The effect of particle size distribution on barite reduction

Article abstract of DOI:10.1016/j.materresbull.2009.02.018

The present study is a follow-up of investigation on barite reduction to barium sulfide as a function of starting particle size distribution and temperature of reaction. In this work we study the high temperature reduction process of barite from the view

Full text of DOI:10.1016/j.materresbull.2009.02.018

Materials Research Bulletin 44 (2009) 1489–1493
Contents lists available at ScienceDirect
Materials Research Bulletin
journal homepage: www.elsevier.com/locate/matresbu
The effect of particle size distribution on barite reduction
A. Salem *, Y. Tavakkoli Osgouei
Mineral Processing Research Center, Chemical Engineering Department, Sahand University of Technology, Tabriz, Iran
A R T I C L E I N F O
A B S T R A C T
Article history:
The present study is a follow-up of investigation on barite reduction to barium sulfide as a function of
starting particle size distribution and temperature of reaction. In this work we study the high
temperature reduction process of barite from the view point of particle size distribution. Conversion–
time data have been obtained using iodometry method in each isothermal condition. A modified kinetic
model used to express the carbothermic reduction process. To obtain the values of activation energy and
frequency factor, the same expression was selected for each sample at all temperatures. The rate of
reaction is found to be related to the particle size distribution and the gasification reaction of coke which
has influence on reduction process. The kinetic parameters calculated from standard analysis of
isothermal kinetic data indicated that the particle size of barite controlled the reaction when it was
coarser than 400 mesh both in presence and absence of catalyst.
Received 8 February 2009
Accepted 20 February 2009
Available online 6 March 2009
Keywords:
A. Inorganic compounds
B. Chemical synthesis
D. Catalytic properties
ß 2009 Elsevier Ltd. All rights reserved.
1
. Introduction
Thus, the reduction rate of barite is controlled by the rate of CO
generated by Boudouard reaction. Recently, some investigators
Barium component such as BaCO
3
, BaCl
2
and Ba(NO
3
)
2
are
have been studying the reduction of barium sulfate by carbon in
the presence of inorganic materials like sodium carbonate, ferric
nitrate and sodium vanadate in which these additives act as
promoters for the gasification reaction [3–6].
important chemical materials that are manufactured from natural
barite. These components of barium widely use in ceramic and
glass industries, drilling mud composition and nanostructure
materials. Conversion of barite to barium sulfide by coke which is
On the other hand, the particle size distribution of barite
influences the activation energy of reactions. In fact two factors are
important in reduction process, the particle size of barite and the
gasification of coke. In this study firstly a modified kinetic model
has been developed for reduction of barite. Various factors
affecting on reduction process have been investigated experimen-
tally. The sodium carbonate and ferric nitrate have been selected as
catalysts to study the reaction mechanism. In second step, the
influence of barite particle size distribution was taken into
consideration to investigate its role on reaction kinetic.
‘
‘black ash process’’ is very important reaction in manufacturing
above barium components [1]. This reaction frequently performs
in rotary kiln or in fluidized bed reactors at high temperature
ranging 950–1100 8C in the presence of carbon or methane gas.
Reduction of barite by coke can be represented by following
reaction [2]:
BaSO
4
þ 4C ! BaS þ 4CO
(1)
The major reducing agent is gaseous intermediate CO which
influences the reduction of barite as:
2. Materials and methods
BaSO
4
þ 4CO ! BaS þ 4CO
2
(2)
The influence of particle size distribution of barite on reduction
process has been investigated. At first, an industrial barite ore
was selected and the complete chemical and mineralogical
analyses were identified by X-ray method. As shown in Table 1,
the main mineral was barite and small quantities of quartz were
found. Also, an industrial coke was used in reduction process
whose chemical analysis was presented in Table 2.
In fact the generated CO diffuses and reacts with barite which is
not in contact with carbon. The CO obtained through the reaction
2
diffuses into carbon to generate more CO according to Boudouard
reaction:
C þ CO
2
! 2CO
(3)
In order to evaluate the effect of the barite particle size on
carbothermic process, coarse-sized (ꢀ140, +230 mesh) middle-
sized (ꢀ230, +400 mesh) and fine-sized (ꢀ400 mesh) were used in
*
Corresponding author. Tel.: +98 4123443800/9144108973 (mobile);
fax: +98 4123444355.
4
stoichiometric mixes of barite and coke (i.e., BaSO + 4C). Also, 5%
E-mail address: salem@sut.ac.ir (A. Salem).
sodium carbonate and ferric nitrate were doped on coke as
0025-5408/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.materresbull.2009.02.018

1
490
A. Salem, Y.T. Osgouei / Materials Research Bulletin 44 (2009) 1489–1493
Table 1
Chemical and mineralogical analysis of barite ore.
Chemical analysis
SiO
2
BaO
Na
2
O
K
2
O
MgO
0.08
CaO
0.08
MnO
0.03
Al
2
O
3
Fe
2
O
3
SO
3
2
.75
60.27
0.67
0.01
0.23
0.33
35.53
Mineralogical analysis (calculated by chemical analysis)
Barite
1.8
Quartz
2.75
Others
5.48
9
catalyst, based on dry weight of coke. The carbothermic reduction
of samples have been performed in laboratory electrical kiln under
isothermal conditions at 950, 1000, 1050 and 1100 8C in order to
obtain the values of kinetic parameters. The zero point of time scale
was defined as starting time at which samples enter the kiln. The
cooled samples were weighed and barium sulfide percentage was
measured by iodometry method [7]. The producer for determining
BaS content is as follow:
Fig. 1. Conversion–time plots for the reduction of barite at different particle size
distribution in absence of catalyst.
(
i) the iodine solution and hydrochloric acid solution was
prepared by mixing 50 ml of 0.5N iodine solution and 50 ml
of 1N hydrochloric acid solution.
with the rise in reduction temperature in all conditions. These are
due to gasification promotion of Boudouard reaction at higher
temperature. The conversion–time plots show that the rate of
reaction increases as particle size of barite was finer. As previously
discussed, the values of kinetic parameters obtained by the
logarithmic method allow the assessment of the modified kinetic
model and choice of kinetic expression for the reduction of
barite in all cases. To obtain the values of activation energy, it
is necessary to find the values of rate constants at different
temperatures. The choice of the kinetic equation allows the
calculation of the rate constant values, k, from the logarithmic
form of general kinetic expression as follows:
(
ii) 2 g of black ash was mixed with 100 ml deionized hot water
and stirred to prevent agglomeration and effective dissolution.
(
iii) the black ash suspension was added into the iodine and
hydrochloric acid solution and stirred for 3 min.
(
iv) the solution was titrated with 0.1N sodium thiosulfate to
obtain straw color in presence of starch indicator. The BaS
percentage was calculated by following equation:
ꢀ
ꢁ
N
1
V
1
ꢀ N
2 2
V
C ¼ 8:17
(4)
W
where C is the BaS percentage, N
iodine and sodium thiosulfate respectively. W also is the initial
1
V
1
and N
2
V
2
are total moles of
ln½ꢀlnð1 ꢀ xÞꢁ ¼ ln k þ m ln t
(6)
weight of sample. The fractional conversion of BaSO
calculated according Eq. (1).
4
was
Fig. 2 shows the logarithmic plots of Eq. (5) and the calculated
values of k and m were reported in Table 3 for each isothermal
runs. The values of rate constant are then used in the plot of ln k
versus 1/T which is logarithmic form of Arrhenius equation:
3
. Results and discussion
The experiments, which lead to the exponential construction
ꢀ
ꢁ
Ea
k ¼ k
0
exp ꢀ
(7)
plots of conversion versus time, enable us to elaborate on the
mechanism of reduction process and to estimate the values of
activation energy. To compare the effect of particle size
distribution on reduction of barite a modified kinetic model,
which is related to carbothermic reduction, was applied for all
conversion–time data [3]. The kinetic model can be represented
as following:
RT
m
x ¼ 1 ꢀ expðꢀkt Þ
(5)
where x is the fractional conversion, t is reduction time, k and m are
constant rate and power of equation, respectively.
The conversion–time variations were plotted in Fig. 1 for three
groups of barite particle size distribution in absence of catalyst. As
can be observed in this figure, the conversion values increase
Table 2
Specifications of used coke.
Materials
Percentage
Particle size distribution (mesh)
Percentage
Fixed carbon
Ash
70
10
16
4
+16
0
20
80
ꢀ16, +140
ꢀ140
Volatile matter
Moisture
Fig. 2. Application of modified kinetic model to the isothermal reduction of barite at
different particle size distribution in absence of catalyst.

A. Salem, Y.T. Osgouei / Materials Research Bulletin 44 (2009) 1489–1493
1491
1100
Table 3
The constant rate and m values of modified kinetic model.
Catalyst type
Particle size distribution (mesh)
m
k
r
950
1000
1050
1100
950
1000
1050
1100
950
1000
1050
Without catalyst
ꢀ140, +230
0.50
0.39
0.36
0.35
0.30
0.28
0.24
0.20
0.20
0.31
0.44
0.44
0.66
0.84
0.83
1.16
1.25
1.21
0.99
0.98
0.96
0.98
0.97
0.96
0.97
0.99
0.98
ꢀ
230, +400
400
ꢀ
5% Na
2
CO
3
ꢀ140, +230
0.52
0.50
0.48
0.40
0.46
0.36
0.26
0.31
0.30
0.29
0.37
0.36
0.66
0.57
0.69
1.19
1.17
1.17
0.99
0.98
0.99
0.99
0.99
0.99
0.99
0.98
0.97
ꢀ
230, +400
400
ꢀ
5% Fe(NO
3
)
3
ꢀ140, +230
0.53
0.46
0.34
0.42
0.28
0.30
0.25
0.23
0.16
0.31
0.41
0.47
0.63
0.96
0.83
1.16
1.20
1.36
0.99
0.98
0.99
0.99
0.99
0.99
0.98
0.98
0.99
ꢀ
230, +400
400
ꢀ
Fig. 4. Conversion–time plots for the reduction of barite at different particle size
distribution in presence 5% Na CO as catalyst.
2
3
Fig. 3. Arrhenius plots for reduction of barite in absence of catalyst for different
particle size distribution.
where k
0
is the frequency factor, E
a
is the activation energy, R is the
the gasification of coke. As a matter of fact, in the case of finer size
of barite particles, the activation energy of reduction reaction is
approximately 9.0 kcal less than coarse size and the trends of
activation energy and frequency factor show an overall positive
effect in the rate constants. This phenomenon shows that reaction
was controlled by the particle size of barite when it is coarser than
400 mesh.
In the first part of the investigation it was inferred that the
carbothermic reaction of barite is physically related to gasifica-
tion process of coke throughout the Boudouard reaction and
intermediate gaseous phase diffusion i.e. CO is the rate limiting
step in overall reaction. The formation of barium sulfide involves
with value of CO that is controlled by gasification of coke.
Therefore, decreasing particle size of barite less than ꢀ400 mesh
has no effect on rate constants and activation energy conse-
quently.
gas constant, and T is the absolute temperature. Fig. 3 presents the
Arrhenius plots for three groups of particle size in absence of
catalyst. Also the values of activation energy and exponential
factor were reported in Table 4.
The values of activation energy are not significantly different for
each group of samples containing different particle size distribu-
tion. It is observed that the value of activation energy decreases as
particle size distribution of barite is shitted to finer sizes.
Activation energy falls down, approximately 9.0 kcal, when the
barite particle range changes from (ꢀ140, +230 mesh) to (ꢀ230,
+400 mesh), but more decrease in particle size has little effect on
activation energy. The trends of kinetic parameters show an overall
positive effect in the rate enhancement of reduction when particle
size is coarser than 400 mesh. The difference can be tentatively
explained by two factors that control the reduction of barite. One of
them is the effect of particle size of barite which increases the
reaction rate due to an increase in surface area and the second is
Figs. 4 and 5 present the plots of conversion–time data for
barite reduction in presence of 5% sodium carbonate and ferric
Table 4
The values of kinetic parameters in absence and presence of catalyst.
Catalyst type
E
a
(kcal/mol)
k
0
r
ꢀ140, +230
ꢀ230, +400
ꢀ400
ꢀ140, +230
ꢀ230, +400
ꢀ400
ꢀ140, +230
ꢀ230, +400
ꢀ400
7
5
5
6
6
Without catalyst
45.5
45.1
46.1
36.6
37.1
37.9
35.3
37.6
37.0
2.0 ꢂ 10
8.5 ꢂ 10
5.1 ꢂ 10
1.9 ꢂ 10
1.1 ꢂ 10
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.99
0.96
7
6
5% Na
2
CO
3
3.4 ꢂ 10
1.6 ꢂ 10
7
6
5% Fe(NO
3
)
3
2.5 ꢂ 10
1.4 ꢂ 10

1
492
A. Salem, Y.T. Osgouei / Materials Research Bulletin 44 (2009) 1489–1493
Fig. 5. Conversion–time plots for the reduction of barite at different particle size
distribution in presence of 5% Fe(NO as catalyst.
3
)
3
Fig. 7. Arrhenius plots for reduction of barite at different particle size distribution in
presence of 5% Fe(NO as catalyst.
3 3
)
nitrate. Also, the values of m and rate constants are reported in
Table 3. As may be observed the rate constants of carbothermic
reaction increases with addition of sodium carbonate, but it
obviously seems that ferric nitrate has negligible effect on k
values of reaction. To verify the influence of catalysts on
activation energy the values of ln k are plotted versus 1/T in
Figs. 6 and 7 for all particle size distribution. Also, the values of
kinetic parameters in presence of catalysts were reported in
Table 4. As previously discussed particle size distribution
changes the values of activation energy only 8.0 kcal both in
presence of sodium carbonate and ferric nitrate when the
particle size distribution decreases from (ꢀ140, +230 mesh) to
catalyst. Furthermore, the rate of reaction in all cases increases
with addition of Na CO . According to Hamilton et al. [8], in
2
3
catalyst gasification, the rate is proportional to both the number
of active sites and pores development in the carbon matrix
where reaction takes place. However, an increase in the surface
area alone is not sufficient to increase the reaction rate. The
presence of active sites is more important. Thus the catalysts
play a beneficial role in increasing the number of active sites
generated in carbon matrix. This is evident from the fact that the
0
value of frequency factor, k , which is a measure of the number
of active sites, is more in presence of ferric nitrate compared to
the non-catalytic case. However, the sodium carbonate increases
the values of frequency factor in reduction of barite more than
ferric nitrate.
(
ꢀ230, +400 mesh). The decreasing particle size less than
00 mesh changes the value of activation energy negligibly.
4
This effect is because of controlling reduction process by the
surface area of barite.
The activation energy estimated for reduction of barite in
presence of catalysts in three groups of particle size distribution
is only slightly different from the values reported in absence of
The trend of kinetic parameters shows an overall positive effect
in rate enhancement of barite reduction when particle size of
barite ranges ꢀ230, +400 mesh. The pattern of variation in kinetic
parameters show special phenomenon called isokinetic effect [6].
The existence of isokinetic effect has been proved by linear
Fig. 6. Arrhenius plots for reduction of barite at different particle size distribution in
Fig. 8. The effect of particle size on isokinetic parameters in absence and presence of
presence of 5% Na
2
CO
3
as catalyst.
catalyst.

A. Salem, Y.T. Osgouei / Materials Research Bulletin 44 (2009) 1489–1493
1493
Table 5
absence of catalyst. From the results obtained by using a modified
kinetic model, the following conclusions can be drawn:
The values of isokinetic parameters for catalytic and non-catalytic conditions.
Catalyst type
T
iso (8C)
k
iso
ꢃ
ꢃ
The presence of 5 wt.% sodium carbonate in the barite composi-
tion strongly favors reaction rate compared to the same value of
ferric nitrite.
The clear variation in particle size distribution and measuring
isothermic conversion–time data indicated that the activation
Without catalyst
1140
1033
1173
1.82
0.98
2.73
5% Na
2 3
CO
5% Fe(NO )
3 3
relationship between E
a
and k
0
as following equation:
energy decreases only 9.0 kcal. By using a particle size
distribution range of ꢀ230, +400 mesh the minimum activation
energy was obtained for catalytic and non-catalytic condition.
Further, the use of barite with particle size between ꢀ230 and
E
a
ln k
0
¼ ln kiso
þ
(8)
RT
+400 mesh allows reaching the best results in presence of
where kiso and Tiso are isokinetic rate constant, respectively.
Isokinetic temperature is described as temperature in which rate
constants at different particle size distribution have the same
values. It means that at isokinetic temperature the rate
constants are unaffected by the particle size of barite. To
sodium carbonate due to the modification in active site and
kinetic parameters. The finer size of barite, ꢀ400 mesh, could not
improve the barite reduction.
The addition of ferric nitrate to mixtures of barite and coke could
raise the number of active sites of coke. The reason for the
increase in of active sites number is the difference in frequency
factor of Arrhenius equation.
ꢃ
0 a
evaluate isokinetic effect, the plot of ln k versus E in semi-
logarithmic scale was made as shown in Fig. 8. These plots show
to be straight line, in agreement with isokinetic equation which
predicts a relationship between frequency factor and activation
energy. In Table 5, the values of kiso and Tiso are reported for
catalytic and non-catalytic conditions. The values of Tiso are
close to each other in absence and presence of ferric nitrate and
are around 1155 8C. This indicates that the particle size of barite
has no role in the reduction of barite when process was
performed at around this temperature. Table 5 data indicates
that presence of sodium carbonate as a catalyst in initial
composition may be able to decrease isokinetic temperature
considerably.
Acknowledgment
The authors would like to thank Dr. M. J. Mohammadzadeh for
his assistance with X-ray diffraction.
References
[1] J.J. Mcketta, Executive Ed. Encyclopedia of chemical processing and design, vol. 4,
Marcel Dekker, New York, 1977.
2] A.A. Ravdel, N.A. Novikova, J. Appl. Chem. USSR 36 (7) (1963) 1384.
3] S.B. Jagtap, A.R. Pande, A.N. Gokarn, Ind. Eng. Chem. Res. 29 (1990) 795.
4] A.F. Lozhkin, V.N. Pashcenko, F.V. Povar, J. Appl. Chem. USSR 47 (5) (1974) 1031.
[
[
[
4
. Conclusion
[5] Y. Pelovski, Ir. Gruchavov, I. Dombalov, J. Therm. Anal. 32 (1987) 1743.
[
[
6] A.N. Gokarn, S.D. Prodhan, G. Pathak, S.S. Skulkarni, Fuel 79 (2000) 821.
7] F.D. Snell, L. Hilton, Encyclopedia of Industrial Chemical Analysis, vol. 6, Wiley, New
York, 1984, p. 569.
In the present study, the possible use of different particle sizes
of barite in reduction process was investigated in presence and
[8] T. Hamilton, A.S. David, F. Shadman, Fuel 63 (1984) 1008.