlation and cooling to room
40,000 466
temperature
40,000 at a rate 508 of 40,000 K s−1, the sample of αCD contracts to more round shape
*remaining
Heating rates colorless andand
of 0.0667, 0.1667 crystalline,
0.5 K s−1 (4,Figure
10 and 302b, which
K min shows that
−1 , respectively) were α-cyclodextrin melts
used in DSC experiments,
− 1
withoutKdecomposition
30–40,000 s in FSC experiments. at its Tm value found.

Figure2.2.(a)
Figure (a)Correlation
Correlationofofonset
onsettemperatures
temperatures(T(T ) ofmain
o )oof mainendothermic
endothermic peak
peak with
with thethe heating
heating rate
rate β
β in
in DSCDSC
andandFSCFSC experiments
experiments for for
drydry native
native cyclodextrins
cyclodextrins (the(the
lineslines are given
are given to guide
to guide the eye);
the eye); (b)
(b) the
the microscope
microscope pictures
pictures of αCD
of αCD samplesample in polarized
in polarized lightlight before
before (upper)
(upper) andand
afterafter (lower)
(lower) melting
melting by
by heating to 520
◦ °C and cooling to room temperature with the rate
heating to 520 C and cooling to room temperature with the rate of 40,000 K s . of 40,000 −K1 s −1.

The observed
Melting of αCD differences
was revealedin melting
also points of native
by polarized CDsmicroscopy.
light are consistent with heating
Upon thermo-
dynamic
above properties
its melting of anhydrous
point ◦ CDs. The
of 520 C derived anhydrous
from To vs. βγCD has much
correlation andmore negative
cooling to roomen-
thalpy of solution
temperature in water
at a rate of 40,000 K s−1 , the sample of αCD compared
and N,N-dimethylformamide, contracts toto more
αCD and roundβCD [22].
shape
Presumingcolorless
remaining that all CDs
and have a similar
crystalline, state2b,
Figure in solution,
which showsthe higher energy of solidmelts
that α-cyclodextrin γCD
can explain
without its lower melting
decomposition at its Tpoint.
m value found.
To calculate
The observed the kinetic parameters
differences of CDs
in melting points decomposition,
of native slope of the
CDs are consistent withplot ‘lgβ vs.
thermody-
1000/Tproperties
namic o’ was determined for each
of anhydrous CDs.studied CD, Figure
The anhydrous γCD 3. has
Thismuch
correlation is linearenthalpy
more negative for DSC
of
andsolution
FSC data in water and N,N-dimethylformamide,
for heating rates below 2000 K s−1 for αCD, compared to αCD
and below 500and −1 for [22].
K sβCD βCDPre-and
suming that all CDs
γCD. According have
to the a similar state in solution,
Flynn–Wall–Ozawa the higher
method [21], energy
its slope gives of solid γCD
activation can
energy
explain
Ea of theitsinitial
lowerstage
melting point. decomposition (SI). The activation energy of this process
of thermal
is equal to 163, 105, and 160 parameters
To calculate the kinetic of CDs
kJ mol−1 for αCD, decomposition,
βCD, slope of the plot ‘lgβ vs.
and γCD, respectively.
1000/To ’ was determined for each studied CD, Figure 3. This correlation is linear for DSC
and FSC data for heating rates below 2000 K s−1 for αCD, and below 500 K s−1 for βCD and
γCD. According to the Flynn–Wall–Ozawa method [21], its slope gives activation energy
Ea of the initial stage of thermal decomposition (SI). The activation energy of this process is
equal to 163, 105, and 160 kJ mol−1 for αCD, βCD, and γCD, respectively.
Int.
Int. J.J. Mol.
Mol. Sci. 2022, 23,
Sci. 2022, 23, 13120
x FOR PEER REVIEW 44 of
of 88

Figure 3.
Figure 3. Fitting by
by Flynn–Wall–Ozawa
Flynn–Wall–Ozawamethod
methodfor
fordependence
dependenceof
ofTToo on
on heating
heating rate.
rate.

The
The TG/DSC
TG/DSC data obtained obtained also
alsoallow
allowto estimateEE
toestimate a avalue
valuefrom fromthe thethermogravimetric
thermogravimet-
data
ric data in the wide range of decomposition degree. Approximation of thesethermokinetic
in the wide range of decomposition degree. Approximation of these thermokinetic
data
data by modelmodel methods
methodsshows showsthatthatforforthe thenative
nativeCDs CDs studied
studied thethemostmost appropriate
appropriate ki-
kinetic model is CnB which corresponds
netic model is СnB which corresponds to the nth order to the nth order reaction with autocatalysis, the
the
n
corresponding
corresponding conversion functionisisf f(α)
conversion function (α) ==(1 (1−− α)α)n ··(1
(1 ++ K Kcat
catα). In this
α). In this case,
case, the the optimal
optimal
− 1 respectively.
activation energies EEaa are
activation energies are 147,
147, 177,
177, and
and 156
156 kJ kJ mol
mol−1, ,respectively.
The
The TG TG datadata were
were also
alsoanalyzed
analyzedusingusingisoconversional
isoconversionalmodel-free model-free methods:
methods: dif-
differ-
ferential Friedman and integral Flynn–Wall–Ozawa. These
ential Friedman and integral Flynn–Wall–Ozawa. These methods do not require any as- methods do not require any
assumptions
sumptions ofofthe thekinetic
kineticequation
equationother
otherthan
thanthe theArrhenius-type
Arrhenius-type temperature
temperature dependence
dependence
of
of the reaction rate [23]. In the range of conversion degrees 0.1–0.7, these methods give
the reaction rate [23]. In the range of conversion degrees 0.1–0.7, these methods give
−1 −1
average EEaa,, respectively,
average respectively,149 149and
and146146kJkJmolmol−1for αCD,169
for αCD, 169and and174 174kJkJmol
mol−1 for βCD, 154
for βCD, 154
and 156 kJ mol −1 for γCD (SI). These E values are nearly the same and, in the case of αCD
a
and 156 kJ mol for γCD (SI). These Ea values are nearly the same and, in the case of αCD
−1
and γCD, are
and γCD, are close
close to the EEaa values
to the values calculated
calculated from from DSC DSC and and FSCFSC datadata for
for the
the initial
initial stage
stage
of decomposition.
of decomposition.
The
The activation
activation energies
energies calculated
calculated fromfrom calorimetric
calorimetric and and TG TG datadata are
are close
close for
for αCD
αCD
and γCD —the difference does not exceed 10–15 kJ mol−−11 . For βCD, the difference between
and γCD —the difference does not exceed 10–15 kJ mol . For βCD, the difference−between 1 . Such
activation energies calculated by different methods is significant: 65–73 kJ mol
activation energies calculated by different methods is significant: 65–73 kJ mol−1. Such dif-
difference can be explained by the different onset temperatures observed in TG and DSC
ference can be explained by the different onset temperatures observed in TG and DSC
measurement. For βCD, the average difference between TG and DSC onset points is 33 K,
measurement. For βCD, the average difference between TG and DSC onset points is 33 K,
while as for αCD and γCD it is equal to 12 and 6 K, respectively. For βCD, earlier DSC
while as for αCD and γCD it is equal to 12 and 6 K, respectively. For βCD, earlier DSC
onset can be explained by preliminary phase transition or intramolecular reaction without
onset can be explained by preliminary phase transition or intramolecular reaction without
a change in mass. Moreover, for βCD, a phase transition is observed at 222 ◦ C (SI), which
a change in mass. Moreover, for βCD, a phase transition is observed at 222 °C (SI), which
has been observed also elsewhere [10]. All these effects can increase the energy of βCD
has been observed also elsewhere [10]. All these effects can increase the energy of βCD
thus decreasing activation energy of initial step of its main thermal degradation.
thus decreasing activation energy of initial step of its main thermal degradation.
The FSC experiment for γCD allows determination of its fusion enthalpy ∆Hm . This
The FSC experiment for γCD allows determination of its fusion enthalpy ΔHm. This
cyclodextrin has the lower melting point than the other CDs studied, Table 1, and its melt-
cyclodextrin has the lower
ing and decomposition peaksmelting point than
are separate theheating
at the other CDs ratestudied,
of 10,000TableK s−1, 1 , and its 4.
Figure melt-
So,
ing molar
the and decomposition
fusion enthalpy peaks
∆Hm are
ofseparate
γCD canatbethe heating rate
calculated. Forof this10,000
study, K wes , additionally
−1 Figure 4. So,
the molar fusion
determined enthalpy
the molar heatΔH m of γCD can be calculated. For this study, we additionally
capacity Cp,m of this cyclodextrin in the range of 80–220 ◦ C
determined the molar heat
(SI), which is needed to find its samplecapacity C p,m of this cyclodextrin in the range of 80–220 °C (SI),
mass in FSC experiment. The temperature depen-
dence of this parameter is expressed by in
which is needed to find its sample mass theFSC experiment.
equation Cp,m (JThe moltemperature
−1 K−1 ) = 781 dependence
+ 9.13T −
of this parameter
2 is expressed by the equation C p,m (J mol−1 K−1) = 781−+
0.0043T . At 298 K, this equation gives the value of Cp,m = 1556 J mol K , which 1 9.13T
− 1 − 0.0043T²
is in.
aAtgood
298 K, this equation with gives the value determined
of Cp,m = 1556value J molof K , which
= 1568isJ in a good
−1 K−agree-
−1 −1 1 [24].
agreement the previously Cp,m mol
ment with the previously determined value of C p,m = 1568
Two runs of FSC experiments for γCD at a rate of 10,000 K s give molar fusion en- J mol −1− K1−1 [24]. Two runs of FSC

experiments
thalpy of ∆Hfor γCD at a rate of−10,000
m = 221 ± 9 kJ mol
1 (SI). KThe s−1corresponding
give molar fusion enthalpy
molar fusionofentropy
ΔHm = 221±9 of γCD kJ
mol −1 (SI). The ◦ corresponding
(at Tm = 474 C) is ∆Sm = 296 ± 12 J mol K . molar fusion− 1 entropy
− 1 of γCD (at T m = 474 °C) is ΔS m = 296±12
J mol−1 K−1.
Int. J. Mol. Sci. 2022, 23, x FOR PEER REVIEW 5 of 8
Int. J. Mol. Sci. 2022, 23, 13120 5 of 8

K ss−
Figure 4. FSC curve for γ-cyclodextrin at heating rate of 10,000 K −11
..

3. Materials and
3. Materials and Methods
Methods
3.1. Materials
3.1. Materials
α-Cyclodextrin and γ-cyclodextrin were obtained commercially from Sigma-Aldrich
α-Cyclodextrin and γ-cyclodextrin were obtained commercially from Sigma-Aldrich
with Cat. Nos. 28705, 779431, respectively. β-Cyclodextrin was obtained from ICN, Cat.
with Cat. Nos. 28705, 779431, respectively. β-Cyclodextrin was obtained from ICN, Cat.
No. 190053.
No. 190053.
3.2. Simultaneous TG/DSC Experiment
3.2. Simultaneous TG/DSC Experiment
Before the experiment, all CD samples were dried at 140 ◦ C and 100 Pa in a vacuum
ovenBefore
for 8 h.the
The experiment, all CD samples
residual hydration were dried
of CD samples wasatno
140more
°C andthan 100
2%Pa in or
wt. a vacuum
1.1 mol
oven for 8 h. The residual hydration
water per mol CD according to TG data. of CD samples was no more than 2% wt. or 1.1 mol
waterTheperdevice
mol CD according to TG data.
of simultaneous thermogravimetry and differential scanning calorimetry
The device
(TG/DSC) Netzsch of simultaneous
STA 449 C Jupiter thermogravimetry and differential
was used to determine scanning calorimetry
thermal decomposition curves
(TG/DSC)
of Netzsch STAwith
dried cyclodextrins 449heating
C Jupiter was
rates of used
4, 10, to
anddetermine
30 K minthermal
− 1 in andecomposition
argon flow of
curves
75 mL min −1 . Incyclodextrins
of dried this experiment, withthe
heating rates of(10–15
CD samples 4, 10, and
mg)30 K min
were −1 in an argon flow
studied in aluminum
of 75 mL (40
crucibles minµL)
−1 . In thislids
with experiment, the CD
having 3 holes of samples
0.5 mm in(10–15 mg) Before
diameter. were studied
heating,inthe
aluminum
samples
crucibles (40 μL) with lids having 3 holes of 0.5 mm in diameter. Before
were purged with argon at room temperature inside the device until the constant weight. heating, the sam-
ples were purged with argon at room temperature inside the device until the constant
3.3. Model-Free Methods of Thermokinetic Analysis
weight.
Two model-free methods of Friedman [25] and Ozawa–Flynn–Wall [26,27] were used
3.3.approximation
for Model-Free Methods of theof experimental
Thermokinetic dataAnalysis
according to the recommendation of Interna-
tionalTwo
Confederation for Thermal Analysis
model-free methods of Friedman [25] and and Calorimetry [28].
Ozawa–Flynn–Wall [26,27] were used
In Friedman method, for a linear non-isothermal program,
for approximation of the experimental data according to the recommendation the next equation is used:
of Interna-
tional Confederation for Thermal Analysis and Calorimetry [28].
In Friedman method, i ·(dα/dT)
ln [βfor a linearα,inon-isothermal − Ea /RTthe
] = ln A + ln f (α)program, α,i next equation is used:

where Ea is activation energy, is heating

ln [βiβ·(dα/dT) ln A +i marks
α,i] = rate, ln f(α) −anEaindividual
/RTα,i heating rate, A is
pre-exponential factor and α is the extent of conversion. The Ea and pre-exponential factor
where Ea is activation energy, β is heating rate, i marks an individual heating rate, A is
are calculated from the slope of the plots of ln (dα/dT) vs. 1/T.
pre-exponential
The Flynn−factor
Wall−and α is method
Ozawa the extent is of conversion.method
a model-free The Ea andthat pre-exponential factor
requires the detection
are calculated from the slope of the plots of ln (dα/dT) vs. 1/T.
of temperatures corresponding to isoconversional values of α from experiments at different
Therates
heating Flynn−Wall−Ozawa
β [23]: method is a model-free method that requires the detection
of temperatures correspondinglntoβisoconversional
= 5.523 − 1.052(E values of α from experiments at differ-
a /RT)
ent heating rates β [23]:
Thus, the plot of ln β vs 1/T gives straight line with slope −1.052(Ea /R).
ln β = 5.523 − 1.052(Ea/RT)
3.4. DSC Experiment
Thus, the plot of ln β vs 1/T gives straight line with slope −1.052(Ea/R).
The molar heat capacity Cp,m of dry γ-cyclodextrin was measured using Netzsch
DSC204 F1 Phoenix differential scanning calorimeter. This procedure using sapphire disk
as a standard sample was described in detail earlier [12].
Int. J. Mol. Sci. 2022, 23, 13120 6 of 8

3.5. Fast Scanning Calorimetry

FSC experiments were performed using Flash DSC2+ (Mettler Toledo, Switzerland)
with MultiSTAR UFH1 sensors. Before use, each UFH1 sensor was conditioned and cor-
rected according to the procedure defined by the manufacturer. The measured temperature
of the sensors was calibrated using organic compounds having well known melting tem-
peratures as described elsewhere [29]. In FSC experiment, crystal aggregates with a total
mass of 20–50 ng were placed at the center of sensor. Temperature scan of these samples
was performed in argon dynamic atmosphere with 80 ml/min flow rate.
To get rid of the dehydration thermal effect, which was observed elsewhere at tem-
peratures up to 160 ◦ C [30–32], all CD samples were preheated directly on FSC chip to a
temperature 200 ◦ C in argon flow. Such heating was repeated several times until the heat
flow in FSC curves reached a constant level.
Activation energies of initial stage of CDs decomposition were calculated by Flynn–
Wall–Ozawa method described in detail for other carbohydrate biopolymers [21].
The molar fusion enthalpy of γCD was determined using FSC as described else-
where [11]. In this experiment, the sample of γCD was cyclically heated and cooled 5 times
in the range of 70–200 ◦ C before melting. The heating and cooling curves in these runs
and the molar heat capacity Cp,m of γCD from the DSC experiment were used to calculate
the mass of sample on FSC chip. The molar enthalpy of melting was calculated from the
thermal effect in FSC curve for the same sample heated up to 700 ◦ C.

4. Conclusions
The melting points of thermally unstable α-, β-, and γ-cyclodextrin and the kinetic
parameters of their thermal decomposition were determined using fast scanning calorime-
try, which gives a possibility to shift decomposition temperature of thermally unstable
compound above its fusion process using very fast heating rates. The same approach may
be extended to crystalline inclusion complexes of native cyclodextrins with various medical
drugs. β-Cyclodextrin was found to have a lower activation energy of decomposition what
may be explained by preliminary phase transition and the following chemical reaction
without loss of mass. Such specific feature of β-cyclodextrin thermal stability must be
considered during preparation of its complexes involving heating.
The determined fusion enthalpy and entropy of γ-cyclodextrin may be used to check
the theoretical models predicting these properties for organic compounds from structural
parameters of their molecules. Melting points and enthalpy obtained in this work are
needed for models predicting aqueous solubility of CDs which can be helpful for their
applications as excipients enhancing the solubility and dissolution rate of active pharma-
ceutical ingredients.

Supplementary Materials: The following supporting information (FSC and TG/DSC curves, thermoki-
netic analysis, molar heat capacity and ∆Hm calculations for γ-cyclodextrin) can be downloaded at:
https://www.mdpi.com/article/10.3390/ijms232113120/s1. Reference [21] is cited in the supple-
mentary materials.
Author Contributions: A.K.G. conceptualization, formal analysis, writing—original draft; I.A.G.
investigation; A.V.B. investigation; M.A.Z. validation, methodology; T.A.M. supervision, writing—
review and editing, V.V.G. supervision, writing—review and editing. All authors have read and
agreed to the published version of the manuscript.
Funding: The work was supported by the RSF grant No. 22-23-00367.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.
Int. J. Mol. Sci. 2022, 23, 13120 7 of 8

References
1. Chua, Y.Z.; Do, H.T.; Kumar, A.; Hallermann, M.; Zaitsau, D.; Schick, C.; Held, C. The Melting Properties of D-α-Glucose,
D-β-Fructose, D-Sucrose, D-α-Galactose, and D-α-Xylose and Their Solubility in Water: A Revision. Food Biophys. 2022, 17,
181–197. [CrossRef]
2. Jain, N.; Yalkowsky, S.H. Estimation of the Aqueous Solubility I: Application to Organic Nonelectrolytes. J. Pharm. Sci. 2001, 90,
234–252. [CrossRef]
3. Manson, A.; Sefcik, J.; Lue, L. Temperature Dependence of Solubility Predicted from Thermodynamic Data Measured at a Single
Temperature: Application to α, β, and γ-Glycine. Cryst. Growth Des. 2022, 22, 1691–1706. [CrossRef] [PubMed]
4. Yoshihashi, Y.; Kitano, H.; Yonemochi, E.; Terada, K. Quantitative Correlation between Initial Dissolution Rate and Heat of Fusion
of Drug Substance. Int. J. Pharm. 2000, 204, 1–6. [CrossRef]
5. Rimer, J.D.; Trofymluk, O.; Navrotsky, A.; Lobo, R.F.; Vlachos, D.G. Kinetic and Thermodynamic Studies of Silica Nanoparticle
Dissolution. Chem. Mater. 2007, 19, 4189–4197. [CrossRef]
6. Do, H.T.; Chua, Y.Z.; Kumar, A.; Pabsch, D.; Hallermann, M.; Zaitsau, D.; Schick, C.; Held, C. Melting Properties of Amino Acids
and Their Solubility in Water. RSC Adv. 2020, 10, 44205–44215. [CrossRef]
7. Wyttenbach, N.; Niederquell, A.; Kuentz, M. Machine Estimation of Drug Melting Properties and Influence on Solubility
Prediction. Mol. Pharm. 2020, 17, 2660–2671. [CrossRef]
8. Saokham, P.; Muankaew, C.; Jansook, P.; Loftsson, T. Solubility of Cyclodextrins and Drug/Cyclodextrin Complexes. Molecules
2018, 23, 1161. [CrossRef]
9. Viernstein, H.; Wolschann, P. Cyclodextrin Inclusion Complexation and Pharmaceutical Applications. ScienceAsia 2020, 46, 254.
[CrossRef]
10. Hădărugă, N.G.; Bandur, G.N.; David, I.; Hădărugă, D.I. A Review on Thermal Analyses of Cyclodextrins and Cyclodextrin
Complexes. Environ. Chem. Lett. 2019, 17, 349–373. [CrossRef]
11. Abdelaziz, A.; Zaitsau, D.H.; Mukhametzyanov, T.A.; Solomonov, B.N.; Cebe, P.; Verevkin, S.P.; Schick, C. Melting Temperature
and Heat of Fusion of Cytosine Revealed from Fast Scanning Calorimetry. Thermochim. Acta 2017, 657, 47–55. [CrossRef]
12. Safiullina, A.S.; Buzyurov, A.V.; Ziganshina, S.A.; Gerasimov, A.V.; Schick, C.; Gorbatchuk, V.V.; Ziganshin, M.A. Using Fast
Scanning Calorimetry to Study Solid-State Cyclization of Dipeptide L-Leucyl-L-Leucine. Thermochim. Acta 2020, 692, 178748.
[CrossRef]
13. Gabdulkhaev, M.N.; Ziganshin, M.A.; Buzyurov, A.V.; Schick, C.; Solovieva, S.E.; Popova, E.V.; Gubaidullin, A.T.; Gorbatchuk,
V.V. Smart Control of Calixarene Polymorphic States. CrystEngComm 2020, 22, 7002–7015. [CrossRef]
14. Cebe, P.; Thomas, D.; Merfeld, J.; Partlow, B.P.; Kaplan, D.L.; Alamo, R.G.; Wurm, A.; Zhuravlev, E.; Schick, C. Heat of Fusion of
Polymer Crystals by Fast Scanning Calorimetry. Polymer 2017, 126, 240–247. [CrossRef]
15. Do, H.T.; Chua, Y.Z.; Habicht, J.; Klinksiek, M.; Hallermann, M.; Zaitsau, D.; Schick, C.; Held, C. Melting Properties of Peptides
and Their Solubility in Water. Part 1: Dipeptides Based on Glycine or Alanine. RSC Adv. 2019, 9, 32722–32734. [CrossRef]
16. Corvis, Y.; Wurm, A.; Schick, C.; Espeau, P. Vitreous State Characterization of Pharmaceutical Compounds Degrading upon
Melting by Using Fast Scanning Calorimetry. J. Phys. Chem. B 2015, 119, 6848–6851. [CrossRef]
17. Kons, A.; Mishnev, A.; Mukhametzyanov, T.A.; Buzyurov, A.V.; Lapuk, S.E.; Bērzin, š, A. Hexamorphism of Dantrolene: Insight
into the Crystal Structures, Stability, and Phase Transformations. Cryst. Growth Des. 2021, 21, 1190–1201. [CrossRef]
18. Abdelaziz, A.; Zaitsau, D.H.; Kuratieva, N.V.; Verevkin, S.P.; Schick, C. Melting of Nucleobases. Getting the Cutting Edge of
“Walden’s Rule”. Phys. Chem. Chem. Phys. 2019, 21, 12787–12797. [CrossRef]
19. Mathot, V.; Pyda, M.; Pijpers, T.; Vanden Poel, G.; van de Kerkhof, E.; van Herwaarden, S.; van Herwaarden, F.; Leenaers, A.
The Flash DSC 1, a Power Compensation Twin-Type, Chip-Based Fast Scanning Calorimeter (FSC): First Findings on Polymers.
Thermochim. Acta 2011, 522, 36–45. [CrossRef]
20. Petitprez, J.; Legrand, F.-X.; Tams, C.; Pipkin, J.D.; Antle, V.; Kfoury, M.; Fourmentin, S. Huge Solubility Increase of Poorly
Water-Soluble Pharmaceuticals by Sulfobutylether-β-Cyclodextrin Complexation in a Low-Melting Mixture. Environ. Chem. Lett.
2022, 20, 1561–1568. [CrossRef]
21. Thomas, A.; Moinuddin, K.; Tretsiakova-McNally, S.; Joseph, P. A Kinetic Analysis of the Thermal Degradation Behaviours of
Some Bio-Based Substrates. Polymers 2020, 12, 1830. [CrossRef] [PubMed]
22. Danil de Namor, A.F.; Traboulssi, R.; Lewis, D.F.V. Host Properties of Cyclodextrins towards Anion Constituents of Antigenic
Determinants. A Thermodynamic Study in Water and in N,N-Dimethylformamide. J. Am. Chem. Soc. 1990, 112, 8442–8447.
[CrossRef]
23. Venkatesh, M.; Ravi, P.; Tewari, S.P. Isoconversional Kinetic Analysis of Decomposition of Nitroimidazoles: Friedman Method vs
Flynn–Wall–Ozawa Method. J. Phys. Chem. A 2013, 117, 10162–10169. [CrossRef] [PubMed]
24. Briggner, L.-E.; Wadsö, I. Heat Capacities of Maltose, Maltotriose, Maltotetrose and α-, β-, and γ-Cyclodextrin in the Solid State
and in Dilute Aqueous Solution. J. Chem. Thermodyn. 1990, 22, 1067–1074. [CrossRef]
25. Friedman, H.L. Kinetics of Thermal Degradation of Char-Forming Plastics from Thermogravimetry. Application to a Phenolic
Plastic. J. Polym. Sci. Part C Polym. Symp. 1964, 6, 183–195. [CrossRef]
26. Flynn, J.H.; Wall, L.A. General Treatment of the Thermogravimetry of Polymers. J. Res. Natl. Bur. Stand. Sect. A Phys. Chem. 1966,
70A, 487–523. [CrossRef]
27. Ozawa, T. Estimation of Activation Energy by Isoconversion Methods. Thermochim. Acta 1992, 203, 159–165. [CrossRef]
Int. J. Mol. Sci. 2022, 23, 13120 8 of 8

28. Gatiatulin, A.K.; Osel’skaya, V.Y.; Ziganshin, M.A.; Gorbatchuk, V.V. Smart Control of Guest Inclusion by α-Cyclodextrin Using
Its Hydration History. RSC Adv. 2019, 9, 37778–37787. [CrossRef]
29. Buzyurov, A.V.; Nagrimanov, R.N.; Zaitsau, D.H.; Mukhametzyanov, T.A.; Abdelaziz, A.; Solomonov, B.N.; Schick, C. Application
of the Flash DSC 1 and 2+ for Vapor Pressure Determination above Solids and Liquids. Thermochim. Acta 2021, 706, 179067.
[CrossRef]
30. Gatiatulin, A.K.; Osel’skaya, V.Y.; Ziganshin, M.A.; Gorbatchuk, V.V. Size Exclusion Effect in Binary Inclusion Compounds of
α-Cyclodextrin. Phys. Chem. Chem. Phys. 2018, 20, 26105–26116. [CrossRef]
31. Gorbatchuk, V.V.; Gatiatulin, A.K.; Ziganshin, M.A.; Gubaidullin, A.T.; Yakimova, L.S. Unusually High Efficiency of β-
Cyclodextrin Clathrate Preparation by Water-Free Solid-Phase Guest Exchange. J. Phys. Chem. B 2013, 117, 14544–14556.
[CrossRef]
32. Li, H.; Zhang, G.; Wang, W.; Chen, C.; Jiao, L.; Wu, W. Preparation, Characterization, and Bioavailability of Host-Guest Inclusion
Complex of Ginsenoside Re with Gamma-Cyclodextrin. Molecules 2021, 26, 7227. [CrossRef]