A study on isothermal kinetics of thermal decomposition of cobalt oxalate to cobalt

Article abstract of DOI:10.1016/j.tca.2008.04.005

Thermal decomposition of cobalt oxalate (CoC₂O₄·2H₂O) was studied using TG, DTA and XRD techniques. Non-isothermal studies revealed that the decomposition occurred in two main stages: removal of water and next decomposition of CoC₂O₄ to Co. Isothermal kinetic studies were conducted for the second stage of decomposition at six different temperatures between 295 and 370 °C in reducing (He + 15 vol% H₂) atmosphere. Kinetic equation was found to obey g(α) = [-ln(1 - α)]^1/n = kt relationship with varying values of n between 2 and 4. The variation of activation energy between 73.23 and 119 kJ mol^-1 from the lowest to the highest temperature of decomposition indicated the multi-step nature of the process. SEM analysis revealed the formation of nano-size powder in this temperature range. Thermal decomposition at 550 °C for 30 min was found to be optimum condition producing fine size, non-pyrophoric and spherical powder in larger scale of production. Both hcp and fcc phases were present in the cobalt powder at room temperature.

Full text of DOI:10.1016/j.tca.2008.04.005

Thermochimica Acta 473 (2008) 45–49
Contents lists available at ScienceDirect
Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca
A study on isothermal kinetics of thermal decomposition of
cobalt oxalate to cobalt
S. Majumdar∗, I.G. Sharma, A.C. Bidaye, A.K. Suri
Materials Group, Bhabha Atomic Research Centre, Trombay, Mumbai 400085, India
a r t i c l e i n f o
a b s t r a c t
Article history:
Thermal decomposition of cobalt oxalate (CoC2O4·2H2O) was studied using TG, DTA and XRD techniques.
Non-isothermal studies revealed that the decomposition occurred in two main stages: removal of water
and next decomposition of CoC2O4 to Co. Isothermal kinetic studies were conducted for the second stage
of decomposition at six different temperatures between 295 and 370 ◦C in reducing (He + 15 vol% H2)
atmosphere. Kinetic equation was found to obey g(˛) = [−ln(1 − ˛)]1/n = kt relationship with varying val-
ues of n between 2 and 4. The variation of activation energy between 73.23 and 119 kJ mol−1 from the
lowest to the highest temperature of decomposition indicated the multi-step nature of the process. SEM
analysis revealed the formation of nano-size powder in this temperature range. Thermal decomposition at
550 ◦C for 30 min was found to be optimum condition producing fine size, non-pyrophoric and spherical
powder in larger scale of production. Both hcp and fcc phases were present in the cobalt powder at room
temperature.
Received 13 December 2007
Received in revised form 31 March 2008
Accepted 7 April 2008
Available online 16 April 2008
Keywords:
Kinetics
Isothermal
Thermal decomposition
Cobalt oxalate
© 2008 Elsevier B.V. All rights reserved.
1. Introduction
powder at room temperature has also been reported earlier [5]. The
non-isothermal studies on the mechanism and kinetics of thermal
Cobalt is well established for its applications in the field of
cemented carbides, tool steels, magnetic materials, paint pigments,
catalysts and production of artificial ␥-ray sources. Pure cobalt in
the form of slugs (6 mm × 25 mm) and pellets (1 mm × 1 mm) are
used in the nuclear reactors to prepare Co60 ␥-radiation source. The
key uses of Co60 slugs and pellets are food preservation and radia-
tion therapy in medical applications, respectively. These shapes are
fabricated through the powder processing route involving appro-
priate combination of compaction, sintering [1] and hot working
techniques, such as hot extrusion, hot swaging under reducing
atmosphere. The properties of sintered metallic products depend
on the size, morphology and purity of the powders. Various routes
such as: (1) aqueous cobaltous hydroxide slurry under hydro-
gen reduction conditions using palladium chloride catalyst [2], (2)
reduction of cobalt oxide, chloride or sulfate [3], (3) decomposition
of ammonium cobalt fluoride [4] and (4) electro-winning [3] have
been explored for the preparation of cobalt powder.
decomposition of cobalt oxalate have been reported earlier [6,7].
However, isothermal kinetics of the decomposition in reducing
atmosphere and its effect on powder morphology is not available in
the literature. Some isoconversional methods for analyzing thermal
decomposition have been described earlier [8,9].
In the present investigation, isothermal kinetics of the decom-
position of cobalt oxalate to metallic cobalt powder was studied in
a thermogravimetric (TG) and differential thermal analysis (DTA)
equipment Setsys Evolution 24, Setaram Instrumentation, France.
Based on the non-isothermal studies carried out in both helium
and helium +15 vol% H2 atmospheres, six different temperatures
between 295 and 370 ◦C were selected. The isothermal studies were
conducted in the atmosphere containing 15 vol% hydrogen and bal-
ance helium. Powder morphology and phase analysis were carried
out by SEM and XRD technique, respectively. Co-relation with the
larger scale decomposition experiments was established and opti-
mum condition for preparing cobalt powder suitable for making
slugs and pellets was predicted.
Thermal decomposition of cobalt oxalate is the most commonly
used technique for the preparation of high purity, fine size cobalt
metal powder. The co-existence of both hexagonal closed packed
(hcp) and face centered cubic (fcc) phases in as produced cobalt
2. Experimental procedure
2.1. Preparation of cobalt oxalate
∗
99.9% pure cobalt sulfate (CoSO4·7H2O) supplied by M/s s. d.
Corresponding author. Tel.: +91 22 25590183; fax: +91 22 25505151.
E-mail address: sanjib@barc.gov.in (S. Majumdar).
fine chemicals was used as starting feed. Cobalt sulfate solutions
0040-6031/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.tca.2008.04.005

46
S. Majumdar et al. / Thermochimica Acta 473 (2008) 45–49
were prepared by maintaining the cobalt concentration at 30 and
60 g/L. Cobalt oxalate precipitation reactions were conducted by
mixing oxalic acid solution with cobalt sulfate solution under vig-
orous stirring. Studies were conducted by varying the experimental
parameters such as temperature, pH and concentration of oxalic
acid. The optimum condition producing a maximum cobalt recov-
ery of 99% was found to be T 40 ◦C, pH 1.1 and 40% excess oxalic
acid solution. Cobalt oxalate precipitate was filtered, neutralized
and dried under infrared lamp. Finally, cobalt oxalate dihydrate
(CoC2O4·2H2O) powder having a pink color was obtained after dry-
ing.
2.2. TG–DTA study
Thermogravimetric studies were conducted using the TG–DTA
equipment; model SETSYS EVOLUTION 24 from Setaram Instru-
mentation, France. The equipment consists of cylindrical graphite
furnace, graphite reaction tube, TG–DTA sensor (microbalance and
DTA transducers), carrier and auxiliary gas inlet, furnace thermo-
couple for temperature control by PID controller, etc. Tungsten
DTA transducer (rod) was used for carrying out the experiments
in reducing (He + 15% H2) as well as inert (pure helium) atmo-
spheres. Criado and Perez-Maqueda [10] have reported that the
use of sample controlled thermal analysis (SCTA) methods present
two important advantages with regard to the more conventional
rising temperature experiments. Because they have a higher reso-
lution power for discriminating among the reaction kinetic models
and also is a powered tool for minimizing the influence of the
experimental conditions on the forward reaction. In the current
study, around 30 mg cobalt oxalate was charged in each experi-
ment in a tungsten (sample) crucible and placed in the housings
made in the TG–DTA rod along with the blank reference crucible.
The depth of the cobalt oxalate layer inside the crucible was 6 mm.
The chamber was evacuated up to 10−2 mbar after charging the
sample followed by filling with required carrier gas. Initial exper-
iments were conducted under the non-isothermal heating rate of
5 ◦C min−1 up to 550 ◦C in pure helium (He) and helium +15 vol%
hydrogen atmospheres, respectively. Cobalt oxalate produced from
both 30 and 60 g/L solutions were studied. The temperature range
for the final stage of the decomposition i.e. CoC2O4 to Co formation
was detected by these initial studies. Next set of experiments was
conducted for studying the kinetics of the final stage of decom-
position of cobalt oxalate prepared from 60 g/L initial solution
under He + 15% H2 atmosphere. Isothermal decomposition stud-
ies were conducted at six different temperatures between 295 and
370 ◦C. Initially the samples were heated at the rate of 20◦ min−1
to reach the desired temperature of decomposition followed by
holding for 4 h at all the different attempted temperatures. The
kinetic data obtained from thermogravimetric study was analyzed
and activation energy was calculated by establishing the kinetic
law.
Fig. 1. Decomposition of CoC2O4·2H2O in helium. Heating rate 5 ◦C min−1
.
tion peaks were analyzed and indexed according to the diffracting
planes of different phases.
3. Results and discussions
3.1. Non-isothermal studies
TG and DTA data obtained during the decomposition of cobalt
oxalate in helium and helium with 15 vol% H2 are presented in
Figs. 1 and 2, respectively. Thermal decomposition of CoC2O4·2H2O
is described by the following reaction:
CoC2O4·2H2O = Co + 2CO2 + 2H2O
(1)
As observed in Fig. 1, the decomposition in helium takes place
mainly in two stages. In the first stage that takes place below 200 ◦C
water of crystallization gets removed as indicated by a drop in
weight and endothermic DTA peak at 182 ◦C. The second stage is the
decomposition of CoC2O4 to metallic cobalt and formation of CO2
gas. The endothermic DTA peak is at 365 ◦C and the decomposition
is complete at 395 ◦C.
A different nature of the DTA curve (Fig. 2) is observed for
decomposition in the atmosphere containing 15 vol% H2 and bal-
ance He. The first stage, removal of water of crystallization showed
the similar behavior as obtained in case of He atmosphere, whereas
in the second stage there are two additional exothermic peaks at
350 and 372 ◦C, respectively, coupled with one endothermic peak
at 365 ◦C. The area of the endothermic peak (365 ◦C) has been
reduced because of some other exothermic reactions occurring
2.3. Scanning electron microscopy
The particle size and morphology of the cobalt powder was stud-
ied in a Hitachi make SEM. Powder was dispersed on a self-adhesive
carbon tape and placed inside the chamber of the SEM. The chamber
was evacuated to 10−5 mbar vacuum and samples were analyzed at
15 kV accelerating voltage and 12.5 mm working distance (WD).
2.4. X-ray diffraction
Panalytical make XRD equipment was used for X-ray diffrac-
˚
tion study. Mo K␣ radiation of wavelength 0.7093 A was used and
sample scanning was done between the angles of 10–50◦. Diffrac-
Fig. 2. Decomposition of CoC2O4·2H2O in He + 15 vol% H2. Heating rate 5 ◦C min−1
.

S. Majumdar et al. / Thermochimica Acta 473 (2008) 45–49
47
Fig. 4. Kinetic plots for isothermal decomposition.
Fig. 3. TG plot for decomposition of oxalate prepared from 30 to 60 g/L sulfate
solutions. Heating rate 5 ◦C min−1
.
Table 1
Kinetic models used for plotting t/t0.5 vs. ˛ in Fig. 5
simultaneously. The gas phase reactions between CO2 and H2
forming hydrocarbons are responsible these exothermic peaks. For-
mation of methane as per the reaction
Model
Integral form, g(˛)
A2—Avrami–Erofe’ev (J–M)
A3—Avrami–Erofe’ev
A4—Avrami–Erofe’ev
R2—Contracting area
R3—Contracting volume
D1—1D diffusion
[−ln(1 − ˛)]1/2
[−ln(1 − ˛)]1/3
[−ln(1 − ˛)]1/4
1 − (1 − ˛)1/2
1 − (1 − ˛)1/3
CO2 + 4H2 = CH4 + 2H2O
(2)
is thermodynamically feasible (ꢀG◦r = −13.66 kcal mol−1) [11] and
2
␣
has been confirmed by Maciejewski et al. [7].
D2—2D diffusion
(1 − ␣) ln(1 − ˛)+␣
Fig. 3 is the TG plot showing the comparative weight change
for the thermal decomposition reactions of cobalt oxalates from 30
to 60 g/L sulfate solutions, respectively. No significant difference
in decomposition kinetics is detected though oxalate from 60 g/L
solution is expected to have larger particle size. This result has got
the significance particularly on the aspects of cobalt metal recovery
from the starting feed. It is because the decomposition kinetics is
independent of the concentration of the feed solution; the recov-
ery of the cobalt values can be maximized using the feed solution
containing highest possible concentration of cobalt. Cobalt oxalate
from 60 g/L starting solution was, therefore, used for subsequent
studies on isothermal kinetics of second stage of decomposition.
scale (reduced time)
ꢀ
ꢁ
t
g(˛) = A
(4)
t0.5
where A is a calculated constant dependent on the form of the
function g(˛). This expression is independent of the kinetic rate
constants and is dimensionless. Calculations were performed for
evaluating t/t0.5 at different values of ˛ for different tempera-
tures. The values were compared with standard values for different
kinetic models [13].
The reduced time plots for the decomposition kinetics at 295
and 310 ◦C are presented in Fig. 5 along with the theoretical plots
for different classical models (Table 1). At lower decomposition
temperatures (295 and 310 ◦C), the experimental curves are closely
matching with the equation [−ln(1 − ˛)]1/2 = kt (A2). The model for
3.2. Isothermal kinetics
Isothermal decomposition studies were conducted in
He + 15 vol% H2 atmospheres to avoid traces of oxygen pick
up by the freshly produced cobalt powder. Based on the non-
isothermal studies, six different temperatures 295, 310, 325, 340,
355 and 370 ◦C were selected for conducting isothermal kinetic
studies. The samples were heated at 20 ◦C min−1 to the desired
temperature followed by isothermal holding for 4 h. Fig. 4 shows
the variation of degree of decomposition (˛) of CoC2O4 to Co with
time at different isothermal conditions. Up to the decomposition
temperature of 325 ◦C incubation period is observed and decom-
position at 295 ◦C is not complete within 4 h of holing time. Beyond
340 ◦C, the decomposition is just initiated and at 370 ◦C the value
of ˛ is 0.05 at t = 0. Sigmoidal shape of ˛–t plot indicates that the
kinetics of the decomposition reaches a maximum followed by a
decrease in the rate. Reduced time plot approach [12] was used for
evaluating the kinetic equation for the decomposition. The kinetic
relationship is expressed in the form:
g(˛) = kt
(3)
where ˛ is the fraction reacted, k the rate constant and t the time. If
t0.5 be the time required to obtain 0.5 fraction reacted (˛ = 0.5) then
one obtains Eq. (3) in an altered form in terms of dimensionless time
Fig. 5. Reduced time plots for different kinetic models and experimental conditions.

48
S. Majumdar et al. / Thermochimica Acta 473 (2008) 45–49
Table 2
Residual sum of squares (s2) for the kinetic data obtained at different isothermal
experimental conditions for different kinetic models
Temp (◦C)
A2
A3
A4
R2
R3
295
310
325
340
355
370
0.00113a
0.002237a
0.018291
0.105476
0.050805
0.01015a
0.339388
0.016221
0.003946a
0.027427
0.005104
0.011171
0.303281
0.043416
0.017089
0.013298a
0.005037a
0.032828
0.813425
0.111119
0.175287
0.368703
0.26031
1.018416
0.190598
0.26832
0.509876
0.37762
0.142792
0.231028
a
Minimum s2 values.
this equation known as J–M model named after Johnson and Mehl
[14], is well known and commonly used to describe the kinetics of
recrystallization [15].
For obtaining the kinetic equation at higher temperatures, a sta-
tistical test has been performed for different kinetic models and
residual sum of squares (s2) given by
Fig. 7. Dependence of activation energy on the extent of conversion (˛). E˛ values
obtained from ln t˛ vs. 1/T plots.
ꢀ
ꢁ
2
ꢂ
1
t
g(˛)
g(˛0.5)
s2
=
−
(5)
n − 1
t0.5
for each model has been calculated for different temperatures and
presented in Table 2. It is observed that at lower temperatures (295
and 310 ◦C), s2 is minimum for A2 whereas the minimum residual
sum of squares at 325, 340, 355 and 370 ◦C was observed for the
models A3, A4, A4 and A2, respectively, as indicated by asterisk (*)
marks in Table 2.
Therefore, thermal decomposition of cobalt oxalate follows
Avrami–Erofe’ev model
g(˛) = [− ln(1 − ˛)]1/n = kt
(6)
where the exponent ‘n’ varies between 2 and 4 with increase in the
decomposition temperature.
The value of activation energy was calculated by plotting ln t˛
against 1/T as because the decomposition was not following any
particular kinetic law throughout the temperature range. Fig. 6
represents a typical ln t˛ vs. 1/T plot for ˛ = 0.4. From the slope
of linear curve, the activation energy at ˛ = 0.4 was found to be
93.8 8.31 kJ mol−1. Similarly, the values of activation energy at
different ˛ were calculated by analyzing ln t˛ vs. 1/T plots for differ-
ent ˛. Fig. 7 shows the variation of activation energy (E˛) with the
extent of conversion (˛). The increase in E˛ with ˛ indicates that
the thermal decomposition of cobalt oxalate to cobalt is a multi-
step process [16,17]. The activation energy varied between 73.23
and 119 kJ mol−1 from the lowest to the highest temperature of
decomposition conditions.
Fig. 8. SEM image showing nano-size cobalt powder formation at 340 ◦C.
3.3. Morphology and phase of cobalt
Detailed SEM analysis of the cobalt powder produced at different
isothermal conditions was carried out. The powder produced below
340 ◦C immediately caught fire while removing from the ther-
mobalance. Since during the decomposition of the cobalt oxalate
particle six foreign atoms leave its lattice for each cobalt atom, the
product becomes porous thus having a very high specific surface
Fig. 6. Plot of ln t˛ vs. 1/T for ˛ = 0.4.
Fig. 9. SEM image showing nano-powder chain formation at 370 ◦C.

S. Majumdar et al. / Thermochimica Acta 473 (2008) 45–49
49
varied starting from 400 ◦C. The optimum condition producing
non-pyrophoric, spherical powder of minimum agglomeration was
found to be as temperature 550 ◦C, time 30 min under reducing
atmosphere. Fig. 10 shows the formation of 1–1.5 ␮m spherical
powder at the optimized conditions. The powder was easily flow-
able and thus suitable for preparation of cobalt slugs and pellets
by powder metallurgical processes. Fig. 11 is the XRD plot of the
cobalt powder produced at 550 ◦C showing the presence of both
hcp and fcc phases in the powder. Strong (1 1 1) reflections from fcc
cobalt was detected along with the (1 0 0), (1 0 1), etc., hcp reflec-
tions. Though the allotropic transition from hcp to fcc phase is at
417 ◦C, significant amount of high temperature fcc phase is retained
in the cobalt crystals. The peak broadening at each reflection is also
the indicative of formation of fine size cobalt powder.
4. Conclusions
Isothermal kinetics of thermal decomposition of cobalt oxalate
to cobalt was found to obey Avrami-Erofe’ev equations. The
increase in activation energy with increasing temperature revealed
that the decomposition involved multi-step reaction mechanisms.
The powder size was less than 100 nm at the decomposition
temperatures below 370 ◦C. Decomposition at 550 ◦C for 30 min
produced the cobalt powder suitable for making slugs and pellets
by powder metallurgical processing techniques. Existence of both
hcp and fcc phases was observed in the cobalt powder at room
temperature.
Fig. 10. Micron size cobalt powder produced after decomposition at 550 ◦C for
30 min.
Acknowledgements
The authors wish to thank Prof. I. Samajdar and Dr. S. L. Kamat
of Indian Institute of Technology Bombay, Powai, Mumbai, India
for extending his help in carrying out XRD and SEM experiments,
respectively.
References
[1] S. Majumdar, I.G. Sharma, R. Kapoor, J.K. Chakraborty, A.K. Suri, Metall. Mater.
Trans. B 37 (2006) 633–639.
[2] D.J. Kim, H.S. Chung, K. Yu, Mater. Res. Bull. 37 (2002) 2067–2075.
[3] K.K. Sharma, T.V.L. Narshima Rao, Non-Ferrous Metals Strategy Cum Source
Book, 4, 2003, pp. 175–187.
[4] I.G. Sharma, P. Alex, S. Majumdar, J. Kishor, A.K. Suri, J. Alloy. Compd. 437 (2007)
231–237.
Fig. 11. XRD plot showing co-existence of both hcp and fcc phases in cobalt powder.
[5] J.Y. Huang, Y.K. Wu, H.Q. Ye, Appl. Phys. Lett. 66 (1995) 308–310.
[6] B. Malecka, E. Drozdz-Ciesla, A. Malecki, J. Therm. Anal. Cal. 60 (2002) 819–
831.
[7] M. Maciejewski, E. Ingier-Stocka, W.D. Emmerich, A. Baiker, J. Therm. Anal. Cal.
60 (2000) 735–758.
[8] S. Vyazovkin, W. Linert, J. Solid State Chem. 114 (1995) 392–398.
[9] P.J. Skrdla, R.T. Robertson, Thermochim. Acta 453 (2007) 14–20.
[10] J.M. Criado, L.A. Perez-Maqueda, J. Therm. Anal. Cal. 80 (2005) 27–33.
[11] I. Barin, O. Knacke, Thermochemical Properties of Inorganic Substances,
Springer-Verlag Berlin Heidelberg, New York, 1973.
area. This makes the powder highly pyrophoric in nature. Fig. 8
shows the SEM image of the cobalt powder produced at 340 ◦C. The
individual particle size was less than 100 nm which is the single
domain region of metallic cobalt. It becomes very difficult to sepa-
rate or resolve individual cobalt particles of this nano-size range.
As the temperature of decomposition is increased to 370 ◦C,
coarsening of the fine particles takes place. As evident in Fig. 9,
cobalt crystallite size is bigger than that produced at 340 ◦C. The
highly ferromagnetic character of the powder makes them aligned
at particular directions. As a result the acicular morphology of the
aggregate is observed in all the powders. From the point of view of
larger scale production, the powder is still difficult to handle.
Larger scale (60 g) experiments were conducted in a hori-
zontal tubular furnace keeping the cobalt oxalate powder in a
molybdenum tray. Temperature and time of decomposition was
[12] H.S. Ray, Kinetics of Metallurgical Reactions, Oxford & IBH Publishing Co. Pvt.
Ltd., 1990, pp. 39–45.
[13] E.M. Kurian, J. Therm. Anal. 35 (1989) 1111–1117.
[14] W.A. Johnson, R.F. Mehl., Trans. AIME 135 (1939) 416–441.
[15] F.J. Humphreys, M. Hatherly, Recrystallization and Related Annealing Phenom-
ena, Pergamon Press, Oxford, 1995, pp. 188–189.
[16] S. Vyazovkin, N. Sbirrazzuoli, J. Phys. Chem. B 107 (2003) 882–888.
[17] M.E. Brown, M. Maciejewski, S. Vyazovkin, R. Nomen, J. Sempere, A. Burnham,
J. Opfermenn, R. Strey, H.L. Anderson, A. Kemmler, R. Keuleers, J. Janssens, H.O.
Desseyn, C.-R. Li, T.B. Tang, B. Roduit, J. Malek, T. Mitsuhashi, Thermochim. Acta
355 (2000) 125–143.