SINGH et al.
Table 4 Ignition delay D
i
, activation energy for thermal ignition (E*) and correlation coefficient (r) for the complexes
D
i
/s at temperature/°C
–1
Complex
E*/kJ mol
r
3
75±1
400±1
425±1
450±1
475±1
[
[
[
Co(bipy)
2
](NO
](NO
](NO
3
)
2
⋅4H
2
O
74
123
105
55
102
88
48
80
71
42
65
62
35
49
53
28.6
36.8
27.8
0.9907
0.9944
0.9986
Ni(bipy)
2
3
)
2
2
⋅2H O
Cu(bipy)
2
3 2
)
Thus, the plausible mechanistic pathways for the
thermolysis of this complex may be proposed as
to correlate particular activation energy for a particu-
lar process in thermolysis.
To examine the effect of rapid heating on the
1
87–218°C
[
Cu(bipy) ](NO ) ⎯ ⎯ ⎯ ⎯ →
2
3 2
complexes, D data reported in Table 4 indicates that
i
[
Cu (bipy)](NO ) +bipy
3 2
the time required for the thermal ignition at a particular
temperature, increases in the order Co<Cu<Ni com-
plex. Thus, showing the relative thermal stability of the
complexes in the same order under the condition of
rapid heating. Activation energy for the thermal igni-
tion of the complex (Table 4) was found to be in the in-
creasing order as Co<Cu<Ni complex which is also in
agreement with the thermal stability order of the com-
plexes. The values of activation energies for slow ther-
mal decomposition (E) and rapid decomposition (E*)
are not same, but the increasing order is same in both
the cases. Less stability of the cobalt complex may be
attributed to the less nuclear charge and greater ionic
3
05–324°C
[
Cu(bipy)](NO ) ⎯ ⎯ ⎯ ⎯ →CuO+gaseous products
3
2
Monobipyridine cobalt and copper nitrate is
formed as intermediates and separation of stages is due
to the relatively higher thermal stability of this mono-
ligand intermediate in comparison to that of nickel com-
plexes. The formation and stability of such monoligand
intermediate compounds have also been seen earlier
during the thermal studies of bis(ethylenediamine)cop-
per chloride/bromide monohydrates [20], bis(ethylene-
diamine)copper nitrate [9, 21] and bis(ethylenedi-
amine)metal perchlorate complexes [22].
The analysis of the kinetics from isothermal TG,
using model fitting methods lead to uncertainty.
However, values of activation energy E, obtained
from different models for particular samples are
nearly equal irrespective of the equations used. An
average value of E equal to 45.6, 116.3 and
2+
size of Co . Moreover, due to higher nuclear charge
2+
2+
and small size of Cu than Ni the stability of Cu
complex should be greater as compared to Ni complex.
However, the thermal stability of Ni complexes over
Cu complexes is also reported in earlier studies, where
–
9.5 kJ mol have been evaluated for cobalt, nickel
1
[Ni(en) ](ClO ) have greater thermal stability than
2 4 2
5
[
Cu(en) ](ClO ) [22].
2 4 2
and copper complexes, respectively. In case of nickel
complex, the value of E is much higher than cobalt
and copper complexes because in this case, the E is
calculated for the loss of two bipyridine and two
water molecules, while in case of cobalt and copper, E
is calculated for the loss of one bipyridine ligand
along with water molecules.
Conclusions
The thermal studies carried out on the two bipyridine
metal complexes using TG-DTA, DSC (in flowing ni-
trogen) and non-isothermal TG in air, indicates that all
these complexes decomposes in multisteps, last step of
which are rapid and exothermic. Both, nickel and cop-
per complex decompose in two steps, whereas cobalt
complex decomposes in four steps. Application of
model fitting method for the kinetic analysis from iso-
thermal data, fails to explain the complexity of thermo-
lysis as it gives single value of activation energy which
can not be assigned to a particular process. On the
other hand, isoconversional method describes well the
complexity of even single step decomposition that
yields a series of E values as a function of extent of
conversion. The ignition delay and E* have been found
to be maximum for the nickel complex and minimum
for the cobalt and copper complexes.
Application of isoconversional method to the
isothermal TG data indicates that the decomposition
of these complexes is not simple, as indicated by
model fitting methods. Values of E for these com-
plexes are found to vary with α. As we have just stud-
ied the kinetics of the removal of water molecules and
one bipyridine ligand (~30% mass loss), Fig. 10 rep-
resents the dependencies of activation energies on α
up to 0.30. The E for nickel complex in this α range is
lower than the cobalt copper complexes. For these
complexes, E decreases in α range 0.09 to 0.21 which
may be due to the predominance of exothermic
changes over endothermic changes. However, the
variation of activation energy with extent of conver-
sion is changing at every stage, which may be the re-
sult of many competeting effects and is very difficult
430
J. Therm. Anal. Cal., 85, 2006