Non-isothermal decomposition of NiC2O4-FeC2O4 mixture aiming at the production of NiFe2O4

Article abstract of DOI:10.1016/S0022-3697(03)00163-X

The non-isothermal decomposition of NiC₂O₄·2H₂O-FeC₂O₄ ·2H₂O (1:2 mole ratio) mixture was studied on heating to the formation of NiO-Fe₂O₃ mixture at 350 °C in air atmosphere using thermogravimetry. Kinetic analysis of data according to the integral composite method showed that the oxidative decomposition of FeC₂O₄ and NiC₂O₄ are best described by the three-dimensional phase boundary model. The activation parameters were calculated and discussed. The solid products at different decomposition stages were identified using XRD, Mo?ssbauer and FT-IR spectroscopic techniques. Some characteristic XRD lines of NiFe₂O₄ start to appear at 800 °C beside the characteristic lines of NiO and Fe₂O₃, whereas at 1000 °C, only the characteristic lines of single phase cubic NiFe₂O₄ appeared. The Mo?ssbauer spectrum at 1000 °C fitted into two Zeeman sextets characteristic of Fe³⁺ on the tetrahedral (A) and octahedral (B) sites of NiFe₂O₄ inverse spinel. Consistent results were obtained using FT-IR where the absorption bands appeared at 602 and 407 cm^-1 for the mixture calcined at 1000 °C. These can be assigned to the intrinsic vibrations of tetrahedral and octahedral sites of NiFe₂O₄, respectively.

Full text of DOI:10.1016/S0022-3697(03)00163-X

Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
www.elsevier.com/locate/jpcs
Non-isothermal decomposition of NiC2O4–FeC2O4 mixture
aiming at the production of NiFe2O4
*
M.A. Gabal
Department of Chemistry, Faculty of Science, Benha University, Benha, Egypt
Received 21 January 2003; accepted 10 March 2003
Abstract
The non-isothermal decomposition of NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole ratio) mixture was studied on heating to the
formation of NiO–Fe2O3 mixture at 350 8C in air atmosphere using thermogravimetry. Kinetic analysis of data according to the
integral composite method showed that the oxidative decomposition of FeC2O4 and NiC2O4 are best described by the three-
dimensional phase boundary model. The activation parameters were calculated and discussed. The solid products at different
¨
decomposition stages were identified using XRD, Mossbauer and FT-IR spectroscopic techniques. Some characteristic XRD
lines of NiFe2O4 start to appear at 800 8C beside the characteristic lines of NiO and Fe2O3, whereas at 1000 8C, only the
¨
characteristic lines of single phase cubic NiFe2O4 appeared. The Mossbauer spectrum at 1000 8C fitted into two Zeeman sextets
characteristic of Fe3þ on the tetrahedral (A) and octahedral (B) sites of NiFe2O4 inverse spinel. Consistent results were obtained
using FT-IR where the absorption bands appeared at 602 and 407 cm21 for the mixture calcined at 1000 8C. These can be
assigned to the intrinsic vibrations of tetrahedral and octahedral sites of NiFe2O4, respectively.
q 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
The specific physical behaviour of NiFe2O4 can be
considered as a function of its specific chemical and
structural properties. For this reason, many researchers
have paid attention to the different methods of its synthesis.
Music et al. [4] investigated the chemical and structural
properties of the system Fe2O3–NiO prepared by the
thermal treatment of co-precipitated hydroxides or the
Ferrites are a broad class of complex magnetic oxides of
considerable technological importance. Their remarkable
electrical and magnetic properties depend up on the nature
of the ions, their charges and their distribution among
tetrahedral (A) and octahedral (B) sites [1].
solid-state reaction between NiO and a-Fe2O3 using X-ray
¨
Nickel ferrite is an inverse spinel in which eight units of
NiFe2O4 go into a unit cell of the spinel structure. Half of the
ferric ions preferentially fill the tetrahedral sites (A-sites)
and the other occupies the octahedral sites (B-sites).
Thus, the compound can be represented by the formula
57
diffraction, FT-IR and Fe Mossbauer spectroscopic
techniques. They found that the heating temperature rather
than the heating time had the ultimate effect on the
microstructure of NiFe2O4. Elmasry et al. [5] prepared
nickel ferrite through the thermal decomposition of mixed
nickel(II) and iron(III) nitrates at 1500 8C. Nanometre size
NiFe2O4 particles were synthesized by the citrate precursors
technique. Control of solid-state reactivity yields mono-
phasic ultrafine NiFe2O4. It was characterized by X-ray
diffraction and BET surface area measurements [6]. Sol–gel
method [7] using polyacrylic acid as a chelating agent was
used to prepare NiFe2O4 nanoparticles by calcining of the
gel precursors in air at 300 8C for 2 h. The formation rate of
3þ
2þ 3þ
(Fe )[Ni Fe ]O242, where the round and square brackets
1.0
1.0 1.0
represent A and B sites, respectively [1]. Nickel ferrite
proved to be a good catalyst for the benzoylation of toluene
with benzoyl chloride [2]. Also, it has a peculiar and
extraordinary gas sensing capabilities towards low concen-
trations of chlorine gas in air [3].
*
Tel.: þ20-1010-48600; fax: þ20-132-22578.
E-mail address: mgabalabdo@yahoo.com (M.A. Gabal).
0022-3697/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0022-3697(03)00163-X

1376
M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
NiFe2O4 prepared by mechanical alloying process of Fe2O3/
NiO simple oxide powder system using high energy ball
milling has been investigated using X-ray diffraction and
to assure complete homogeneity. The wetted mixture was
then dried at 100 8C for about 2 h.
Samples of the oxalate mixture were inserted into an
electrical oven at room temperature, then the temperature
was raised to the desired value (obtained from DTA-TG
studies) and maintained for specified intervals. Samples
were then removed and cooled in a desiccator to room
temperature.
¨
Mossbauer spectroscopy [8]. Shock wave treatment of
nanosized mixture powders of ferric oxide and nickel oxide
prepared by the co-precipitation method yields nanosized
nickel ferrite with increased magnetization than that
prepared by the conventional calcinations process [9].
The thermal decomposition of oxalates results in the
production of oxides or metal possessing the characteristic
necessary for its function as reactive solids. Therefore, much
research has been done on their structures, properties and
thermal behaviour. Various mechanisms of decomposition
have been proposed [10]. A considerable amounts of
information on the kinetics of the thermal dehydration and
decomposition of the individual as well as mixed metal
2.2. Apparatus
Simultaneous DTA-TG experiments were performed
using a Shimadzu DT-40 thermal analyser. The experiments
were performed at different specified heating rates of 1, 2, 3
and 5 8C min21 in air at a flowing rate of 3.0 lh21. The
sample mass in the Pt cell was kept at about 8 mg in all
experiments, in order to assure a linear heating rate and
accurate temperature measurements.
oxalates have been reported [11]. Diefallah et al. used DTA-
¨
57
TG, XRD and Fe Mossbauer spectroscopy to study the
thermal decomposition reactions in air of CdC2O4–FeC2O4
[12] and MgC2O4–FeC2O4 [13] mixtures both in with 1:2
mole ratio. They found that the ferrite formation starts to
appear at 800 8C.
In the present work, a solid-state mixture of NiC2O4·2-
H2O–FeC2O4·2H2O (1:2 mole ratio) mixture with the aim
of the production of NiFe2O4 spinel was prepared using the
impregnation technique. The decomposition products at the
different decomposition stages as well as the spinel ferrite
formation were characterized using DTA-TG, XRD, FT-IR
XRD patterns were obtained at room temperature using a
Philips PW 1710 diffractometer which was equipped with
Ni-filter and generates a beam of Co K a radiation ðl ¼
ꢀ
1:7889 AÞ: The operational setting for the all XRD scans are
voltage: 40 kV; current: 30 mA; 2u ¼ range 20–808; scan-
ning speed ¼ 58 min21 and split width ¼ 0.028. For
identification purposes, the relative intensities ðI=I0Þ and
˚
the d-spacing (A) were compared with standard diffraction
patterns of the ASTM powder diffraction files.
57
¨
Mossbauer spectra were recorded at room tempera-
and 57Fe Mossbauer spectroscopy. Mossbauer spectroscopy
was also used to determine the cation distribution at both A
and B sites in the system. The kinetics of the non-isothermal
oxalate decomposition processes was studied in air.
The composite method of kinetic integral analysis was
used both to identify the mechanistic equation describing
the decomposition reaction and to calculate the activation
parameters. The results were compared and discussed with
respect to other integral methods of analysis of the non-
isothermal data.
¨
¨
¨
ture using an Austin Science Mossbauer effect spec-
trometer. A 57Co–Rh source (Amersham) was used with
an initial activity of 50 mCi. a-Fe spectrum is used for
calibration of both observed velocities and hyperfine
magnetic fields. The absorber thickness is approximately
10 mg cm22 of natural iron. The hyperfine interaction
parameters were computed using an interactive least
square ‘Mos-90’ program [14].
FT-IR spectra were obtained at a resolution 4 cm21
,
between 4000 and 200 cm21, using a model FT-IR 310
Jasco spectrometer (Japan). The specimens were pressed in
KBr discs. The spectra were represented as relative
transmittance versus wave number (cm21).
2. Experimental procedure
2.1. Materials
3. Kinetic consideration
Individual metal oxalates, NiC2O4·2H2O or FeC2O4·2-
H2O were prepared by co-precipitation method. Analyti-
cally pure reagents of NiCl2·6H2O of FeSO4·7H2O were
dissolved in bidistilled water, and an equivalent amount
of analytical reagent oxalic acid was added till complete
precipitation occurred. The fine precipitate obtained in
each case was filtered, washed with distilled water and
dried.
The kinetics of the non-isothermal decomposition is
considered here in view of three integral methods: the
Ozawa method [15], the Coats–Redfern method [16] and the
Diefallah composite method [12,13,17], based either on the
modified Coats–Redfern equation [18]; (composite method
I) or on Doyle’s equation [19]; (composite method II).
According to the composite method I, when use was
made of the modified Coats–Redfern equation [18]
NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole ratio) mixture
was prepared by the impregnation technique [18,19] in
which few drops of bidistilled water was added to the proper
mole ratios of the two metal oxalates with vigorous stirring
gðaÞ
AR
¼
e2E=RT
;
ð1Þ
T2
bE

M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
1377
the equation was rewritten in the form and the frequency factor is calculated from the equation
ꢀ
ꢂ
ꢃꢁ
ꢀ
ꢁ
ꢂ
ꢃ
E
E
bgðaÞ
AR
E
E
log A ¼ log gðaÞ 2 log
p
bR RT
:
ð6Þ
ln
¼ ln
2
;
ð2Þ
T2
RT
In the Coats–Redfern method [16], the function gðaÞ is
where gðaÞ is the kinetic model function calculated for the
fraction reacted a at temperature T and heating rate b; and is
given by the equation
approximated to the form:
ꢀ
ꢁ
ART2
2RT
gðaÞ ¼
1 2
e2E=RT
:
ð7Þ
bE
E
ðT
gðaÞ
A
R
¼
expð2E=RTÞdT;
ð3Þ
T2
This equation has been written in the form:
ꢀ
0
ꢁ
ꢀ
ꢁ
gðaÞ
AR
2RT
E
where A is the frequency factor and E is the activation
energy. The dependence of ln½bgðaÞ=T2ꢀ; which was
calculated for different a-values at their respective b-
values, on 1=T must give rise to a single master straight line
for the correct form of gðaÞ; and hence a single activation
energy and frequency factor can be calculated readily. In a
similar way, according to composite method II, the Doyle’s
equation has been rewritten in the form:
2ln
¼ 2ln
1 2
þ
:
ð8Þ
T2
bE
E
RT
The quantity lnðAR=bEÞð1 2 2RT=EÞ is reasonably con-
stant for most values of E and in the temperature range over
which most reactions occur.
4. Results and discussion
ꢀ
ꢁ
AE
R
E
logðbgðaÞÞ ¼ log
2 2:315 2 0:4567
:
ð4Þ
4.1. DTA-TG studies
RT
Thermal decomposition studies of NiC2O4·2H2O–FeC2-
O4·2H2O (1:2 mole ratio) mixture show that the mixture
dehydrate and decompose through three well defined steps
in the temperature range from 140 to 350 8C. A typical
DTA-TG thermogram measured at 5 8C min21 is shown in
Fig. 1. The thermogravimetry (TG) curve implies the
dehydration of the mixture at 140–190 8C with a mass
loss of 19.0% compared with expected mass loss of 19.9%
corresponding to the complete dehydration of the mixture
with the formation of anhydrous oxalate mixture, NiC2O4–
FeC2O4. This was followed immediately by a mass loss of
24.5% up to about 250 8C attributed to the oxidative
decomposition of FeC2O4 in the mixture with the formation
of NiC2O4–Fe2O3 mixture (theoretical mass loss, 23.6%).
The mixture is thermally stable up to a temperature of
290 8C after which the nickel oxalate content starts to
decompose giving a mass loss of 13.5% at 330 8C. This
Again, the dependence of log gðaÞb; calculated for the
different a-values at their respective b-values on 1=T must
give rise to a single master straight line for the correct form
of gðaÞ: A computer program [20] has been written to
perform the data analysis using the different models of
heterogeneous solid-state reactions [17] listed in Table 1.
The results of the calculation allow choosing the kinetic
mechanism which best fit the data and give the highest
correlation coefficient and the lowest standard deviation.
The program also calculates the activation energy ðEÞ and
the frequency factor ðAÞ from the slope and intercept of the
linear fit line.
In the Ozawa method [15] the activation energy is
calculated from the thermogravimetric data obtained at
different heating rates using the equation
2logðbÞ ¼ 0:4567ðE=RTÞ þ constant
ð5Þ
Table 1
Kinetic equations examined in this work
Reaction model
gðaÞ
Symbol
One-dimensional diffusion
a2
D1
D2
D3
D4
R2
R3
F1
Two-dimensional diffusion
a þ ð1 2 aÞ lnð1 2 aÞ
1=3
2
Jander equation, three-dimensional diffusion
Ginsling–Braunshtein equation, three-dimensional diffusion
Two-dimensional phase boundary reaction
Three-dimensional phase boundary reaction
First order kinetic, Mampel unimolecular law
Random nucleation: Avrami equation
Random nucleation: Erofeev equation
Exponential law
½1 2 ð1 2 aÞ ꢀ
2=3
ð1 2 2=3aÞ 2 ð1 2 aÞ
1=2
½1 2 ð1 2 aÞ
½1 2 ð1 2 aÞ
ꢀ
ꢀ
1=3
½2lnð1 2 aÞꢀ
½2lnð1 2 aÞꢀ
½2lnð1 2 aÞꢀ
ln a
1=2
A2
A3
E1
1=3

1378
M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
Fig. 1. DTA-TG curves of NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole ratio) mixture in air at heating rate of 5 8C min21
.
makes the total mass loss to 57.0%, which is close to that
expected (56.8%) for the over all conversion of NiC2O4·2-
H2O–FeC2O4·2H2O (1:2 mole ratio) mixture to NiO–Fe2O3
mixture.
stages and following the ferrite formation at 30 min at
250 8C, 30 min at 350 8C, 2 h at 600 8C, 2 h at 800 8C and
2 h at 1000 8C. For simplicity, these different calcined
mixtures are denoted in the text by NiFe, followed by the
calcination temperature. Thus NiFe-250 indicates a mixture
calcined at 250 8C.
DTA curve shows peaks that closely correspond to the
mass loss observed on the TG trace. The two broad
overlapped endothermic peaks in the temperature range
150–190 8C with peak’s temperature at 170 and 183 8C are
attributed to the successive dehydration of NiC2O4·2H2O and
FeC2O4·2H2O in the mixture. These dehydration tempera-
tures suggest that the water molecules present in the
crystalline hydrate as co-ordinately linked as well as a
crystal water [17]. The two broad overlapped exothermic
DTA peaks in the temperature range of 200–240 8C (at 213
and 220 8C) accompanied the FeC2O4 decomposition step,
which suggests that the decomposition is through two steps.
Macklen [21] proposed that the oxidation of ferrous ion was
the first step in the decomposition of ferrous oxalate in an
oxidizing atmosphere, and this was immediately followed by
bond rupture of the relatively unstable oxidation product to
evolve CO2. The broad exothermic peak in the temperature
range 305–324 8C with the peak temperature at 318 8C
characterizes the thermal decomposition step of NiC2O4. The
exothermic nature of the decomposition is attributed to the
oxidation of CO evolved on decomposition to CO2 [10].
From the above results it was obvious that the
decomposition behaviour of NiC2O4·2H2O or FeC2O4·2H2O
in their mixture is similar to that of the individual metal
oxalates [10,21] which indicate that each metal oxalate
behaves as it was present alone in the mixture, and that no
chemical reactions can takes place between the two metal
oxalates. Also, DTA-TG curves gives information about the
minimum temperatures needed for the calcination pro-
cesses. Consequently, samples of the mixture were calcined
to characterize the decomposition products at different
4.2. X-ray diffraction
The characteristic parts of the X-ray powder diffraction
patterns of (NiFe) mixture annealed at various temperatures
are shown in Fig. 2. The parent mixture, NiFe gave the
individual characteristic lines of both NiC2O4·2H2O
(JCPDS file No. 25-582) and FeC2O4·2H2O (JCPDS file
No. 23-293). A nearly amorphous pattern was obtained for
the NiFe-250 mixture (Fig. 2(b)). This behaviour is
attributed to the formation of fine grained (or possibly
even amorphous) iron(III) oxide. For the NiFe-350 mixture
(Fig. 2(c)), the obvious broadening of the all diffraction
peaks suggests that the size of iron(III) oxide particles is still
relatively small. The increase in the annealing temperature
yields an increased sharpness of the major peaks, that is, the
growth of the grain size of iron(III) oxide. This behaviour is
clear in the XRD of NiFe-600 mixture (Fig. 2(d)) which
shows the characteristic XRD lines of Fe2O3 with larger
grain size (JCPDS file No. 13-534) and NiO (JCPDS file No.
4-835). The unit cell parameters [22] of NiFe2O4 is very
close to twice as large as that of NiO. Both compounds
posses the fcc crystal structure and for this reason, NiO does
not exhibit independent diffraction lines. No separation
between overlapped diffraction lines of these two com-
pounds was detected for sample where both compounds
were present even at high Bragg angles. For such samples
the changes in the fraction of NiO and NiFe2O4 can be
followed by comparison of the intensities of overlapped

M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
1379
Fig. 2. Characteristic parts of XRD patterns of NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole ratio) mixture calcined at different temperatures. (a)
parent mixture, (b) mixture calcined at 250 8C, (c) mixture calcined at 350 8C, (d) mixture calcined at 600 8C, (e) mixture calcined at 800 8C and
(f) mixture calcined at 1000 8C. Phases: (K) NiC2O4·2H2O, (þ) FeC2O4·2H2O, (W) Fe2O3, ( £ ) NiO and (X) NiFe2O4.
diffraction lines belonging to both compounds with the
intensities of diffraction lines belonging only to NiO. Thus,
comparing the XRD patterns of NiFe-600 and NiFe-800
mixtures (Fig. 2(d) and (e), respectively) it is clear that, the
new two peaks appear in the pattern of NiFe-800 together
with Fe2O3 and NiO peaks are attributed to the formation of
spinel ferrite, NiFe2O4 (JCPDS file No. 10-325). Also, the
overlapping of NiFe2O4 peaks with NiO peaks leads to an
increase in the peak’s intensities compared to NiFe-600
pattern. X-ray diffraction pattern of NiFe-1000 mixture

1380
M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
¨
4.3. Mossbauer spectroscopy
(Fig. 2(f)) shows only the characteristic peaks of the single-
phase ferrite, NiFe2O4, crystallized in the cubic system. At
the same time, no peaks attributed to Fe2O3 and/or NiO were
detected.
¨
The room temperature Mossbauer spectra of NiFe
mixture annealed at different temperatures are shown in
Fig. 3. The spectral parameters such as isomer shift ðdÞ;
quadrupole splitting ðDEQÞ and hyperfine magnetic split-
ting ðHÞ are computed and summarized in Table 2.
Spectrum of NiFe mixture (Fig. 3(a)) is fitted with a
doublet that is assigned to divalent iron in FeC2O4·2H2-
¨
O. The obtained Mossbauer parameters are in good
agreements with the values reported by Brady and Duncan
[23]. Spectrum of Nife-250 and NiFe-350 mixtures
(Fig. 3(b) and (c), respectively) are fitted with a doublet
with a very small magnetically split pattern and a sextet
subspectra, which are characteristic of iron(III) oxide
having two different magnetic states. The doublet is
owing to the paramagnetic finely divided (nearly amor-
phous) iron(III) oxide, while the sextet is attributed to
magnetically oriented iron(III) oxide. Diefallah et al. [17]
have defined a similar paramagnetic behaviour in ultrafine
a-Fe2O3 particles prepared by calcination of CdC2O4–
FeC2O4 mixture at 250 8C. The ratio of the different
magnetic phases is based on the calculated absorption ratio
of the two subspectra. In agreement with the previous XRD
results, which show an evolution in the crystallinity of
Fe2O3 as the temperature increased, the ratio of the
paramagnetic phase (Table 2) was observed to decrease
with rising the annealing temperature. The hyperfine
magnetic splitting obtained at 350 8C (497 kOe) is
significantly less than the reported value [24], for iron(III)
oxide with a larger particle size, of 515 kOe. The
paramagnetic component disappears completely by raising
the annealing temperature to 600 8C (Fig. 3(d)). The
hyperfine magnetic splitting of the obtained sextet
(510 kOe) is very close to the normal value of iron(III)
¨
oxide. The Mossbauer spectrum of NiFe-800 mixture
Table 2
¨
Mossbauer parameters for NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole
ratio) mixture calcined at different temperatures
Mixture
Sub-
Area Isomer
Quadrupole Hyperfine
spectra (%) shift ðdÞ splitting
filed ðHÞ
(^0.02
ðDEQÞ
(^5 kOe)
mm s21
)
(^0.02
mm s21
)
NiFe
Doublet 100 1.26
1.70
20.68
0.49
–
NiFe-250
Doublet
Sextet
23 0.41
77 0.51
0.40
–
546
–
NiFe-350
Doublet
Sextet
6
20.45
20.20
20.21
20.19
0.06
94 0.44
100 0.44
85 0.44
15 0.27
49 0.44
51 0.32
497
510
512
486
488
518
NiFe-600
NiFe-800
Sextet
¨
Fig. 3. The Mossbauer spectra of NiC2O4·2H2O–FeC2O4·2H2O (1:2
Sextet
mole ratio) mixture calcined at different temperatures: (a) parent
mixture, (b) mixture calcined at 250 8C, (c) mixture calcined at
350 8C, (d) mixture calcined at 600 8C, (e) mixture calcined at
800 8C and (f) mixture calcined at 1000 8C.
(A) site
NiFe-1000 (A) site
(B) site
0.01
0.01

M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
1381
(Fig. 3(e)) is fitted into two six-line patterns. The first sextet is
assigned to iron(III) oxide as revealed from XRD, and the
other one is attributed to Fe3þ ions at one of the two distinct
crystallographic sites of NiFe2O4. It can be assigned to the
tetrahedral (A) site on the bases of the isomer shift and the
hyperfine magnetic splitting values obtained (Table 2). In
most magnetically ordered spinel ferrites the hyperfine
magnetic splitting due to Fe3þ ions at A sites is usually
smaller than that of Fe3þ ions at B sites [25]. The spectrum of
NiFe-1000 mixture (Fig. 3(f)) exhibits the normal two clearly
split Zeeman sextets. One is due to Fe3 ions at the tetrahedral
(A) sites, and the other is due to Fe3þ ions at the octahedral
(B) sites. The relative number of Fe3þ ions on A and B sites is
determined by knowing the area under the resonance lines
due to these ions. Since the spectrum is well resolved into two
Zeeman sextets with nearly equal areas of A and B subspectra
(Table 2), then thecationdistribution maybedescribed bythe
3þ
2þ 3þ
chemical formula: (Fe )[Ni Fe ]O422
.
1.0
1.0 1.0
Fig. 4. FT-IR spectra of NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole ratio) mixture calcined at different temperatures: (a) parent mixture, (b)
mixture calcined at 250 8C, (c) mixture calcined at 350 8C, (d) mixture calcined at 600 8C, (e) mixture calcined at 800 8C and (f) mixture
calcined at 1000 8C.

1382
M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
Waldron [30] attributed the band n1 at around 600 cm21 to
4.4. FT-IR spectral studies
the intrinsic vibration of tetrahedral metal oxygen com-
plexes and the band n2 at around 400 cm21 to the intrinsic
vibration of octahedral metal oxygen complexes. Thus, the
band appears at 596 cm21 is assigned to the intrinsic
vibration of the tetrahedral complex in agreement with the
The present FT-IR spectroscopic results obtained for
NiFe mixture calcined at different temperatures were
summarized in Fig. 4. Consistent with the presence of the
oxalate moiety in the parent mixture; NiFe (Fig. 4(a)), the
bands appeared (at 1317 and 1362 cm21) and (at
1632 cm21) are assigned to O–C–O symmetric and
asymmetric stretches, respectively. The bands obtained at
812 and 493 cm21 are assigned to CyC–O and O–C–O
bending vibration, respectively [26]. The broad band at
3385 cm21 is assigned to the water of hydration in the
mixture. FT-IR spectrum of NiFe-250 mixture (Fig. 4(b))
shows the same bands as those appear in the parent mixture,
which are assigned to the presence of nickel oxalate.
Schwertmann and Taylor [27] suggested that the character-
istic IR bands at 540, 470 and 345 cm21 could be used as
fingerprints in the identification of hematite. The presence of
the foreign ions as well as the degree of crystallinity
strongly affects the location of these bands. Yariv and
Mendelovici [28] showed that three strong absorption bands
characterize the poorly crystalline hematite at 530, 445 and
308 cm21, which were assigned to O22 displacement. Thus,
the FT-IR bands that appear at 550, 450 and 325 cm21 in the
spectrum of NiFe-250 mixture (Fig. 4(b)) are assigned to
poorly crystalline iron(III) oxide. The band at 3400 cm21 is
assigned to the stretching O–H band which is attributed to
the absorption of moisture from the air by the fine granules
of iron(III) oxide [29]. The FT-IR spectrum of NiFe-350 and
NiFe-600 mixtures (Fig. 4(c) and (d), respectively) are
characterized by two dominant bands at 540 and 470 cm21
and a weak intensity band at 390 cm21. Two shoulders at
626 and 430 cm21 are also visible. These bands are typical
for a-Fe2O3 [22]. Also, the band observed in the range
450–475 cm21 was interpreted as the nickel–oxygen
stretching vibration. In the FT-IR spectrum of NiFe-800
mixture (Fig. 4(e)), the bands appear at 530, 460, 430 and
390 cm21 are assigned to Fe2O3. The appearance of the new
band 596 cm21 is attributed to the presence of NiFe2O4.
¨
Mossbauer results previously reported. Two dominant bands
at 602 and 407 cm21 are observed in the FT-IR spectrum of
NiFe-1000 mixture (Fig. 4(f)), due to the presence of
NiFe2O4 [22].
4.5. Kinetic studies
Analysis of dynamic data for the oxalate decomposition
steps in NiFe mixture (viz. the second and the third
decomposition steps) were carried out at heating rates of
1, 2, 3 and 5 8C min21. The fractional reaction remaining as
a function of temperature for the oxalate decomposition
reactions are shown in Fig. 5. The composite method of
analysis performed a complete analysis of all non-
isothermal curves into a single master curve. A deviations
from the straight-line relationship are interpreted in the
terms of multi-step reaction mechanisms [15,17]. The
method allows us to choose the kinetic model which best
fits the data, and gives the highest correlation coefficient and
the lowest standard deviation. The results always show that
[13,17] both composite methods equivalent curves and
nearly identical values for the activation parameters. The
values of the activation parameters for the non-isothermal
decomposition of FeC2O4 and NiC2O4 in their mixture that
were calculated according to the composite method II,
assuming the different kinetic models are reported in
Table 3. From the table, it is clear that, the oxalate
decomposition reactions are best described by the two- and
three-dimensional phase boundary and random nucleation
models. The exponential low and the other models gave less
satisfactory fit of data. Fig. 6 shows the results of the kinetic
data analysis for the oxalate decomposition reactions
calculated using the three-dimensional phase boundary
Fig. 5. Dynamic measurements for the thermal decomposition of NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole ratio) mixture in air, (a)
decomposition of FeC2O4 to Fe2O3 and (b) decomposition of NiC2O4 to NiO. Heating rate: curve A, 1 8C min21; curve B, 2 8C min21; curve C,
3 8C min21 and curve D, 5 8C min21
.

M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
1383
Table 3
Activation parameters of the non-isothermal decomposition in air of FeC2O4 and NiC2O4 in their mixture, calculated according to the composite
method II, assuming different kinetic models
Model
Decomposition of FeC2O4 (step II)
Decomposition of NiC2O4 (step III)
E (kJ mol21
)
log A (min21
)
rp
E (kJ mol21
)
log A (min21
)
rp
D1
D2
D3
D4
R2
R3
F1
223 ^ 12
245 ^ 13
264 ^ 14
253 ^ 13
134 ^ 5
140 ^ 5
153 ^ 7
81 ^ 1
23.0 ^ 1.4
24.9 ^ 1.5
27.0 ^ 1.5
25.1 ^ 1.5
13.3 ^ 0.6
13.8 ^ 0.6
15.8 ^ 0.6
8.3 ^ 0.4
0.879
0.891
0.905
0.896
0.935
0.972
0.956
0.920
0.818
0.625
357 ^ 22
370 ^ 24
407 ^ 29
388 ^ 26
221 ^ 8
227 ^ 9
244 ^ 12
150 ^ 8
120 ^ 6
98 ^ 29
30.5 ^ 2.1
32.2 ^ 2.3
34.1 ^ 2.6
32.4 ^ 2.4
18.7 ^ 0.8
19.1 ^ 0.9
21.0 ^ 1.1
12.7 ^ 0.5
9.9 ^ 0.6
0.862
0.849
0.832
0.843
0.940
0.970
0.914
0.957
0.898
0.285
A2
A3
E1
57 ^ 1
5.9 ^ 0.5
94 ^ 30
11.0 ^ 3.6
9.3 ^ 2.8
(R3) model according to the composite method based on
Doyle’s equation. Analysis of the dynamic data was also
carried out using the Ozawa and Coats–Redfern methods,
assuming also the R3 model, and the results were compared
with those obtained using the composite method of analysis.
The results of the different methods are shown in Table 4.
The three methods of calculation show similar activation
parameters for the thermal decomposition of FeC2O4. The
values calculated by the Coats–Redfern method for the
decomposition of NiC2O4 within experimental errors agree
with that calculated by the composite method. However,
these results do not show agreement with values calculated
by the Ozawa method. In many cases [13,29], the activation
energies estimated by the composite method are, within
Fig. 6. Composite analysis of TG data for the non-isothermal decomposition of NiC2O4·2H2O–FeC2O4·2H2O (1:2 mole ratio) mixture in air
based on Doyle’s equation assuming: (1) R3 model and (2) E1 model. (A) Decomposition of FeC2O4 and (C) decomposition of NiC2O4.

1384
M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
Table 4
Activation parameters of the non-isothermal decomposition in air of FeC2O4 and NiC2O4 in their mixture calculated according to R3 model
Method of analysis
Decomposition of FeC2O4 (step II)
Decomposition of NiC2O4 (step III)
E (kJ mol21
)
log A (min21
)
E (kJ mol21
)
log A (min21
)
Composite method II
Coats–Redfern
Ozawa
140 ^ 5
140 ^ 7
140 ^ 10
13.8 ^ 0.6
10.9 ^ 0.6
14.0 ^ 1.4
227 ^ 9
250 ^ 20
157 ^ 14
19.1 ^ 0.9
13.9 ^ 0.8
17.2 ^ 0.9
experimental errors, in good agreement with the values
estimated using the model-free method of Ozawa.
Acknowledgements
Dollimore [10] has been discussed the different
decomposition mechanisms for the metal oxalates and
found that for the oxalates of divalent metals, the extent to
which the metal–O bond is covalent depends on the
electronegativity of the metal. Decomposition will occur
when a temperature is reached at which rupture of the M–O
link is possible, or at which rupture of the C–O bond occurs.
For the oxalates which produce the metal in nitrogen
atmosphere, the decomposition temperature represents the
temperature at which the M–O bond is ruptured and will
depend critically on the size and charge of the metal ion. The
decomposition of NiC2O4 in nitrogen atmosphere gives Ni
metal, and the decomposition of FeC2O4 gives Fe, Fe3O4
and a trace of FeO. In air, the metal formed in the
decomposition is oxidized to metal oxide [11].
The author would like to thank Prof. Dr S.S. Ata-Allah,
Reactor and Neutron Physics Department, Nuclear Research
Center, Atomic Energy Authority for doing the Mo¨ssbauer
experiments. Thanks are also due to Mr Abdalla
O. Mahmoud, Central Metallurgical research and Develop-
ment Institute (CMRDI) for doing the X-ray diffraction
experiments.
References
[1] A. Goldman, Modern Ferrite Technology, Marcel Dekker,
New York, 1993.
[2] C.G. Ramankutty, S. Sugunan, Appl. Catal., A 218 (2001) 39.
[3] C.V.G. Reddy, S.V. Manorama, V.J. Rao, Sens. Actuators, B
55 (1999) 90.
Since the composite method of analysis gives the least
calculated standard deviation for the activation parameters
than the other methods (Table 4), then the data will be
discussed according to this method. The reported activation
energy for the thermal decomposition of FeC2O4 (Table 4)
is in agreement with that present in the literature [11–13].
The activation energy obtained (163 kJ mol21) from the
non-isothermal kinetic of the thermal decomposition of
NiC2O4 found by Palanismy et al. [31] using the Coats–
Redfern method is very small compared with that reported
in Table 4. This is due to the presence of Fe2O3 (produced
from the thermal decomposition of FeC2O4) in the vicinity
of NiC2O4. The electronegativity of Fe3þ (1.9) is higher
than that of Ni2þ (1.8) [32] so that the presence of Fe3þ
ions with NiC2O4 results in an increased negative on the
nickel ion. Thus a more covalent type of bond thus occurs
in the mixed oxalate than in pure NiC2O4. This results in
strengthening of Ni–O bond and an increase in the
activation energy of NiC2O4 decomposition. The electro-
negativity of Ni2þ has no effect on the decomposition of
FeC2O4 since both Ni2þ and Fe2þ has the same electro-
negativity [32].
[4] S. Music, D. Blazar, S. Popopvic, M. Gotic, I. Czako Nagy, S.
Dalipi, Croat. Chem. Acta 70 (1997) 719.
[5] M.A.A. Elmasry, A. Gaber, E.M.H. Khater, J. Therm. Anal.
Cal. (1998) 495.
[6] S. Prasad, N.S.G. Bhiye, J. Alloys Compd. 265 (1998) 87.
[7] D.H. Chen, X.R. He, Mater. Res. Bull. 36 (2001) 1369.
[8] J.Z. Jiang, L. Gerword, S. Morup, Mater. Sci. Forum. 312
(1999) 115.
[9] J. Lin, H. He, X. Jin, Z. Hao, Z. Hu, Mater. Res. Bull. 36
(2001) 2357.
[10] D. Dollimore, Thermochim. Acta 117 (1987) 331.
[11] A. Coetzee, M.E. Brown, D.J. Eve, C.A. Strydom, J. Therm.
Anal. 4 (1994) 357.
[12] El.-H.M. Diefallah, M.A. Gabal, A.A. El-Bellihi, N.A. Eissa,
Thermochim. Acta 376 (2001) 43.
[13] El.-H.M. Diefallah, M.A. Mousa, A.A. El-Bellihi, E.H. El-
Mossalamy, G.A. El-Sayed, M.A. Gabal, J. Anal. Appl.
Pyrolysis 62 (2002) 205.
[14] G. Grosse, Mos-90, version 2.2, second ed., Oskar-Maria-
Graf-Ring, Munchen, 1992.
[15] T. Ozawa, Bull. Chem. Soc. Jpn 38 (1965) 1881. T. Ozawa,
J. Therm. Anal. 2 (1970) 301.
[16] A.W. Coats, J.P. Redfern, Nature (1964) 20.
[17] El.-H.M. Diefallah, Thermochim. Acta 202 (1992) 1.
[18] J.M. Criado, Thermochim. Acta 24 (1978) 186.
[19] C.D. Doyle, J. Appl. Polym. Sci. 5 (1961) 285.
[20] El.-H.M. Diefallah, A.Y. Obaid, A.H. Qusti, A.A. El-Bellihi,
A.M. Abdel Wahab, Thermochim. Acta 274 (1996) 172.
[21] E.D. Macklen, J. Inorg. Nucl. Chem. 29 (1967) 1229.
Generally, in NiC2O4–FeC2O4 mixture, the activation
energy for the thermal decomposition of NiC2O4 is higher
than that of FeC2O4 (Table 4). The radius of Ni2þ (0.72 A) is
˚
smaller than that of Fe2þ (0.76 A), so that Ni2þ would have
˚
a stronger metal-oxygen bond than Fe2þ
.

M.A. Gabal / Journal of Physics and Chemistry of Solids 64 (2003) 1375–1385
1385
[22] S. Music, S. Popovic, S. Dalipi, J. Mater. Sci. 28 (1993) 1793.
[23] P.R. Brady, J.F. Duncan, J. Chem. Soc. (1964) 653.
[24] O.C. Kistner, A.W. Sunyar, Phys. Rev. Lett. 4 (1960) 412.
[25] A.Narayanasamy,L.Haggstrom,SolidStatePhys. 16(1983)591.
[26] V. Moye, K.S. Kane, V.N. Kamat, Dalal, J. Mater. Sci.: Mater.
Electron. 1 (1990) 212.
[28] Sh. Yariv, E. Mendelovici, Appl. Spetrosc. 33 (1979) 410.
[29] El.-H.M. Diefallah, S.N. Basahel, A.A. El-Bellihi, Thermo-
cim. Acta 290 (1996) 123.
[30] R.D. Waldron, Phys. Rev. 99 (1955) 1727.
[31] T. Palanismy, J. Gopalakrishnan, B. Viswanthan, V. Sriniva-
san, V.C. Sastri, Thermochim. Acta 2 (1971) 265.
[32] J.A. Dean, Lange’s Hand Book of Chemistry, twelfth ed., Mc
Graw-Hill, New York, 1979.
[27] U. Schwertmann, R.M. Taylor, Minerals in Soil Environments,
SoilScienceSocityofAmerica, Inc, Madison, WI, 1977, p. 145.