Kinetics of thermal decomposition of mixed-ligand nikkel(II) and copper(II) complexes of 4-acyl pyrazolone derivative and pyridine

Article abstract of DOI:10.1007/s10973-005-6947-6

Thermal behaviors of two mixed-ligand complexes, [Ni(PMPP-SAL)(Py) ₃] and [Cu(PMPP-SAL)Py]?MeOH, (PMPP-SAL=1-phenyl-3-methyl-4- (salicylidene hydrazide)-propenylidene-pyrazolone-5, Py=pyridine), were studied by TG-DTG-DTA in dynamic air atmosph

Full text of DOI:10.1007/s10973-005-6947-6

Journal of Thermal Analysis and Calorimetry, Vol. 89 (2007) 2, 547–553
KINETICS OF THERMAL DECOMPOSITION OF MIXED-LIGAND
NIKKEL(II) AND COPPER(II) COMPLEXES OF 4-ACYL PYRAZOLONE
DERIVATIVE AND PYRIDINE
*
G.-C. Xu, L. Zhang, L. Liu, G.-F. Liu and D.-Z. Jia
Institute of Applied Chemistry, Xinjiang University, Urumqi 830046, P.R. China
Thermal behaviors of two mixed-ligand complexes, [Ni(PMPP-SAL)(Py)
1
3
] and [Cu(PMPP-SAL)Py]×MeOH, (PMPP-SAL=
-phenyl-3-methyl-4-(salicylidene hydrazide)-propenylidene-pyrazolone-5, Py=pyridine), were studied by TG-DTG-DTA in dy-
namic air atmosphere. The complexes show the loss of pyridine molecule is followed by the decomposition of the PMPP-SAL anion
and give respective metal oxides as residues. Meanwhile, the Ozawa–Flynn–Wall model-free analyses and multivariate non-linear
regressions were applied to perform single and overall steps optimization. Kinetic parameters were given and the most probable
mechanism functions were suggested in this study.
Keywords: kinetic studies, mixed-ligand complexes, non-linear regression, thermal analyses
Introduction
the references [22, 23]. The ligand PMPP-SAL was
synthesized according to the procedure described in
the literature [19]. Other chemicals and solvents were
analytical grade and commercially available.
Pyrazolone-5, especially 4-acyl pyrazolones, form an
important class of organic compounds and are widely
used in biological, analytical applications, catalysis and
extraction metallurgy [1–5]. Furthermore, 4-acyl
pyrazolone derivatives have a potential to form different
types of coordination compounds due to the several
electron-rich donor centers [6–9] and tautomeric effect
of enol form and keto form [10–12]. Meanwhile, com-
pounds containing hydrazide moieties and their com-
plexes also possess biological activities, especially serv-
ing as potential inhibitors for many enzymes [13–15].
During the course of our research, we have syn-
thesized a series of 4-acyl pyrazolone derivatives
Syntheses of the complexes
The synthesis procedure for [Ni(PMPP-SAL)(Py)3]
was as follows.
The ligand PMPP-SAL (1 mmol) was dissolved
by the addition of an aqueous solution of 1 M NaOH
in 15 mL methanol. To which 8 mL of methanol solu-
tion of Ni(OAc) ×4H O (1 mmol) and 1.6 mL of a
2
2
mixture of pyridine and acetonitrile with volume
rate 1:1 were added in turn, then the mixture was
stirred slightly and loaded into a Teflon-lined stain-
less steel autoclave, the sealed autoclave was heated
to 120°C and kept isothermal for 48 h. After cooled to
ambient temperature, the red black crystals were col-
lected by filtered and dried in air. Yield: 0.55 g
(83.5% based on nickel).
[16, 17] and reported several transition metal com-
plexes [18–21], but no detailed studies on thermal de-
composition and thermal kinetics have been carried out
so far. Hence, as a continuation of our earlier study and
the interest on the thermal properties of mixed-ligand
complexes, we report in this paper the phenomeno-
logical, kinetic and mechanistic aspects of thermal de-
composition of the nickel(II) and copper(II) complexes
of PMPP-SAL and containing pyridine as co-ligand.
The [Cu(PMPP-SAL)Py]×MeOH complex was
synthesized in a similar way, just using Cu(OAc) ×H O
2
2
instead of Ni(OAc) ×4H O. Yield: 0.42 g (78.2% based
2
2
on copper).
Experimental
Methods
Materials
Elemental analyses of C, H and N were performed us-
ing a PE-2400 analyzer. IR spectra were recorded on
a BRUKER EQUINOX-55 spectrophotometer within
Reagents
–
1
PMPP (1-phenyl-3-methyl-4-propionyl-5-pyrazolone)
and salicylic hydrazide were synthesized according to
4
00–4000 cm with the samples prepared as pellets
*
Author for correspondence: jdz@xju.edu.cn
1
388–6150/$20.00
Akadémiai Kiadó, Budapest, Hungary
Springer, Dordrecht, The Netherlands
©
2007 Akadémiai Kiadó, Budapest

XU et al.
–
1
with KBr. The TG-DTA simultaneous measurements
were carried out on a Netzsch STA 449C Model in-
strument. An alumina crucible was used in dynamic
medium intensity band at around 3249 cm , assigned
to the hydrogen bond NH and OH stretching vibra-
–1
tions. The strong bands at 1663 and 1614 cm are re-
spectively allocated as nC=O of the hydrazide and nC=O
of pyrazolone-ring. From above analysis, it is con-
cluded that the ligand is mainly existed in keto-form in
the solid-state. However, in the spectra of the com-
plexes, the C=O bands are both absent. Meanwhile, a
3
–1
air atmosphere (30 cm min ). The rate of heating
–
1
was 2, 5 and 10°C min and sensitivity of the instru-
ments is 0.1 mg. The crystal structure was determined
by a BRUKER SMART 1000 CCD detector and
SHELXL-97 crystallographic software package.
Solvothermal syntheses experiments were per-
formed in Teflon-lined stainless steel autoclave with
inner volume of 40 mL.
–1
new band is observed around 1440 cm due to n
C– O –
,
–1
the band at around 510 cm is assigned to nM–O. All of
these demonstrate that the oxygen atoms of carbonyl
have formed coordinate bonds with the metal ion.
From these observations, it is concluded that the ligand
reacts in the enol form when coordinated to the metal
ions in the solvent and the two enolic protons are re-
placed by M(II) in the complexes. The results also con-
firmed by the crystal structure analyses.
In the complexes spectra, the NH stretching vi-
bration disappears and two new bands at about 1600
and 1576 cm appear, they are respectively allocated
as n–C=N-N=C– of hydrazone and the coordinated
pyridine nitrogen. All of these demonstrate that the
nitrogen atom of imino-group of hydrazone is coordi-
The theoretical base of the kinetics estimation of the
thermal decomposition
According to non-isothermal kinetic models, the
commonly used kinetic equation of solid thermal de-
composition is given as follows:
da/dT=A/bexp(–E
a
/RT)f(a)
(1)
–
1
where a is fractional extent of reaction at time t, b the
heating rate, E the apparent activation energy, da/dT
a
the rate of the reaction, A the pre-exponential factor,
and f(a) is kinetic differential function. Its integral
form is as follows:
–
1
nated to M(II). The new bands at about 460 cm for
the complexes are assigned to nM–N. It further
confirms that the nitrogen atom of the imino-group
bonds to the transition metal ion.
a
T
G(a ) E
a
da/f(a)=A/b exp(–E
a
/RT)dT
(2)
ò
ò
0
0
In the ligand spectrum, the strong band observed
–
1
at 1308 cm is assigned to nC–OH stretching vibration
of phenolic oxygen. This band shifts very feeble in
the spectra of metal complexes, indicating the
non-coordination of the phenolic oxygen atom to the
central metal ion.
The fifteen reaction models and mechanism
functions [24] were used to fit kinetics curves. The ki-
netic software from Netzsch was used in this analy-
sis [25–28].
Results and discussion
Thermal decomposition
The elemental analyses results for the ligand and
mixed-ligand complexes were listed in Table 1. The
results agree well with the given formulae of the com-
pounds. The complexes are stable in air and soluble in
most of organic solvents.
[Ni(PMPP-SAL)(Py)3]
The simultaneous TG-DTG-DTA curves of the
[Ni(PMPP-SAL)(Py) ] are preserved in Fig. 1. The de-
3
composition of the mixed-ligand complex undergoes
in three stages. The endothermic degradation of two
mole pyridine molecules takes place in the first stage
between 90 and 160°C with a mass loss of 23.96%
Infrared spectra analyses
(
calcd.: 24.02%). The maximum rate of mass loss is in-
The main stretching frequencies of the IR spectra of
the ligand PMPP-SAL and its mixed-ligand complexes
are tabulated in Table 2. The ligand spectrum shows a
dicated by the DTG peak at 142°C. The residue at the
first stage was found to be [Ni(PMPP-SAL)(Py)]. The
Table 1 The elemental analyses for the mixed-ligand complexes
Found (calcd.)/%
–1
Compound
PMPP-SAL
Formula mass/g mol
C
H
N
364.40
658.39
537.07
65.44 (65.92)
63.94 (63.85)
58.32 (58.15)
5.81 (5.53)
5.13 (5.05)
5.07 (5.07)
15.28 (15.37)
14.78 (14.89)
13.21 (13.04)
[Ni(PMPP-SAL)(Py)
3
]
[
Cu(PMPP-SAL)Py]×MeOH
548
J. Therm. Anal. Cal., 89, 2007

DECOMPOSITION OF MIXED-LIGAND Ni(II) AND Cu(II) COMPLEXES
–
1
Table 2 IR bands assignment for ligand and the complexes/cm
Compound
PMPP-SAL
n
O–H
n
N–H
n
C=O
n
C– O–
n
C=N
n
C=N Py
n
C–OH
n
M–O
n
M–N
3249 3249 1663, 1614
–
–
1308
1305
1301
–
–
[Ni(PMPP-SAL)(Py)
3
]
3176
–
–
–
–
1445
1453
1599
1602
1576
1576
506
511
457
464
[
Cu(PMPP-SAL)Py]×MeOH 3313
second stage between 160 and 210°C corresponds to
decomposition of the remaining pyridine molecule and
one endotherm is observed in the DTA trace. The ob-
served mass loss is 11.81% which is consistent with
the theoretical value of 12.01%. With heating, the de-
composition of the ligand PMPP-SAL was followed.
The third stage, which occurs in the temperature
range 340–420°C with DTG peak at 388, 402°C, cor-
responds to the decomposition of the ligand, the ob-
served mass loss (52.12%) is roughly coincide with the
calculated value (52.20%). This step appears in the
DTA curve as a broad exotherm with Tmax=402°C and
–1
DH=12.9 kJ g . The final residue, estimated as nickel
oxide, has the observed mass 12.53% as against the
calculated value of 11.34%.
3
Fig. 2 The molecular structure of [Ni(PMPP-SAL)(Py) ] with
thermal ellipsoids shown at 30%
3
Table 3 Partial bond distances of [Ni(PMPP-SAL)(Py) ]
Bond
distance/
Bond
distance/
Bond
Bond
Ni1–O1
Ni1–N3
Ni1–N5
2.006(3)
1.996(3)
2.132(3)
Ni1–O2
Ni1–N6
Ni1–N7
2.034(3)
2.082(3)
2.148(3)
3
Fig. 1 The TG-DTG-DTA curves of [Ni(PMPP-SAL)(Py) ],
–
1
Therefore, the thermal decomposition process of
Ni(PMPP-SAL)(Py) ] complex can be expressed by
the following scheme:
heating rate: 5°C min
[
3
The
decomposition
procedure
of
Ni(PMPP-SAL)(Py) ] can also be demonstrated by
[
3
[
Ni(PMPP-SAL)(Py) ]®[Ni(PMPP-SAL)Py]®
3
the crystal structure of the complex (Fig. 2). With re-
gard to the bond distances, the Ni–N(pyridine)
[
Ni(PMPP-SAL)]®NiO
(
Ni–N5, Ni–N6 and Ni–N7) distances are longer than
[
Cu(PMPP-SAL)Py]×MeOH
other coordinating bond lengths (Table 3). In theory,
these bonds are less stable and easy to be broken
down. Therefore, the pyridine molecules were consid-
ered to be expelled before the decomposition of the
PMPP-SAL anion. Among the three pyridine mole-
cules, Ni–N5 and Ni–N7 bond lengths (2.132(3),
The thermal model of [Cu(PMPP-SAL)Py]×MeOH is
shown in Fig. 3 and it indicated that the thermal de-
composition takes place in three distinct stages. In the
first stage between 80 to 115°C, the noncoordinated
crystal methanol molecule moves as shown as one
maxima in the DTG curve at 107°C with a 4.63%
mass loss (calcd.: 5.96%). The second stage starts
at 145 and end at 212°C, the corresponding mass loss
2
(
.148(3) ) are longer than Ni–N6 bond length
2.082(3) ), indicating that the bond strengths of
pyridyl nitrogen atom coordinated to Ni(II) are differ-
ent and Ni–N6 is stronger than Ni–N5 and Ni–N7.
Thus, Ni–N6 bond was considered to be broken down
following Ni–N5 and Ni–N7 bonds. The TG curve
shows that two pyridine molecules were lost first and
the removal of the remaining pyridine molecule was
followed.
(
14.51%) is due to the degradation of one mole
pyridine molecule which is in agreement with the cal-
culated value (14.73%). DTA curve shows two endo-
therms at 108 and 199°C for the above two stages.
The third stage, which occurs in the temperature
range 315–405°C, corresponds to the decomposition
J. Therm. Anal. Cal., 89, 2007
549

XU et al.
Fig. 3 The TG-DTG-DTA curves of
Cu(PMPP-SAL)Py]×MeOH, heating rate: 5°C min
–1
[
of the ligand PMPP-SAL, and the observed mass loss
Fig. 4 A plot of E
a
and logA as a function of a for the first
] based on the OFW
(
(
64.45%) coincides with the theoretical value
64.50%). The DTA shows a large exothermic peak
stage of [Ni(PMPP-SAL)(Py)
3
model-free method using TG results. The error limits
specified by bars
at 383°C corresponding to this stage. Mass constancy
is attained at around 420°C. The end product, esti-
mated as CuO, has the observed mass of 15.52%
compared with the calculated value of 14.81%.
Kinetic TG analysis of [Ni(PMPP-SAL)(Py) ]
3
The TG data for the three decomposition stages of this
complex were analyzed by model-free method and
non-linear fit to 15 mechanism functions.
Kinetic analyses
The apparent activation energy of the decomposition
process in non-isothermal conditions can be calculated
by model-free isoconversional method [29–35]. The
model-free kinetic analyses have the advantage of alle-
viating the need to select a specific kinetic model and
hence any dependence on this choice. The Netzsch
thermokinetics software package has the capacity to
perform model-free kinetic analyses. According to our
experience the significance of the model-free analyses
lies in its function as preliminary stage of non-linear
regression. The determination of kinetic parameters by
means of the non-linear regression is an iterative pro-
cess. The preset of start values for the parameters is
necessary. Here the model-free estimation of the acti-
vation energy means a great help for the non-linear re-
gression for being able to deliver start values. Further-
more the multi-step feature of a process often can only
be detected from the dependence of the activation en-
ergy on the reaction degree [36].
Figure 4 shows that the activation energy E of
a
the first stage is nearly independent of the extension
of conversion a. Ozawa–Flynn–Wall model-free
analysis reveals that the average activation energy
–
1
value is about 89 kJ mol . This indicates that the first
decomposition stage is a single-step reaction, so sin-
gle-step reaction mechanism was selected. Consider-
ing fitting quality (characterized by correlation coef-
ficient and relative precision), the C1 B model is the
most suitable one for this reaction. The kinetic param-
eters and statistical characterization after the
multivariate non-linear regression are listed in Ta-
ble 4 and graphic presentation of the curve fitting are
shown in Fig. 5. It can be seen that the experimental
data and the non-linear regression model fit well.
The Ozawa–Flynn–Wall model-free analysis
shows that the average activation energy value of the
second stage varies little with the mass loss, and the
–
1
activation energy value is about 95 kJ mol . This in-
dicates that this stage is also a single-step reaction.
The fitted curves are showed in Fig. 6.
In this study, the Ozawa–Flynn–Wall model-free
analysis and multivariate non-linear regression were
applied to perform single and overall steps optimiza-
tion. Kinetic parameters were given and the most prob-
able mechanism functions were suggested (Table 4).
It is clear in Fig. 7 that the activation energy of
–
1
the third stage assumes a value of 137 kJ mol at the
beginning and, with increasing mass loss, it drops
Table 4 The optimized mechanism functions and kinetic parameters of two mixed-ligand complexes
–1
–1
lgA/s
Reaction
order n
Correction
coefficient
Complex
Stage Mechanism
E
a
/kJ mol
Exponent a
I
II
III
C1 B
Bna
Fn
89.1836
8.5676
10.2767
11.9930
12.1090
1.0000
3.6685
3.1691
7.5229
0.7703
0.99941
0.99935
0.99935
112.5905
176.7993
154.1651
[
[
3
Ni(PMPP-SAL)(Py) ]
Fn
I
Fn
Fn
Bna
361.5583
240.1268
128.3496
48.9240
31.3189
12.0453
0.7256
0.5876
1.1462
0.99957
0.99894
Cu(PMPP-SAL)Py]×MeOH
II
7.1947E-2
550
J. Therm. Anal. Cal., 89, 2007

DECOMPOSITION OF MIXED-LIGAND Ni(II) AND Cu(II) COMPLEXES
Fig. 5 Fit of the first stage of the TG curve of
Ni(PMPP-SAL)(Py) ], simulated with reaction type
C1 B; signs – measured, — – calculated
Fig. 8 Fit of the third stage of the TG curve of
[Ni(PMPP-SAL)(Py) ]; signs – measured, — – calculated
[
3
3
In conclusion, the thermal decomposition of
complex [Ni(PMPP-SAL)(Py) ] occurs in three
3
stages. The Ozawa–Flynn–Wall model-free analysis
and multivariate non-linear regression method are
used to the overall decomposition process. The best
fitted models and kinetic parameters were deter-
mined. The first stage adopts the single-step reaction
model, and the reaction mechanism is C1 B, relevant
–
1
E is 89.1836 kJ mol . The value agrees well with
a
that from OFW analysis. The second stage also adopts
A®B single-step reaction model and the reaction
mechanism function is Bna. The third stage adopts the
A®B®C reaction model and the mechanism func-
tion is Fn–Fn.
Fig. 6 Fit of the second stage of the TG curve of
3
[Ni(PMPP-SAL)(Py) ]; signs – measured, — – calculated
Kinetic TG analysis of [Cu(PMPP-SAL)Py]×MeOH
The stages of the removal of methanol molecule and
pyridine molecule were analyzed by the multivariate
non-linear regression analyses. The start values for
the parameters were taken from the OFW model-free
analyses. The 15 reaction mechanism functions were
tested to fit the curves.
According to the Ozawa–Flynn–Wall model-free
analysis result, the activation energy for the first stage
–1
assumes a value of 310 kJ mol at the beginning and at
–1
the end it drops to a value of 256 kJ mol . The E de-
a
creases with extent of conversion a indicates that the
decomposition reaction contains at least two steps. So
a two-step consecutive reaction A®B®C was se-
lected. Multivariate non-linear-regression method
shows Fn–Fn model is the best fit model with highest
correlation coefficient. Figure 9 shows the experimen-
tal data and the non-linear regression model fit well.
The calculated apparent activation energies are consis-
tent with Ozawa–Flynn–Wall model-free analysis.
Figure 10 shows that the activation energy and
pre-exponential factor of the second stage are inde-
pendent of the extension of conversion, the activation
Fig. 7 A plot of E
a
and logA as a function of a for the third
] based on the OFW
stage of [Ni(PMPP-SAL)(Py)
3
model-free method using TG results. The error limits
specified by bars
–
slowly to a value of 57 kJ mol . This dependence of
1
the activation energy E on a (fractional mass loss) is
a
an indication that the overall reaction contains two
steps. Hence two-step mechanism model was used to
fit it. In this work Fn–Fn model is determined as the
most-probable model. The calculated curves were fit-
ted to the experimental ones (Fig. 8).
–
1
energy value is about 110 kJ mol . This clearly indi-
cates that the second stage is a single-step reaction. So
J. Therm. Anal. Cal., 89, 2007
551

XU et al.
nonlinear regression method. The optimized mecha-
nism function of first stage is A®B®C sequential re-
action model, the mechanism function is Fn–Fn, rele-
–
1
vant E are 361.5583, 240.1268 kJ mol , and logA
a
–
1
are 48.9240 and 31.3189 s . The calculated apparent
activation energies of the fitted model are consistent
with those calculated by isoconversional method us-
ing original data. The second stage adopts single-step
reaction model and the reaction mechanism function
is Bna. We tried to use single-step and multi-step
mechanism model to fit the third stage of decomposi-
tion reaction, unfortunately, a high quality of fit has
not achieved.
Fig. 9 Fit of the first stage of the TG curve of
Cu(PMPP-SAL)Py]×MeOH; signs – measured,
– calculated
[
—
Acknowledgements
This work was supported by the National Natural Science
Foundation of China (No. 20366005 and 20462007) and Sci-
entific Research Foundation in Xinjiang Educational Institu-
tions (XJEDU2005S01).
References
1
N. Raman, A. Kulandaisamy, A. Shunmugasundaram and
K. Jeyasubramanian, Transit. Met. Chem., 26 (2001) 131.
T. Yoshikuni, J. Mol. Catal. A-Chem., 148 (1999) 285.
B. A. Uzoukwu, P. U. Adiukwu, S. S. Al-Juaid,
2
3
Fig. 10 Dependence of E and logA vs. a for the second stage
a
P. B. Hitchcock and J. D. Smith, Inorg. Chim. Acta,
of [Cu(PMPP-SAL)Py]×MeOH based on the OFW
method
2
50 (1996) 173.
Z. Y. Yang, R. D. Yang, F. S. Li and K. B. Yu, Polyhedron,
9 (2000) 2599.
4
5
6
7
8
1
W. F. Yang, S. G. Yuan, Y. B. Xu, Y. H. Xiao and
K. M. Fang, J. Radioanal. Nucl. Ch., 256 (2003) 149.
A. K. El-Sawaf and D. X. West, Transition Met. Chem.,
23 (1998) 417.
N. Kalarani, S. Sangeetha, P. Kamalakannan and
D. Venkappayya, Russ. J. Coord. Chem., 29 (2003) 845.
C. Pettinari, F. Marchetti, C. Santini, R. Pettinari,
A. Drozdov, S. Troyanov, G. A. Attiston and R. Gerbasi,
Inorg. Chim. Acta, 315 (2001) 88.
9
F. Marchetti, C. Rettinar, R. Pettinari, A. Cingolani,
D. Leonesi and A. Lorenzotti, Polyhedron, 18 (1999) 3041.
1
0 O. N. Kataeva, A. T. Gubaidullin, I. A. Litvinov,
O. A. Lodochnikova, L. R. Islamov, A. I. Movchan and
G. A. Chmutova, J. Mol. Stuct., 610 (2002) 175.
Fig. 11 The second stage of the [Cu(PMPP-SAL)Py]×MeOH
decomposition was the best fit to mechanism function;
signs – measured, — – calculated
11 Y. Akama and A. Tong, Microchem. J., 53 (1996) 34.
12 Y. Akama, A. Tong, N. Matsumoto, T. Ikeda and
S. Tunaka, Vib. Spectrosc., 13 (1996) 113.
1
1
1
1
3 M. F. Iskander, L. Sayed, A. F. M. Hefny and S. E. Zayan,
J. Inorg. Nucl. Chem., 38 (1976) 2209.
4 G. H. Havanur and V. B. Mahale, Indian J. Chem.,
single reaction mechanism was used to fit, the results
were satisfied (Fig. 11).
26A (1987) 1063.
In conclusion, the thermal decomposition of
complex [Cu(PMPP-SAL)Py]×MeOH was studied in
air using TG-DTG-DTA analyses. TG-DTA curves
show that the decomposition occurs in three stages.
The most-probable kinetic models of the first and sec-
ond stages have been estimated with multivariate
5 H. Adams, D. E. Fenton, G. Minardi, E. Mura and
M. Angelo, Inorg. Chem. Commun., 3 (2000) 4.
6 X. C. Tang, D. Z. Jia, K. B. Yu, X. G. Zhang, X. Xia and
Z. Y. Zhou, J. Photochem. Photobiol. A: Chem.,
134 (2000) 23.
552
J. Therm. Anal. Cal., 89, 2007

DECOMPOSITION OF MIXED-LIGAND Ni(II) AND Cu(II) COMPLEXES
1
1
1
2
2
7 L. Liu, D. Z. Jia, Y. L. Ji and K. B. Yu, J. Photochem.
28 S. K. Padhi, Thermochim. Acta, 448 (2006) 1.
29 H. L. Friddman, J. Polym. Sci., 6 (1965) 183.
30 T. Ozawa, Bull. Chem. Soc. Jpn., 38 (1965) 1881.
31 K. Uemura, S. Kitagawa, K. Saito, K. Fukui and
K. Matsumoto, J. Therm. Anal. Cal., 81 (2005) 529.
32 V. Alvarez, E. Rodriguez and A. Vazquez,
J. Therm. Anal. Cal., 85 (2006) 383.
Photobiol. A: Chem., 154 (2003) 117.
8 L. Liu, D. Z. Jia and Y. L. Ji, Synth. React. Inorg.
Met.-Org. Chem., 32 (2002) 739.
9 Y. L. Ji, L. Liu, D. Z. Jia and K. B. Yu, Chinese J. Inorg.
Chem., 19 (2003) 345.
0 L. Zhang, L. Liu, D. Z. Jia and K. B. Yu, Struct. Chem.,
15 (2004) 327.
33 L. S. Guinesi, C. A. Ribeiro, M. S. Crespi, A. F. Santos
and M. V. Capela, J. Therm. Anal. Cal., 85 (2006) 301.
34 S. Vyazovkin, Thermochim. Acta, 355 (2000) 155.
35 P. žimon, J. Therm. Anal. Cal., 76 (2004) 123.
36 J. R. Opfermann, E. Kaisersberger and H. J. Flammersheim,
Thermochim. Acta, 391 (2002) 119.
1 L. Zhang, L. Liu, D. Z. Jia, G. C. Xu and K. B. Yu,
Inorg. Chem. Commun., 7 (2004) 1306.
2
2
2 B. S. Jensen, Acta Chem. Scand., 13 (1959) 1668.
3 R. Q. Huang, H. L. Wang and J. Zhou, Preparation of
Organic Intermediates, Chemistry Industrial Press,
China 1997, p. 109.
2
4 D. Dollimore, P. Tong and K. S. Alexander,
Thermochim. Acta, 282–283 (1996) 13.
Received: February 16, 2005
Accepted: January 31, 2007
2
2
5 J. Opfermann, J. Therm. Anal. Cal., 60 (2000) 641.
6 V. Logvinenko, A. Minina, Y. Mikhaylov, Y. Yukhin and
B. Bokhonov, J. Therm. Anal. Cal., 74 (2003) 407.
7 J. H. Hong, T. Y. Yi, J. Min, C. J. Cong and K. L. Zhang,
Thermochim. Acta, 440 (2006) 31.
DOI: 10.1007/s10973-005-6947-6
2
J. Therm. Anal. Cal., 89, 2007
553