Thermodynamic and kinetic stability of inclusion compounds under heating

Article abstract of DOI:10.1007/s10973-007-8473-1

Thermodynamic and kinetic stability of inclusion compounds (so called supermolecular compounds) is discussed. Compounds under study and discussion are clathrates (with coordination compounds matrices) and intercalates (with fluorinated graphite matrices).

Full text of DOI:10.1007/s10973-007-8473-1

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Journal of Thermal Analysis and Calorimetry, Vol. 90 (2007) 1, 23–30
THERMODYNAMIC AND KINETIC STABILITY OF INCLUSION
COMPOUNDS UNDER HEATING
V. Logvinenko*, V. Drebushchak, D. Pinakov and G. Chekhova
Nikolaev Institute of Inorganic Chemistry, Siberian Branch of Russian Academy of Sciences, Lavrentyev Ave. 3
Novosibirsk-90, 630090, Russia
Thermodynamic and kinetic stability of inclusion compounds (so called supermolecular compounds) is discussed. Compounds under
study and discussion are clathrates (with coordination compounds matrices) and intercalates (with fluorinated graphite matrices).
Keywords: clathrates, compounds stability, intercalates, ‘model free’ kinetics, supermolecular chemistry, thermal decomposition
Introduction
[M(Py)3(NO3)2] and [M(Py)2(NO3)2]). The guest-free
matrix is unstable. It is interesting that all but one
studied clathrates (with Mn, Co, Ni, Zn and Cd-matri-
ces) decompose under quasi-equilibrium condition
right away to the [M(Py)3(NO3)2], – because of low
thermodynamic stability of Mn-, Co-, Ni-matrix
[M(Py)4(NO3)2] at vapor pyridine pressure »101 kPa,
Supramolecular compounds
Inclusion compounds are important representatives of
supramolecular systems. Their synthesis (and the exis-
tence itself) are based on their high enough kinetic
(and/or thermodynamic) stability at room temperature
in the processes of thermal dissociation with the break-
down to the host matrix and the guest (volatile) com-
pound. There are no direct connections between the de-
tailed structure of the inclusion compounds (such as
strong and week bonds, etc.) and compounds stability
under heating, their steps of decomposition, stability of
capable intermediate phases, and so on. However, their
thermodynamic stability in thermal dissociation pro-
cess is essentially connected with entropy contribution
in more important way, than with energetic factor.
Clathrates and intercalates are the most studied
groups of supramolecular compounds.
without
liquid
pyridine
support.
Only
[Cu(Py)4(NO3)2]·2Py clathrate decomposes under
quasi-equilibrium conditions up to [Cu(Py)4(NO3)2]
(at Py pressure »101 kPa) at 120°C. It agrees with the
strength of M–Py bonds within the coordination
sphere [1, 2].
Intercalates on the base of fluorinated graphite
The most known intercalates are intercalates with
graphite matrix and with fluorinated graphite matri-
ces. Layered compounds (intercalates) based on
graphite fluoride matrices with various guest mole-
cules are supposed to be a particular group of inclu-
sion compounds.
Clathrates
So as the phases are layered substances, there are
series of different structures: the first stage of filling
(alternating unit layer of the matrix and unit layer of
the guest molecules, Fig. 1a), the second stage of fill-
ing (alternating double matrix layer and one guest
molecule layer, Fig. 1b), the third stage of filling, etc.
Keen interest to these new materials is deter-
mined by their two applications.
Regular microporous materials such as zeolites or co-
ordination polymers (metal-organic frameworks) are
well-known host matrices, which are able to
incapsulate smaller guests into the interlattice voids
to form host–guest inclusion compounds.
The stability of such inclusion compound can be
connected with the stability of the matrix structure it-
self. For example, well known inclusion compounds
They can be used as so-called ‘molecular con-
tainers’ both for different gases (such as CH4, NO2,
(clathrates)
[Cd(Py)4(NO3)2]·2Py can be synthesized in the sys-
tem ‘Zn(NO3)2(crystal)–Py(liquid)’ and ‘Cd(NO3)2(crystal)
Py(liquid)’, although the host compounds
[Zn(Py)4(NO3)2]·2Py
and
HF etc.) and condensed phases (BrF3, n-CnH2n+2
,
–
C2H5OH, C6H6, (CH3)2CO, SOCl2, etc.) [3].
From the other hand, these intercalates with vol-
atile guests are used for the expanded graphite pro-
[Zn(Py)4(NO3)2] and [Cd(Py)4(NO3)2] itself do not
exist in these systems (stable crystal phases are
*
Author for correspondence: val@che.nsk.su
1388–6150/$20.00
Akadémiai Kiadó, Budapest, Hungary
Springer, Dordrecht, The Netherlands
© 2007 Akadémiai Kiadó, Budapest

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
LOGVINENKO et al.
Kinetic stability
In physical chemistry the kinetic stability usually and
traditionally discussed as kinetic lability or reactivity.
The convenient presentation of such a decomposition
process is any thermoanalytical curve, depending on
temperature (Figs 2 and 3 [10]); the term ‘thermal sta-
bility’ originates exactly from the contemplation of
such pictures. The onset temperature Ton (175 or
190°C) on this picture) is found as the visible devia-
tion of TG (DTG, DSC) curve from the baseline. It is
easy to demonstrate that the onset temperature Ton
(being determined from the DTG curves very pre-
cisely) is the temperature of the achievement of a cer-
tain rate constant, the specific constant value being
depended on the used research device (both balance
and monitoring) sensitivity [9]. So the onset tempera-
ture Ton is the kinetic parameter only; it can be used
for the kinetic lability comparison indeed, but only if
all compounds decomposition was studied by means
of one and the same instrument (and the experiment
itself is organized as the ‘kinetic’ one: small sample,
the sufficient shift from the equilibrium position, se-
lected heating rates, etc.); the dependence on heating
rate is clearly evident (Figs 2 and 3).
Fig. 1 Simplified structures of the graphite fluoride inclusion
compounds: a – first stage of filling and b – second
stage of filling
duction (fast increase of the guest pressure under
heating results in the formation of expanded graphite
with low density, such as 3–10 kg m–3) [4, 5].
There are only Van der Waals interaction be-
tween the fluorinated graphite and the intercalated
guest molecules, while graphite inclusion compounds
have mostly donor-acceptor type of interactions
[6, 7]. That was shown by IR spectra, where wave
numbers are considerably shifted from those in pure
graphite and guest components, while wave numbers
of guests, intercalated into fluorinated graphite, are
nearly similar to pure guest spectra. The first stage of
filling is formed usually during the synthetic process;
because of low stability of this phase, the second
stage of filling is the resulting phase after sample
‘drying’ in the air at room temperature.
And what is more, the above proposed measure
of the relative stability of an inclusion compound: the
difference between temperatures of boiling (of liquid
guest) and thermal decomposition (of solid inclusion
compound) (Ton–Tb) [8] is the difference between the
100-
96-
92-
-0
-–1
-–2
-–3
Stability under heating
The peculiar properties of inclusion compounds sta-
bility are discussed in literature. ‘We have found that
the value of the onset temperature, Ton, characterizing
the temperature of guest release, is a reliable measure
of thermal stability. For inclusion compounds of a
given host with the variety of guests, the onset tem-
peratures are clearly a function of both the host–guest
interactions and the intrinsic properties of the guest it-
self. The normal boiling point Tb of the guest is im-
portant, and a useful measure of the relative stability
of an inclusion compound is the difference (Ton–Tb).
This is of interest when trying to encapsulate highly
volatile guests’ [8]. Really the temperature of com-
pound decomposition (both the onset, and maximum
on DTG or DTA, DSC peaks) is a kind of valuation of
‘thermal’ stability, but from the strict physico-
chemical point of view, only thermodynamic stability
or kinetic stability exist. ‘Thermal’ stability depends
on heating rate, the shift from the equilibrium position
(evolved gas pressure, connected with sample holder
shape and inert gas flow) and is susceptible to sample
mass and equipment sensor sensitivity [9].
88-
84-
100
150
200
250
300
Temperature/°C
Fig. 2 TG and DTG curves for Mn(HCOO)2·1/3C4H8O2 de-
composition (dioxane removal), m=7.00±0.02 mg, ar-
gon flow 60 cm3 min–1, heating rate 12°C min–1
100-
96-
92-
88-
84-
-0.0
-–0.4
-–0.8
-–1.2
3
100
150
200
250
300
Temperature/°C
Fig. 3 TG and DTG curves for Mn(HCOO)2·1/3C4H8O2 de-
composition (dioxane removal), m=7.00±0.02 mg, ar-
gon flow 60 cm3 min–1, heating rate 4°C min–1
24
J. Therm. Anal. Cal., 90, 2007

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
INCLUSION COMPOUNDS UNDER HEATING
kinetic and thermodynamic parameters and cannot be
easily attributed either to kinetic stability, nor to ther-
modynamic one.
of compounds with one and the same matrix and dif-
ferent guests (or different matrices and one guest).
As to the simple comparison of stability (Ton–Tb),
above cited, it contains meaningful uncertainty. Re-
ally, the boiling temperature Tb is the equilibrium
temperature of the achievement of 101.3 kPa pressure
of the vapor under liquid guest, and the (thermo-
analytical) decomposition temperature Ton is the tem-
perature of decomposition beginning at casual experi-
mental conditions (guest pressure, heating rate, sam-
ple holder shape) and it fixed the certain (usually un-
known) rate constant.
Thermodynamic stability
The formation of the inclusion compound (from the
compact solid matrix and gaseous guest) can be di-
vided into two stages: 1) the expansion of compact
host matrix, and 2) the inclusion of guest molecules
into the beforehand-expanded (so-called ‘apohost’)
matrix.
The first stage is the endothermic process, and
the system entropy increases (DH1>0; DS1>0). The
second stage is the exothermic process, and the sys-
tem entropy decreases (DH2<0; DS2<0). The disorder
reduces essentially because of the guest molecules
packing into the solid, guest molecules lose
translational, rotational and other degrees of freedom;
so the total entropy change is always negative:
DSf=(DS1+DS2)<0. The inclusion compound is formed
only if the enthalpy sum is negative: DHf=
(DH1+DH2)<0. So then the formation reaction of the
inclusion compound is controlled by enthalpy term. It
is worth to note that the compensational enthalpy–en-
tropy effect is observed under cyclodextrines
clathrates formation [11, 12].
The foregoing measure of inclusion compounds
stability (Ton–Tb) will acquire well defined physical
chemical meaning if the decomposition temperature
will be determined in the thermoanalytical experi-
ment under quasi-equilibrium conditions in one and
the same sample holder (the labyrinth sample holder,
conical sample holder, etc). Such experiment runs at
constant and small decomposition rate (say,
0.2 mg min–1), at almost constant evolved gas pres-
sure (it is maintained at »101 kPa for labyrinth holder
and »20 kPa for the standard sample holder with lid);
this decomposition temperature Teq remains almost
constant (±5–7°C) up to the end of decomposition
step; the steps of the decomposition correspond to the
thermodynamically stable intermediate phases
[2, 15]. This decomposition temperature is the tem-
perature of the achievement of the certain equilibrium
constant (Keq»1 for the reversible decomposition in
the labyrinth sample holder) for so called dissociation
reaction.
The decomposition of the inclusion compound
under heating can be divided into two stages: 1) the
guest molecules removal (without matrix change) and
2) the collapse of the empty matrix (‘apohost’) to the
stable compact structure.
The first stage is the endothermic process, and
the system entropy increases (DH1>0; DS1>0). The
second stage is the exothermic process, and the sys-
tem entropy decreases (DH2<0; DS2<0). The disorder
increases essentially because of the guest molecules
liberation from the solid, so the total entropy change
is always positive: DSd=(DS1+DS2)>0. So as the reac-
tion of decomposition proceeds, therefore
DG=(DHd–TDS)<0; for the endothermic reaction the
enthalpy sum is positive {DHd=(DH1+DH2)>0}, there-
fore the entropy term TDSd>DHd. So then the decom-
position reaction of the inclusion compound is con-
trolled by the entropy term [13].
It is worth to note that the scientists tried to syn-
thesize the ‘ideal inclusion compound’, in which the
interaction enthalpy was assumed to be negligible
with respect to entropic consideration (and so DmixG
could be considered to be ideal) [14]. This approach is
valid only if the pure host matrix remains ‘empty’ af-
ter the thermal decomposition.
For the series of clathrates decomposition (with
identical stoichiometry):
[M(Py)4(NO3)2]·2PyÛ[M(Py)3(NO3)2]+3Py
the row of equilibrium temperatures of decomposition
(Mn<Co<Ni<Cu) fully corresponds to row of reac-
tion enthalpy (DHr values are equal to 53.8, 59.5,
64.6, 74.6 kJ mol–1).
The decomposition of all these clathrates (be-
cause of the matrix itself instability) is connected not
with the inclusion guest molecules evolving, but with
the matrix [M(Py)4(NO3)2] decomposition. Just there-
fore the row of decomposition quasiequilibrium tem-
peratures and the row of decomposition enthalpies
agrees with the strength of M–Py bonds within the
coordination sphere [1, 2].
Naturally, the row of above mentioned tempera-
ture difference (Teq–Tb), with constant Tb, would be
the same.
The direct way to define the thermodynamic sta-
bility of inclusion compounds is tensimetric measure-
ments of the equilibrium guest pressure vs. tempera-
ture [1, 2] (or precise DSC measurements) for series
J. Therm. Anal. Cal., 90, 2007
25

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
LOGVINENKO et al.
Experimental
averaged values of activation energy according to
temperatures of maximum decomposition rate.
Programs ‘Friedman analysis‘ and ‘Ozawa–Flynn–
Wall analysis’ allow calculating the activation energies
for the every experimental point of fractional
conversion (in the interval 0.02<a<0.98). ‘Ozawa–
Flynn–Wall analysis’ is based on the dependence
between the heating rate and inverted temperature;
‘Friedman analysis’ is based on the dependence
between the rate of conversion degree and inverted
temperature. The same set of experimental data can be
used further for searching the topochemical equation
(the selection from 18 equations: nucleation, chemical
reaction on the interface and diffusion). The checking
of the diffusion hindrance absence is very important,
so as the aim of the investigation is the chemical
reaction study. This calculation is made by the im-
proved differential method of Borchardt–Daniels, with
multiple linear regression method, or by nonlinear
regression method (this method allows the consecutive
reaction processing). This part of the calculation
includes the preexponential factor determination.
Special F-test is used for the search of the best kinetic
description [19–25].
The
thermal
decomposition
and
of
clathrates
(Mn(HCOO)2·1/3C4H8O2
Mn(HCOO)2·
1/3C4H8O) was studied under non-isothermal condi-
tions [10, 16]. TG measurements were carried out on
a Netzsch thermal analyzer TG 209. The experiments
were performed in argon flow (100 cm3 min–1), at
heating rates of 4, 8 and 12°C min–1, the sample mass
was kept 7.00±0.02 mg. Thermomechanical analysis
(dilatometry) was carried on a Netzsch thermal ana-
lyzer TMA 202; the powder layer thickness was
1.02–1.03 mm (with sample mass 42–45 mg).
We studied the thermal decomposition of inter-
calates (the 1st stage of filling) on the base of fluori-
nated graphite matrix (with different fluorination de-
gree). Intercalates with acetonitrile were: C2FxBrz·
yCH3CN (x=0.92, 0.87, 0.69 and 0.49, z»0.01, y var-
ied from 0.288 to 0.174) [17, 18].
Intercalates with chloroform were C2FxBrz·
yCHCl3 (x=0.92, 0.87, 0.69 and 0.49, z»0.01, y varied
from 0.204 to 0.139).
Intercalates with dichlorethane were C2FxBrz·
yCH2ClCH2Cl (x=0.92, 0.87, 0.69 and 0.49, z»0.01, y
varied from 0.222 to 0.153).
The thermal decomposition of these intercalates
was studied under isothermal conditions; the choice
of the isothermal experiments (at rather low tempera-
ture) is connected with very low stability of synthe-
sized compounds with the 1st stage of filling. Investi-
gation of inclusion compound thermal decomposition
were carried out by applying thermogravimetric
method in ambient atmosphere, taking inclusion com-
pound sample mass decrease point-by-point, using
thermostatically controlled analytical balance at dif-
ferent temperatures from 14 to 40°C (±0.2°C), sample
mass 30–40 mg. Equilibrium shift was enough for ki-
netic investigations.
The decomposition of the second stage of filling
of intercalates was studied under non–isothermal con-
ditions. TG measurements were carried out on a
Netzsch thermal analyzer TG 209. Thermomecha-
nical analysis (dilatometry) was carried on a Netzsch
thermal analyzer TMA 202; the powder layer thick-
ness was 1.02–1.03 mm (with sample mass
42–45 mg).
Kinetic analysis under isothermal condition
We studied the thermal decomposition of intercalates
on the base of fluorinated graphite matrix (with dif-
ferent fluorination degree).
Experimental data of decomposition degree a vs.
t for each matrix at different temperatures were
mapped by logarithmic form of Avramy–Erofeev’s
equation
(with
Sakovich’s
correlation):
ln[–ln{1–a)]=nln(k/n)+nlnt, where n – nondimen-
sional coefficient, k – generalized rate constant, s–1.
Equation shown above was successfully used by
many authors for kinetic investigations of both ther-
mal decomposition of different compounds (includ-
ing carbamide-n-alkanes clathrates) and phase transi-
tions, solvation and absorption [26, 27]. The general-
ized activation energy is the sum of partial energies of
nucleation, nuclei growth and diffusion. We
meaningly used generalized Avramy–Erofeev’s equa-
tion for the estimation of decomposition kinetic pa-
rameters: values n>1 correspond to reaction with ki-
netic control, n<1 – with diffusion control [26, 27].
Such equation selection gave us the possibility to look
after the imbedding from the reaction kinetics to the
diffusion along the series both of different matrices
(the same guest), and for different guests (the same
matrix).
Kinetic analysis under non-isothermal conditions
The up–to–date convenient experimental path to the
kinetics study is now the ‘model-free’ approach.
Special computer program module ‘model-free’ allows
processing several TG, DTA or DSC curves, obtained
with different heating rates, without the preliminary
information about the kinetic topochemical equations.
Program ASTM E698 enables to calculate the
26
J. Therm. Anal. Cal., 90, 2007