Permethylated β-cyclodextrin/pesticide complexes: X-ray structures and thermogravimetric assessment of kinetic parameters for complex dissociation

Article abstract of DOI:10.1007/s10847-012-0145-5

The X-ray crystal structures of the inclusion complexes formed between three pesticides (two organophosphorus insecticides and one chloroacetanilide herbicide) and permethylated β-cyclodextrin (TRIMEB) are reported. The complexes TRIMEB-fenitrothion (1), TRIMEB-fenthion (2) and TRIMEB-acetochlor (3) are members of a commonly occurring isostructural series. The mode of inclusion of the two organophosphate insecticides is very similar, while the acetochlor molecule, which is structurally quite distinct from the two insecticide molecules, adopts a somewhat different position within the TRIMEB cavity. In addition to the structural elucidation of these complexes, their thermal behaviour was investigated using isothermal and non-isothermal thermogravimetry. The isothermal results showed that the dissociation of the guest molecules from the TRIMEB cavities can best be described by two mechanisms, namely a first-order reaction model and a three-dimensional diffusion model. Both the isothermal and non-isothermal methods allowed the determination of the activation energies of the guest loss process for each complex.

Full text of DOI:10.1007/s10847-012-0145-5

J Incl Phenom Macrocycl Chem (2013) 75:47–56
DOI 10.1007/s10847-012-0145-5
ORIGINAL ARTICLE
Permethylated b-cyclodextrin/pesticide complexes: X-ray
structures and thermogravimetric assessment of kinetic
parameters for complex dissociation
•
Dyanne L. Cruickshank Natalia M. Rougier
•
•
Vaughan J. Maurel Rita H. de Rossi
•
•
•
Elba I. Bujan Susan A. Bourne Mino R. Caira
´
Received: 7 February 2012 / Accepted: 7 March 2012 / Published online: 13 April 2012
Ó Springer Science+Business Media B.V. 2012
Abstract The X-ray crystal structures of the inclusion
complexes formed between three pesticides (two organo-
phosphorus insecticides and one chloroacetanilide herbicide)
and permethylated b-cyclodextrin (TRIMEB) are reported.
The complexes TRIMEB–fenitrothion (1), TRIMEB–fen-
thion (2) and TRIMEB–acetochlor (3) are members of a
commonly occurring isostructural series. The mode of
inclusion of the two organophosphate insecticides is very
similar, while the acetochlor molecule, which is structurally
quite distinct from the two insecticide molecules, adopts a
somewhat different position within the TRIMEB cavity. In
addition to the structural elucidation of these complexes, their
thermal behaviour was investigated using isothermal and
non-isothermal thermogravimetry. The isothermal results
showed that the dissociation of the guest molecules from the
TRIMEB cavities can best be described by two mechanisms,
namely a first-order reaction model and a three-dimensional
diffusion model. Both the isothermal and non-isothermal
methods allowed the determination of the activation energies
of the guest loss process for each complex.
Keywords Pesticides ꢀ Inclusion complexes ꢀ
Permethylated b-cyclodextrin ꢀ Single crystal X-ray
diffraction ꢀ Thermal decomposition kinetics
Introduction
The ability of cyclodextrins (CDs) to form inclusion
complexes with various guest molecules has been studied
extensively over the last few decades especially within the
pharmaceutical industry, as well as the food, cosmetic and
biotechnology industries, and in the field of analytical
chemistry [1]. In 1985 Szejtli [2] reviewed the subject of
CDs and pesticides, stating that within the next decade,
there would be a rapid development in the application of
CDs in pesticide formulations accompanying the decrease
in the price of CDs and the increasing rate of CD pro-
duction. Since then the focus has been on studying the
interactions between pesticides and CDs in solution as well
as in the solid state and evaluating the possible benefits that
could be conferred on pesticides upon their complexation
with biodegradable host molecules such as CDs. In general,
the undesirable physicochemical properties associated with
pesticides include their poor aqueous solubility, chemical,
thermal and physical instability, high volatility, high soil
mobility and poor wettability [3].
Electronic supplementary material The online version of this
article (doi:10.1007/s10847-012-0145-5) contains supplementary
material, which is available to authorized users.
The three pesticides selected for this study are the
insecticides fenitrothion [O,O-dimethyl O-(3-methyl-4-
nitrophenyl) phosphorothioate] and fenthion [O,O-dime-
thyl O-[3-methyl-4-(methylthio)phenyl] phosphorothioate],
and the herbicide acetochlor [2-chloro-N-(ethoxymethyl)-
N-(2-ethyl-6-methylphenyl)acetamide] (Fig. 1). These com-
pounds are oils at room temperature and are currently registered
for use in several countries [4]. The reduced chemical and
thermal stability of these pesticides, as well as the difficulties
D. L. Cruickshank ꢀ V. J. Maurel ꢀ S. A. Bourne ꢀ
M. R. Caira (&)
Department of Chemistry, University of Cape Town,
Rondebosch 7701, South Africa
e-mail: mino.caira@uct.ac.za
´
N. M. Rougier ꢀ R. H. de Rossi ꢀ E. I. Bujan
´
´
Instituto de Investigaciones en Fısico Quımica de Cordoba
(INFIQC), Departamento de Quımica Organica, Facultad de
´
´
´
´
Ciencias Quımicas, Universidad Nacional de Cordoba, Ciudad
Universitaria, X5000HUA Cordoba, Argentina
´
´
123

48
J Incl Phenom Macrocycl Chem (2013) 75:47–56
Single crystals of each inclusion complex were prepared
using the method of co-precipitation. Preparative details for
TRIMEB–fenitrothion (1) were reported previously [8].
For the preparation of the inclusion complex TRIMEB–
fenthion (2), 119 mg of TRIMEB (0.083 mmol) was
weighed into a vial and 2 cm3 of distilled water was added.
An equimolar quantity (23 mg, 0.083 mmol) of the insec-
ticide fenthion was added and the suspension was stirred
for 4 h between 0 and 4 °C. The final clear solution was
then filtered through a 0.45 lm filter into a clean vial and
placed in an oven at 60 °C. The TRIMEB–acetochlor (3)
inclusion complex was prepared by dissolving 106 mg
(0.074 mmol) of the host in 5 cm3 of distilled water at
room temperature, and adding to the solution an equimolar
quantity of acetochlor (20 mg, 0.074 mmol). The resulting
suspension was stirred for 8 h and filtered into a clean vial
which was also placed in an oven set at 60 °C. Single
crystals of 2 and 3 appeared within 24 h.
(b)
(a)
O
O
S
S
P
P
O
O
O2N
O
S
O
(c)
O
N
O
Cl
Fig. 1 The chemical structures of a fenitrothion, b fenthion and
c acetochlor
associated with their handling and transportation, make them
ideal candidates for solid-state formulation prior to application.
Previous studies of the stability of the organophosphate
insecticides fenitrothion and fenthion have shown that the
native CDs have a catalytic effect on the photodegradation
rate of these pesticides in humic water [5]. However, other
investigations have shown that CDs induce an inhibitory
effect on the degradation of fenitrothion in basic media
[6, 7]. No studies on the complexation of the chloroace-
tanilide herbicide acetochlor with CDs in the solid state or
in solution have hitherto appeared in the literature.
The samples used for the solid-state kinetic studies were
prepared by mechanical grinding of equimolar quantities of
the host and guest in a mortar for 30 min in the absence of
water. The mass of the guest was kept constant at 20 mg and
the mass of TRIMEB varied depending on the molecular
weight of each guest compound. The particle size distribu-
Recently we reported the ability of fenitrothion to form
solid inclusion complexes with both hexakis(2,3,6-tri-O-
methyl)-a-CD (TRIMEA) and heptakis(2,3,6-tri-O-methyl)
-b-CD (TRIMEB) [8]. We have since generated two additional
TRIMEB complexes with guests fenthion and acetochlor
which are isostructural with the TRIMEB–fenitrothion com-
plex. In this paper, the similarities and differences amongst
these three structures will be discussed and an attempt made to
reconcile the crystal structures and the results obtained from
the isothermal and non-isothermal kinetic experiments. The
activation energies of complex dissociation obtained from the
solid-state kinetic experiments will provide an indication of the
thermal stability conferred upon the guest molecules by their
encapsulation within CD molecules.
tions of the samples were measured using a MultisizerTM
3
Coulter Counter. The mechanical co-grinding method yiel-
ded samples with particle diameters in the range 20–45 lm,
with a mean diameter of ca. 25 lm.
Single crystal X-ray diffraction
X-ray structural elucidation of complex 1 was described
earlier [8]. Intensity data-collection for complexes 2 and 3
were performed on a Bruker KAPPA APEX II DUO dif-
fractometer. The crystals were maintained at 100(2) K
throughout the data-collection using Cryostream coolers
(Oxford Cryosystems UK). Unit cell refinements and data
reductions were performed using the program SAINT [9].
For both complexes 2 and 3, the Laue system was found to be
mmm, indicating the orthorhombic crystal system and the
common space group P212121 was uniquely identified from
systematic absences. Data were corrected for Lorentz-
polarization effects and for absorption (program SADABS
[10]). Structure solution for complexes 2 and 3 was achieved
by using the host atomic coordinates of the isostructural
complexes TRIMEB–p-iodophenol [11] and TRIMEB–ethyl
laurate [12] respectively, as trial models. All non-hydrogen
atoms, except O6, C7, C8 and C9 of each methyl glucose
unit, served as the input model. The structures were refined
with SHELXH-97 [13] and successive difference Fourier
maps revealed the methoxyl groups and guest molecules. All
non-hydrogen atoms of complex 2 were refined anisotropi-
cally except for one of the primary methoxyl carbon atoms of
Experimental
Materials and solid complex preparation
The host compound heptakis(2,3,6-tri-O-methyl)-b-CD
(TRIMEB) was purchased from Cyclolab (Budapest, Hun-
gary) and was used as received. Fenitrothion was isolated as
a yellow oil from a commercial sample of Sumithion
(Sumitomo Chemical, NY, USA) by column chromatogra-
phy over silica gel and was characterized by H and 31P
1
NMR spectroscopy and GC–MS. The agrochemicals fen-
thion and acetochlor were obtained from Sigma Aldrich
(Germany) and ChemService (USA) respectively.
123

J Incl Phenom Macrocycl Chem (2013) 75:47–56
49
the host, which had a reasonable isotropic thermal parameter
but an unacceptable thermal ellipsoid when refined aniso-
tropically. For complex 3, one primary methoxyl group of the
host molecule was disordered over two positions, as were
atoms C17A and Cl18 of the guest molecule. This disorder
was modelled by allowing the two components in each case
to refine isotropically with site-occupancy factors x and 1 -
x. The final value of x refined to 0.65 and 0.57 for the major
components of the host and guest atoms respectively. All
ordered atoms of complex 3 were refined anisotropically.
Although some of the guest H atoms were apparent in the
difference Fourier maps for complexes 2 and 3 all the H
atoms were added in idealised positions in a riding model
with Uiso values 1.2–1.5 times those of their parent atoms.
is evident from the close similarity in their unit cell
dimensions and the identical space group (P212121) in
which they crystallize. The X-ray crystal structure of 1 has
been described in detail previously [8]; however, for
comparison the monomeric units of complexes 1, 2 and 3
are shown in Fig. 2, which views them from a common
direction. The dimethyl phosphorothioate units of fenitro-
thion and fenthion in complexes 1 and 2 are situated close
to the primary rims of the TRIMEB cavities which are
blocked by several methoxyl groups in each case. The
3-methyl and 4-methylsulfanyl groups positioned on the
aromatic ring in complex 2 protrude from the secondary
rim of the TRIMEB cavity into the interstitial spaces
between neighbouring TRIMEB molecules. This particular
location of the aromatic substituents is similar to that in
complex 1.
Powder X-ray diffraction (PXRD)
When compared with the modes of inclusion of the
organophosphate insecticide guest molecules, the aceto-
chlor molecule assumes a different position and orientation
within the TRIMEB cavity (Fig. 2c). The aromatic moiety
is fully encapsulated by the host molecule, while the ethyl
substituent and chloro-ethoxymethyl acetamide group of
acetochlor extend from the secondary rim of the TRIMEB
cavity. The primary rim is once again ‘sealed off’ by the
primary methoxyl groups which fold over giving the
TRIMEB host molecule a cup-like shape. There are no
hydrogen bonds between the acetochlor and TRIMEB
molecules. Several intramolecular and intermolecular host–
host interactions of the type C–HꢀꢀꢀO exist in all three
complex structures. These hydrogen bonds are essential in
maintaining the overall framework of the crystal structure.
The X-ray structures of 2 and 3 provide the first accurate
representations of the fenthion and acetochlor molecules in
the solid state. It is important to note the geometry of each
guest. The dimethyl phosphorothioate unit of fenthion
adopts the usual tetrahedral conformation and the 4-meth-
ylsulfanyl group displays a slight twist out of the plane of
the aromatic ring (the dihedral angle C–C-S–C is 24.1(7)°).
For acetochlor, the angle between the least-squares plane
through the acetamide group (C2N–C=O) and the phenyl
ring plane is 85.4(3)°. This value is similar to that observed
in the crystal structure of an intermediate used for the
synthesis of acetochlor, namely 78.0(3)° [14]. All other
geometrical parameters for the fenthion and acetochlor
molecules are in the expected ranges.
PXRD patterns were recorded at room temperature using a
Huber Imaging Plate Guinier Camera 670 with nickel-
˚
filtered CuKa1 radiation (k = 1.5405981 A) produced at
40 kV and 20 mA by a Philips PW1120/00 generator. A 2h
range of 4–100° was used with a step size of 0.005° 2h. The
samples were exposed to the X-ray beam for 30 min with
10 multi-scans to collect the data.
Thermal analysis
Thermogravimetric analysis (TGA) and differential scanning
calorimetric (DSC) measurements were performed using TA
instruments (TA-Q500 and DSC-Q200) with Universal
Analysis 2000 software. The TG instrument operated with
dry nitrogen purge gas flowing at a rate of 60 cm3 min-1
while the DSC operated with a flow rate of 50 cm3 min-1
.
For consistency, the sample size for each isothermal and non-
isothermal TG run was kept small and relatively constant
(1.77–1.94 mg). For each TG experiment the complex
material (1, 2 or 3) was placed in an open aluminium pan and
a specific temperature program was applied. DSC measure-
ments were conducted in closed aluminium pans.
Isothermal TG runs were performed by heating the
sample at a rate of 30 or 40 K min-1 to a specific tem-
perature which was then maintained for 300 min. The non-
isothermal (dynamic) TG experiments were performed by
heating the powdered samples from 20 to 300 °C at dif-
ferent heating rates (1.0, 2.5, 5.0, 7.5 and 10.0 K min-1).
Geometrical parameters for the host molecule in com-
plexes 2 and 3 are comparable to those observed in com-
plex 1 [8] and in several other isostructural TRIMEB
complexes [15]. Figure 3 illustrates the packing arrange-
ment for complex 2 as representative. The complex units
pack in a screw-channel mode and are arranged in a head-
to-tail manner parallel to the b-axis. This diagram shows
that the guest molecules are isolated from one another. The
Results and discussion
Single crystal X-ray structures
Table 1 lists crystal data and refinement details for com-
plexes 1, 2 and 3. The three complexes are isostructural as
123

50
J Incl Phenom Macrocycl Chem (2013) 75:47–56
Table 1 Crystal data and
structure refinement for the
complexes 1 [8], 2 and 3
1
2
3
Chemical formula
Formula weight
Crystal system
Unit cell constants
Space group
C63H112O35ꢀC9H12NO5PS C63H112O35ꢀC10H15O3PS2 C63H112O35ꢀC14H20ClNO2
1706.75
1707.84
1699.29
Orthorhombic
Orthorhombic
Orthorhombic
P212121
15.1588(4)
21.1279(4)
27.5575(7)
90
P212121
15.084(2)
21.164(3)
27.491(3)
90
P212121
14.728(2)
21.542(3)
27.604(4)
90
˚
a (A)
˚
b (A)
˚
c (A)
a = b = c (°)
3
Volume (A )/Z
˚
8825.9(4)/4
1.284
8,776(2)/4
1.293
8,758(2)/4
1.289
Density(calc)
(g cm-3
)
Temperature (K)
173(2)
100(2)
100(2)
˚
Wavelength (A)
0.71073
0.71073
0.71073
Dimensions (mm) 0.24 9 0.13 9 0.11
0.30 9 0.12 9 0.05
1.48–24.75
0.30 9 0.27 9 0.07
1.83–25.87
h-range collection 2.69–25.70
(°)
Index range
-18 B h B 18
-25 B k B 25
-33 B l B 33
121,751
-9 B h B 17
-25 B k B 25
-32 B l B 32
73,115
-17 B h B 15
-26 B k B 22
-29 B l B 33
55,784
Reflections
collected
Independent
reflections
16,732
11,175
959
8,481
6,530
1,026
9,159
6,877
1,034
Reflections with
I [ 2r(I)
Number of
parameters
Rint
0.0568
1.021
0.0459
0.977
0.0485
1.025
Goodness of fit
R1 [I [ 2r(I)]
wR2
0.0626
0.1428
0.0519
0.1316
0.64/-0.60
0.0540
0.1301
0.92/-0.69
Largest difference 0.71/-0.74
peak and hole
˚
(e A
-3
)
Thermal analysis
Thermal profiles
isostructurality of complexes 1, 2 and 3 implies a common
arrangement of the host molecules and thus the packing
arrangements of the complex units are identical. This is
evident from the experimental PXRD patterns of the three
complexes that were obtained by kneading (Fig. 4). These
patterns are also consistent with the corresponding com-
puted patterns (see supporting information), confirming
that the single crystals selected for X-ray analyses are
representative of the respective bulk materials. A detailed
comparison of the computed patterns for complexes 1 and
2, containing guests that are close analogues, shows the
patterns to be in very good correspondence. Significant
intensity differences arise for 3, owing to the different
nature of the included acetochlor molecule and its unique
location in the host cavity.
The TG traces for the three TRIMEB inclusion complexes all
display two-step mass losses. The first one corresponds to a
single guest molecule being released from the CD cavity
[exptl. 16.5 0.1, 16.5 0.2 and 16.3 0.2 % (n = 5);
calcd. for a 1:1 host–guest ratio, 16.2, 16.3 and 15.9 % for
complexes 1, 2 and 3, respectively] and the second mass loss
can be attributed to host decomposition (Fig. 5). There was a
considerable difference in the onset temperatures of guest
loss for the samples prepared by co-grinding (see supporting
`
information) vis-a-vis those prepared by co-precipitation, on
which the crystal structure determinations were based. These
123

J Incl Phenom Macrocycl Chem (2013) 75:47–56
51
Fig. 2 Stereoviews of the
complexes 1, 2 and 3 with their
host residue numbering. For
clarity, only the major host and
guest components of complex 3
are shown
differences have been summarised in Table 2 and are a result
of the different average particle sizes of each sample. The co-
precipitation method produced large single crystals that were
gently crushed and blotted dry on filter paper before being
placed in the TG instrument, whereas the materials obtained
by kneading contained particles of smaller average size,
resulting in guest loss occurring at a lower temperature. The
melting points of the complexes (as determined by DSC)
prepared by the two methods differ by a maximum of only
3 K. This minimal temperature difference is due to the
excellent heat transference between the closed sample pans
and the thermoelectric discs on which the pans are positioned.
Despite the early onset temperatures at which guest loss
for the complexes commences, the entire process of guest
loss takes place over a wide temperature range (onset temp.
to *240 °C). Consequently, the processes of guest loss
and complex fusion partially overlap, guest loss preceding
fusion in each case.
Isothermal method
The general equation used to study the kinetics of a solid-
state process at a constant temperature can be represented
by Eq. 1 or by the integral form (Eq. 2),
123

52
J Incl Phenom Macrocycl Chem (2013) 75:47–56
da
dt
¼ kfðaÞ
ð1Þ
ð2Þ
gðaÞ ¼ kt
where a is the fractional mass loss [(mt - mo)/(m? - mo)]
at reaction time t, k is the rate constant and f(a) and g(a)
are functions that describe a particular reaction mechanism.
The a-time curves obtained from the isothermal experiments
conducted for each complex are shown in Fig. 6. These plots
show deceleratory behaviour and were analysed using
various mathematical expressions derived from proposed
mechanisms for a range of solid-state processes [16]. For our
results, the equation based on the first-order reaction model
(F1, Eq. 3) as well as the equation based on the 3-D
diffusion model (3D, Eq. 4), showed a high degree of
linearity and good correlation coefficients when fitted to the
a versus time graphs over an a range of 0.0–0.8 for the F1
model and 0.1–0.9 for the 3D model.
Fig. 3 A packing diagram of complex 2 viewed along [100].
Hydrogen atoms have been omitted for clarity
ꢁlnð1 ꢁ aÞ ¼ kt
ð3Þ
ð4Þ
h
i
2
1=3
1 ꢁ ð1 ꢁ aÞ
¼ kt:
Using both models we were able to obtain two sets of
rate constants (k) at specific temperatures for each complex
studied. An Arrhenius plot (ln k vs. 1/T) provided a straight
line (Fig. 7), from which the activation energy (Table 3)
associated could be determined. We interpret the derived
Ea values here as referring to the composite process of
guest loss and complex fusion owing to the wide a-ranges
spanned in the isothermal experiments.
There are many instances in the literature where more
than one solid-state model can be used to describe a par-
ticular process [17, 18]. The F1 model is based on the
apparent ‘order’ of reaction as seen for rate laws estab-
lished in solution; however, it does not have the same
significance for solid-state reactions. The mechanism of a
first-order reaction involves random nucleation (either at a
particle surface or at an imperfection) and growth of these
nucleation sites that does not advance beyond the indi-
vidual crystallite nucleated [16]. The activation energies
obtained using this model (Table 3) are not statistically
different, suggesting that the kinetics of the process of
guest loss and melting of a given complex is not dependent
on the structure of the guest molecule within the TRIMEB
cavity.
Fig. 4 Experimental PXRD patterns of the three isostructural com-
plexes obtained by kneading TRIMEB and the respective guests.
These patterns match those generated from the single crystal X-ray
data (see supporting information)
The 3D model applies to a diffusion-limited reaction
which takes place at a phase interface that forms between
the product (TRIMEB) and the reactant (TRIMEB–pesti-
cide complex). During the process of dissociation, the
guest molecules diffuse towards the crystal surface and
finally desorb from the surface. The activation energies
obtained from the 3D model (Table 3) are very similar
once again but do display a decreasing trend. One can
Fig. 5 A schematic TG trace for the TRIMEB inclusion complexes
1–3
123

J Incl Phenom Macrocycl Chem (2013) 75:47–56
53
Table 2 The onset temperatures of guest loss and the melting points of the three complexes prepared using two methods
Complex
Onset temperature of guest loss as determined by TGA (°C)a
Temperature of complex fusion determined by DSC (°C)
Co-precipitation
Co-grinding
Co-precipitation
Co-grindingb
1
2
3
a
129
135
109
100
93
160
153
136
158
152
133
83
These temperatures were measured based on the standard 10 K min-1 heating rate
See Supplementary data for the DSC profiles
b
Fig. 6 Isothermal plots of a versus time for complexes a 1, b 2 and c 3
Fig. 7 Arrhenius plots obtained from the isothermal kinetic data for a 1, b 2 and c 3
Table 3 Activation energies
determined from the isothermal
1
2
3
experiments for each complex
using the two best-fitting
models (F1 and 3D)
F1
3D
F1
3D
F1
3D
86
Ea (kJ mol-1
r2
)
126
4
114
1
129 10
0.989
109
9
130 13
0.990
8
0.997
0.999
0.986
0.991
therefore propose that the kinetics of the process is
dependent on the structure or physical properties of the
complexes, e.g. their melting points (decreasing in the
order 1 [ 2 [ 3). Even though the guest loss process
being studied is complicated due to two simultaneous
events occurring (guest loss and complex fusion), the 3D
model appears to be a reasonable one to be applied to the
kinetics of the above events. Prior to the complexes
melting, the 3D model is consistent with the crystal
packing arrangement of the complexes. The guest mole-
cules pack within their respective TRIMEB crystal struc-
tures such that they are isolated from one another and
there are no particular directional constraints on their
diffusion (Fig. 8).
123

54
J Incl Phenom Macrocycl Chem (2013) 75:47–56
Fig. 8 A stereoview showing
the packing in the crystal of
complex 2 viewed along the b-
axis to illustrate the lack of
continuous channels. Hydrogen
atoms have been omitted for
clarity
Fig. 9 Plots of log b versus 1/T for complexes a 1, b 2 and c 3
Non-isothermal method
Fig. 9c the slopes of the plots increase at lower conversion
levels, possibly indicating that the process observed is a
composite one. For all three complexes the activation
energies were calculated by equating the slopes of the
curves to (-0.457Eaa/R) and solving for Eaa (refer to
Eq. 5). These values are reported in Table 4 and have been
averaged for each sample.
A model-free isoconversional method based on the Ozawa,
Flynn and Wall (OFW) method [19, 20] was used to
determine the activation energies for comparison with
those obtained from the isothermal experiments performed
on the three isostructural complexes. Equation 5 allowed
the determination of the activation energies at each con-
version level, where ba is the heating rate, Aa the frequency
factor, Eaa is the activation energy, Ta is the temperature at
each conversion level and g(a) refers to the kinetic model.
The variation in Ea values obtained for complexes 1 and
2 is random with no trend evident as the conversion level
increases. The relative percentage errors for these activa-
tion energies are less than 10 %, which is within the con-
ventionally accepted range for isoconversional methods
[18]. The decrease in the activation energy for dissociation
of complex 3 as a function of a can be explained by the
significantly lower melting point of 3 (133 °C) in com-
parison to complexes 1 (158 °C) and 2 (152 °C); and thus,
one, two and three reciprocal temperature values recorded
at the 12, 16 and 20 % conversion levels, respectively,
were lower than the reciprocal fusion temperature of 3 (see
supporting information).
AaEaa
Eaa
logba ¼ log
ꢁ 2:315 ꢁ 0:457
ð5Þ
gðaÞR
RTa
Activation energies were evaluated at the conversion
levels 8, 12, 16 and 20 %. These levels were chosen since
the vast majority of the temperatures derived at the
different heating rates for the specific conversion levels
were lower than the respective melting points of
the complexes, thus precluding the influence of melting
on the dissociation kinetics. The plots of log ba versus the
reciprocal temperature values are shown in Fig. 9.
Figure 9a and b illustrate fairly linear plots whereas in
With regard to the activation energies obtained from the
isothermal and non-isothermal experiments, the latter are
nominally larger; however if their values are examined for
123

J Incl Phenom Macrocycl Chem (2013) 75:47–56
55
diffusion-based models are obviously dependent on the
packing arrangement of the complex units and are different
for the b-CD and TRIMEB complexes discussed here.
It is important to assess the thermal stability of CD
inclusion complexes in the solid state. These studies show
that a relatively high energy barrier is associated with guest
loss from a TRIMEB molecule and that these solid-state
complexes should remain intact during normal handling
and storage of the material. Finally, the method of
mechanical co-grinding employed here, involving a solid
host and a guest which is an oil at ambient temperature,
may be feasible in an industrial setting to produce large
quantities of the complex, thereby stabilising oil-based
agrochemicals in a solid form [2].
Table 4 Activation energies obtained at specific conversion levels
for the three inclusion complexes using non-isothermal thermo-
gravimetric methods
Conversion level (%)
Eaa (kJ mol-1
)
1
2
3
8
142.9
130.1
185.5
12
158.2
155.3
157.8
16
167.7
160.6
138.0
20
166.8
150.4
123.0
Average
158 (12)
149 (13)
151 (27)
each complex, the largest difference in Ea magnitude cor-
responds to only *2.5 combined standard deviations. On
the other hand, direct comparison of the two sets of acti-
vation energies is not strictly justified due to the different
methodologies used to obtain them (model-fitting versus
model-free/isoconversional methods) and furthermore, the
conversion level ranges over which the activation energies
were determined were different for the two methods.
We consider the non-isothermal kinetic method as the
preferred method to study the complicated systems pre-
sented by the three isostructural inclusion complexes since
the activation energies calculated from these experiments
depend mainly on the initial stages of guest loss, with
minimal interference from the process of complex fusion.
Supporting information
The CIF files for the complexes 2 and 3 have been
deposited with the Cambridge Crystallographic Data Cen-
tre (CCDC 857692-857693). These data may be obtained
free of charge on request from the CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK. Tel.: ?44-1233-336408; Fax:
?44-1233-336033; E-mail: deposit@ccdc.cam.uk. Addi-
tional data: DSC and TG profiles of the co-ground material
as well as the computed PXRD patterns generated from the
single crystal X-ray structures of the three complexes are
supplied. The reciprocal temperature values for the non-
isothermal experiments have been tabulated.
Conclusion
Acknowledgments This material is based on research supported by
the National Research Foundation (NRF) under Grant number 67381.
MRC and DLC express their thanks to the NRF and the University of
Cape Town for financial assistance. Financial assistance from Cons-
The X-ray crystal structures of complexes 2 and 3 provide
the first accurate molecular parameters for the guest mole-
cules fenthion and acetochlor. These complexes are also
new members of a well-established isostructural TRIMEB
inclusion complex series and their addition extends further
the already wide variety of guest molecules capable of being
included within the host TRIMEB [15]. Furthermore, the
present report is the first in which an attempt is made to
reconcile kinetic data for the dissociation of guest molecules
from permethylated-CD inclusion complexes with the
crystal structures of the complexes. There are several studies
that have investigated the dissociation of a guest molecule
from native CDs [21–23]. The activation energies reported
for the thermal dissociation of benzyl alcohol and cinna-
maldehyde from their b-CD inclusion complexes, for
example, were reported as 158 and 160 kJ mol-1 respec-
tively [21, 23], which are very similar to the values of Ea
determined here for the three TRIMEB inclusion complexes
using the method of non-isothermal TGA. The guest loss
process for the b-CD complexes containing benzyl alcohol
and cinnamaldehyde occurs via a diffusion-based mecha-
nism, the latter being proposed here as applying also to the
TRIMEB inclusion complexes. The dimensionalities of the
´
ejo Nacional de Investigaciones Cientıficas y Tecnicas (CONICET),
Fondo para la Investigacion Cientıfica y Tecnologica (FONCYT),
´
´
´
´
´
Ministerio de Ciencia y Tecnologıa (MINCyT), Cordoba (Argentina)
and support from the National University of Cordoba are greatly
´
´
acknowledged. This work was carried out as part of a bilateral
cooperation project supported by the NRF (South Africa) and
MINCyT (Argentina). NMR is a grateful recipient of a fellowship
from CONICET (Argentina).
References
1. Szejtli, J.: Past, present and future of cyclodextrin research. Pure
Appl. Chem. 76, 1825–1845 (2004)
2. Szejtli, J.: Cyclodextrins in pesticides. Starch 37, 382–386 (1985)
3. Morillo, E.: Applications of cyclodextrins in agrochemistry. In:
Dodziuk, H. (ed.) Cyclodextrins and Their Complexes,
pp. 459–466. Wiley-VCH Verlag GmBH & Co. KgaA, Weinheim
(2006)
4. Kegley, S.E., Hill, B.R., Orne, S.: PAN pesticides database—
pesticide registration status. http://www.pesticideinfo.org (2011).
Accessed 15 Oct 2011
123

56
J Incl Phenom Macrocycl Chem (2013) 75:47–56
5. Kamiya, M., Kameyama, K., Ishiwata, S.: Effects of cyclodex-
trins on photodegradation of organophosphorus pesticides in
15. Cambridge Structural Database and Cambridge Structural Data-
base System. Conquest, Version 1.13, Cambridge Crystallographic
Data Centre, University Chemical Laboratory: Cambridge, Eng-
land (2011)
humic water. Chemosphere 42, 251–255 (2000)
´
6. Vico, R.V., Bujan, E.I., de Rossi, R.H.J.: Effect of cyclodextrin
on the hydrolysis of the pesticide fenitrothion [O, O-dimethyl O-
(3-methyl-4-nitrophenyl) phosphorothioate]. Phys. Org. Chem.
15, 858–862 (2002)
16. Galwey, A.K., Brown, M.E.: Kinetic Background to thermal
analysis and calorimetry, Chap. 3. In: Brown, M.E. (ed.) Hand-
book of Thermal Analysis and Calorimetry, pp. 147–224. Else-
vier Science, Amsterdam (1998)
17. Vyazovkin, S., Wight, C.A.: Isothermal and nonisothermal
reaction kinetics in solids: in search of ways toward consensus.
J. Phys. Chem. A 101, 8279–8284 (1997)
18. Vyazovkin, S., Wight, C.A.: Isothermal and non-isothermal
kinetics of thermally stimulated reactions of solids. Int. Rev.
Phys. Chem. 17, 407–433 (1998)
19. Ozawa, T.: A new method of analysing thermogravimetric data.
Bull. Chem. Soc. Jpn. 38, 1881–1886 (1965)
7. Rougier, N.M., Cruickshank, D.L., Vico, R.V., Bourne, S.A.,
´
Caira, M.R., Bujan, E.I., de Rossi, R.H.: Effect of cyclodextrins
on the reactivity of fenitrothion. Carbohydr. Res. 346, 322–327
(2011)
8. Cruickshank, D.L., Rougier, N.M., Vico, R.V., de Rossi, R.H.,
´
Bujan, E.I., Bourne, S.A., Caira, M.R.: Solid-state structures and
thermal properties of inclusion complexes of the organophos-
phate insecticide fenitrothion with permethylated cyclodextrins.
Carbohydr. Res. 345, 141–147 (2010)
9. Program SAINT, Version 7.60a, Bruker AXS Inc., Madison, WI,
USA (2006)
10. Program SADABS, Version 2008/1, Bruker AXS Inc., Madison,
WI, USA (2008)
11. Harata, K., Uekama, K., Otagiri, M., Hirayama, F.: Conformation
of permethylated cyclodextrins and the host–guest geometry of
their inclusion complexes. J. Inclusion Phenom. Mol. Recognit.
Chem. 1, 279–293 (1984)
12. Mentzafos, D., Mavridis, I.M., Schenk, H.: Crystal structure of
the 1:1 complex of heptakis(2,3,6-tri-O-methyl)cyclomaltohep-
taose (permethylated b-cyclodextrin) with ethyl laurate. Carbo-
hydr. Res. 253, 39–50 (1994)
20. Flynn, J.H., Wall, L.A.: A quick, direct method for the deter-
mination of activation energy from thermogravimetric data.
J. Polym. Sci. Part B Polym. Lett. 4, 323–328 (1966)
21. Zhang, N., Li, J.-H., Cheng, Q.-T., Zhu, M.-W.: Kinetic studies
on the thermal dissociation of b-cyclodextrin-benzyl alcohol
inclusion complex. Thermochim. Acta 235, 105–116 (1994)
22. Neoh, T.L., Yamauchi, K., Yoshii, H., Furuta, T.: Kinetic study
of thermally stimulated dissociation of inclusion complex of
1-methylcyclopropene with a-cyclodextrin by thermal analayis.
J. Phys. Chem. B 112, 15914–15920 (2008)
23. Li, J.-H., Zhang, N., Li, X.-T., Wang, J.-Y., Tian, S.-J.: Kinetic
studies on the thermal dissociation of the inclusion complex of b-
cyclodextrin with cinnamic aldehyde. J. Therm. Anal. 49,
1527–1533 (1997)
13. Sheldrick, G.M.: A short history of SHELX. Acta Crystallogr.
Sect. A. A64, 112–122 (2008)
14. Song, Z.-W.: 2-Chloro-N-chloromethyl-N-(2-ethyl-6-methyl-
phenyl)acetamide. Acta Crystallogr. Sect. E. E64, o991 (2008)
123