Chem. Eng. Technol.

19 (1996) 373-377 373

Kinetics of Calcium Carbonate Decomposition

T. Rajeswara Rao*

Experiments on thermal decomposition of calcium carbonate were carried out in a thermo-

gravimetric analyser under non-isothermal conditions of different heating rates (10 to
100 "C/min). A new technique for determining the kinetic parameters from non- isothermal
thermogravimetric data was described. The activation energy and frequency factors were
determined from the proposed method and also by the widely used Coats and Redfern
method. The kinetic parameters and the kinetic compensation effect between the activation
energy and frequency factors obtained from both the methods were found to be very consis-
tent and are in very good agreement with the literature values. The activation energy and
frequency factors were also determined from isothermal experiments in the temperature
range from 680 to 875 "C. The activation energy and frequency factors determined from
isothermal data using initial rate method were also found to be in very good agreement with
the above results. It is also found that the kinetic parameters determined by isothermal
analysis were consistent with the values determined by non-isothermal analysis.

1 Introduction tors and are discussed by Rao et al. [I91 and Rao [20]. The
intrinsic reaction rate constant determined by Rao et al. [I91
Several investigations have been carried out on the thermal
and Rao [20] was based on structural (grain) model while it
decomposition of calcium carbonate. Different techniques
was based on shrinking core model by other investigators.
have been reported in literature for determining the kinetic
The associated heat and mass transfer considerations were
parameters, viz., activation energy and frequency factors
not taken into account by these investigators in determining
from non-isothermal thermogravimetric data. Doyle [I],
the reaction rate constant.
Zsako [2], Horowitz and Metzer [3], Coats and Redfern [4]
have developed integral methods for determining the kinetic In the present work a simple method is proposed for deter-
parameters. Newkirk [5], Ingraham and Marier [6], and mining the rate parameters from non-isothermal thermo-
Sharp and Wentworth [7] have used differential methods. gravimetric data. The kinetic parameters are also deter-
Freeman and Corrol [8] proposed a difference-differential mined from widely used Coats and Redfern [4] method for
method which involves differences in rates for determining comparison. The reaction rate parameters are also deter-
the kinetic parameters. Chatterjee [9] proposed a technique mined from isothermal experimental data using initial rate
which requires two thermogravimetric curves from samples method.
of different weights carried out under identical conditions.
Zsako [ 101 has discussed some problems in determining
2 Experimental
kinetic parameters from thermogravimetric data and con-
cluded that Coats and Redfern [4] method is simple and
A Perkin-Elmer TGA-7 thermogravimetric analyser which
more accurate compared to other integral methods. Criado
gives sample weight changes as a function of time and tem-
and Morales [I I] have discussed some possibilities of misin-
perature to an accuracy of f mg and + 2 K, respec-
terpretation of the mechanism of reaction from thermo-
tively, has been used. Analar grade calcium carbonate (May
gravimetric data and found that contracting volume
and Baker, Ltd., Dagenham, England) has been used in the
(shrinking core) mechanism is appropriate for calcium car-
experiments. The average grain size determined by measur-
bonate decomposition.
ing large number of individual grains using electron micro-
It has been found by several investigators, Gallagher and scope is 1.07-10-5 m. In case of non-isothermal experi-
Johnson [12, 131, Zsako and Arz [14], Caldwell et al. [15], ments the sample weight is approximately 80 mg and nitro-
Criado et al. [16], Khraisa and Dugwell [17], and Bapat et gen gas flow rate is 50 ml/min at normal temperature and
al. [ t 81 that a kinetic compensation effect which is a linear pressure. Experiments are conducted at heating rates of 10,
relationship between activation energy and logarithmic val- 25, 50, 75, and 100°C per minute. In case of isothermal ex-
ue of frequency factor exists for thermal decomposition of periments, the data reported by Rao et al. [I91 at tempera-
solids. tures 680,692,722,737,777,810,840, and 875 "C are taken
and the initial rates are taken from the TGA thermograms.
The thermal decomposition of calcium carbonate under
isothermal conditions has been studied by several investiga-
3 Results and Discussion
* T. R. Rao, Associate Professor, Department of Chemical Engineer-
ing, Indian Institute of Technology Delhi, H a w Khas, New The experimental results obtained from the thermobalance
Delhi-110016, India. are in the form of plots of sample weight percent [wt.-To]

0 VCH Verlagsgesellschaft mbH, D-69451 Weinheim, 1996 0930-7516/96/0408-0373 $10.00+ .25/0

374 Chem. Eng. Technol. 19 (1996) 373-377

and its first derivative [%/min] against temperature corre- It is observed from Fig. 1 that all the four lines intersect at
sponding to a heating rate. It is observed from the thermo- a point corresponding to a temperature of 738 "Cwhich can
gravimetric curves that the temperature of the sample corre- be considered as isokinetic temperature. It is also found that
sponding to a particular value of weight percent of the sam- an empirical linear relationship exists between the activa-
ple or conversion increases with increase in heating rate. tion energy and the conversions, i.e., weight percent of the
Thus, the instantaneous reaction rates at any level of con-
version depend on temperature and, hence, rate constants at
different temperatures corresponding to a fixed conversion
value can be determined.

As the particles of calcium carbonate are very small it is rea-

sonable to assume that the reaction in the particles proceeds
according to shrinking core model, i.e., contracting volume
mechanism. Gallagher and Johnson [ 131 have reported that
contracting volume mechanism, i.e., reaction order 2/3,
gave the best fit for their data on decomposition of calcium
carbonate in carbon dioxide. Criado and Morales [I I], by
analysing both isothermal and non-isothermal thermogravi-
metric data of calcium carbonate decomposition, reported
that contracting volume mechanism is the appropriate
mechanism. Thus, the instantaneous reaction rate accord-
ing to contracting volume mechanism is given by ')

dX/dt = (3 K , / e R , ) (1 - X ) 2 / 3 (1)

The above equation can be written in the following form for

calculating the activation energy, and frequency factors.
0.8 0.9 1.0
In (d W/dt) = In [3 KOA W, (1 - X)2/31- E ( R T ) (2) 1/T x103 [K-'1
Figure 1. Plots of (dW/dt) vs. 1/T for different weight percentages of
According to Eq. (2), a plot of In (dW/dt) vs. l/Tgives the the solid. Weight percent: 0 = 95 wt.-70, V = 90 wt.-70, = 80 wt.-70,
kinetic parameters, activation energy and frequency fac- A = 70 wt.-%.
tors, The experimental data corresponding to four different
weight percents of the sample are plotted as shown in Fig. 1 I
and the activation energy and frequency factor values ob- 6
tained are given in Tab. 1. It is seen from Tab. 1 that as the
activation energy values decrease, the frequency factor val-
ues also decrease indicating the existence of kinetic compen- 5
sation effect. The kinetic compensation law is generally ex-
pressed by the following equation:

log K O = a E + b (3) 4

The activation energy and frequency factor values obtained sr"

are shown by square points according to Eq. (3) in Fig. 2. a3
-
0

Table 1. Kinetic parameters for thermal decomposition of calcium car-

2
bonate (present technique).

Weight of E x ~ O - ~
reacting solid [J/kmol]
[wt.-701
1

95 175.52 5.432
90 157.38 4.525 0
80 130.35 3.235
70 109.81 2.349
E xlOd [J kmol"]
Figure 2. Kinetic compensation effect for the thermal decomposition of
1) List of symbols at the end of the paper. calcium carbonate in nitrogen gas.
Chem. Eng. Technol. 19 (1996) 373-377 375

19.5

19.0

Lu
-c

18.5

I I I
0-8 0.9 1.0 1.1
18.C ~

100 90 80 70 6 1 / T x103 [K-'1

Figure 4. Plots of In [(I -(1 - X ) " 3 J / T Z ]vs. 1 / T for different heating
rates. Heating rate: 1 = lO"C/min, 2 = 25"C/min, 3 = 5O"C/min,
4 = 75 " C h i n , 5 = 100 "C/min.

samples which is shown in Fig. 3. On extrapolation to zero by Zsako and Arz [14] for calcium carbonate decomposi-
conversion, an activation energy value of 1.91 * lo8 J/kmol tion. The kinetic parameters obtained are given in Tab. 2.
is obtained which can be considered as the value corre- The variation of activation energy and frequency factors ac-
sponding to isothermal condition based on initial reaction cording to Eq. (3) are shown in Fig. 2. The inverted trian-
rates. gular points refer to the initial kinetic stage for 10, 15, and
25 "C/min heating rate while the circular points correspond
According to Coats and Redfern [4]method assuming con- to the other stage of the reaction. It is seen that the kinetic
tracting volume mechanism, the following equation gives compensation relation is consistent for both the stages of
the values of activation energy and frequency factors. the reaction as well as with the results obtained from the
proposed method. The kinetic compensation law for calci-
In [(l -(1 - X ) " 3 J / T 2 ]= ln((KoR/j?E) (1 - 2 R T/E)J um carbonate decomposition reported by various investiga-
tors and the present work is given in Tab. 3. It is seen that
-E / ( R T ) (4) the values of a and b obtained from this study are in good
agreement with the values reported by other investigators.
Fig. 4 shows the results obtained. It is observed that lines 1,
2, and 3 of the lower heating rates are composed of two dis-
tinct kinetic stages. A similar trend has also been reported The reaction rate constant from values of initial rates can
be obtained from Eq. (1) as follows.

Table 2. Kinetic parameters for thermal decomposition of calcium car-

bonate (Coats and Redfern method).

Heating rate E x ~ O - ~ 1% KO These results are shown in Fig. 5 in the form of Arrhenius
[k min- '1 [J/kmol]
plot. It is seen that there are two distinct regions, one below
10* 103.85 1.557 737°C and the other above it. The equations for the rate
10 169.55 5.108 constant for the two regions are:
25 * 80.53 0.467
25 159.10 4.635 K, = 5.22 lo4 exp ( - 1.86 10s/[R 71) from 682 to 737 "C
50 * 76.50 0.352 K , = 16.68 exp ( - 1.18 108/[R f l ) from 737 to 875°C
50 140.12 3.653
15 129.68 3.150
100 126.07 3.006 A decrease in reaction rates and activation energies in the
higher temperature range is usually attributed to diffusional
* Initial stage of decomposition effects. In the present case, as the reaction rate constants
376 Chem. Eng. Technol. 19 (1996) 373-377

, -
Table 3. Kinetic compensation law parameters for thermal decomposi-
tion of calcium carbonate: log KO= a E + b.

Investigator a x lo6 b Experimental

lo4-
[kmol/J] conditions

Y . H. Khraisha, 0.0634 ~ 4.32 cauldon limestone in

D.R. Dugwell (1989) NZ
-
0.0487 - 2.18 cauldon limestone in
CO,/O, and CO,/N,
-6

are determined from initial rates before the onset of diffu-

Table 4. Temperatures and temperature drops AT in the pellet.
sional resistances, the reduction in activation energy cannot
be due to diffusional effects. From a careful analysis of the Initial Calculated AT AT* AT**
data it is found that the decrease in rate constants at higher pellet pellet calculated exp. exp.
temperatures is due to cooling of the solid pellet from the temp. [K] temp. [K] IKI [Kl [Kl
initial temperature condition.
1148 1092 56 57 57
1113 1070 43 46 43
It is reported by Hills 1211 and Rao et al. [I91 that the tem- 1050 1034 16 25 17
perature of the solid pellet immediately decreases to a low
value due to endothermic effect and gradually rises to the * Maximum temperature drop in the pellet (Figs. (13 and 14) in [19]).
reaction gas temperature towards the end of the reaction. ** Temperature drop during the first 100 seconds.
Also temperature drop in the solid decreases with decrease
in reaction temperature. In the present case up to 737 "C the
temperature drop in the pellet can be assumed to be very
In order to verify the consistency of the activation energy
small and, consequently, the rate constants obtained can be
and frequency factor values from isothermal and non-iso-
considered corresponding to the reaction gas temperatures.
thermal experiments, the frequency factor values are com-
The activation energy and frequency factor values for this
pared. The frequency factor value, corresponding to 1.86
region are in very good agreement with the literature values.
1O8 J/kmol activation energy, is calculated for non-isother-
Using these values, the temperatures of the solid pellet
ma1 case using the kinetic compensation Eq. (3) and the fol-
which gives the observed rate constants at 777, 840, and
lowing equation:
875 "C are calculated. These values together with measured
temperature drops in the pellets as reported by Rao et al.
[I91 are given in Tab. 4.

It is observed from Tab. 4 that the calculated and measured These are:
temperature drops in the solid pellet are in very good agree-
ment except for 777 "C reaction temperature. At 777 "C the Kb = 5.22 lo4 for isothermal data
minimum temperature is attained over a longer period of Kb = 3.77 lo4 for non-isothermal data
time as compared to the higher reaction temperatures.
Hence, the temperature drops during the initial approxi- The above values are in close agreement with each other.
mately 100 seconds period over which the reaction rates are The temperature 737 "C which can be considered again as
measured are given in Tab. 4 which shows excellent agree- isokinetic temperature is very close to the value 738 "C ob-
ment with the calculated temperature drops. served in case of non-isothermal data analysis.
Chem. Eng. Technol. 19 (1996) 373-377 377

4 Conclusions Greek symbols

It is found that the present method of determining kinetic e [kmol/m3] molar density of the solid
parameters from non-isothermal thermogravimetric data is D [K min-'1 heating rate
quite satisfactory. It is also shown that the initial reaction
rates method for isothermal experimental data is quite satis-
factory for determining the reaction rate constants. From References
the present analysis a temperature of 737 "C can be regard-
ed as isokinetic temperature of calcium carbonate decompo- [l] Doyle, C.D., J . Appl. Polym. Sci. 5 (1961) pp. 285-292.
sition reaction. The kinetic parameters obtained from both [2] Zsako, J., J. Phys. Chem. 72 (1968) pp. 2406-2411.
the methods are found to be very consistent and the rate [3] Horowitz, H.H., Metzger, G., Anal. Chem. 35 (1963)
constants and the kinetic compensation law obtained are in pp. 1464- 1468.
[4] Coats, A.W., Redfern, J.P., Nature201 (1964) pp. 68-69.
very good agreement with literature values. [5] Newkirk, A.E., Anal. Chem. 32 (1960) pp. 1558- 1563.
[6] Ingraham, T.R., Marier, P., Can. J. Chem. Eng. 42 (1964)
pp. 161 - 163.
Received: March 20, 1995 [CET 7421 [7] Sharp, J.H., Wentworth, S.A., Anal. Chem. 41 (1969)
pp. 2060 - 2062.
[8] Freeman, E.S., Carroll, B., J. Phys. Chem. 62 (1958)
pp. 394 - 397.
[9] Chatterjee, P.K., J. Polym. Sci. 3 (1965) pp. 4253 -4262.
[lo] Zsako, J., Therm. Anal. 5 (1973) pp. 239-251.
[ l l ] Criado, J.M., Morales, J., Thermochim. Acta 19 (1977)
pp. 305-317.
Symbols used [12] Gallagher, P.K., Johnson, D.W., Thermochim. Acta 6 (1973)
pp. 67-83.
[kmol/J] constant defined by Eq. (3) [13] Gallagher, P.K., Johnson, D.W., Thermochim. Acta 14 (1976)
[-I constant defined by Eq. (3) pp. 255 - 261.
[J/kmol] activation energy [14] Zsako, J., Arz, H.E., J . Therm. Anal. 6 (1974) pp. 651 -656.
lS-7 frequency factor [15] Caldwell, K.M., Gallagher, P. K., Johnson, D. W., Thermochim.
[ k m ~ l m - ~ s - ~frequency
] factor Acta 8 (1977) pp. 15- 19.
[kmol m2 s- '1 reaction rate constant [16] Criado, J.M., Rouquerol, F., Rouquerol, J., Thermochim. Acta
[J/(kmol K)] gas constant 38 (1980) pp. 1117-1213.
[ml radius of the particle [17] Khraisa, Y. H., Dugwell, D.R., Chem. Eng. Res. Des. 67 (1989)
[sl time pp. 48-51.
[KI temperature [18] Bapat, J.D., Visvesvaraya, H.C., Rao, T.R., Rao, D.S.,
[Kl temperature drop in the solid pellet Zem.-Kalk-Gips 3 (1994) pp. 156- 160.
PJOI weight percent of the solid [19] Rao, T.R., Gunn, D. J., Bowen, J. H., Chem. Eng. Res. Des. 67
[qol percentage weight loss at completion of the (1989) pp. 38-47.
reaction [20] Rao, T.R., Can. J. Chem. Eng. 71 (1993) pp. 481 -484.
[-I fractional conversion [21] Hills, A. W.D., Chem. Eng. Sci. 23 (1968) pp. 297-320.