Indian Journal of Chemical Technology
Vol. 13, July 2006, pp. 391-397
Determination of the optimum conditions for dissolution of
magnesite with H2SO4 solutions
Yüksel Abalia*, Mehmet Çopurb & Mesut Yavuza
a
Celal Bayar University, Science and Arts Faculty, Chemical Department, 45030 Manisa, Turkey
b
Atatürk University, Engineering Faculty, Chemical Engineering Department, 25240 Erzurum, Turkey
Email: [email protected]
Received 6 May 2005; revised received 9 May 2006; accepted 15 May 2006
Basic data on leaching of magnesite (MgCO3) with sulphuric acid are of interest from the point of view of the industrial
process for obtaining pure MgSO4. The Taguchi method was used to determine the optimum conditions for the dissolution
of magnesite in H2SO4 solutions. The experiments were performed within the ranges mentioned herein i.e. 20-65°C for
reaction temperature, 0.5/100-10/100 g/mL for solid-to-liquid ratio, 0.2-5 M for acid concentration, 5-60 min for reaction
time and 150-750 rpm for stirring speed. The optimum conditions for these factors were found to be 65°C, 5/100 g/mL, 2 M,
60 min and 300 rpm, respectively. Under these conditions, the dissolution mass fraction of MgCO3 in H2SO4 solutions were
w = 96.32%.
Keywords: Optimisation, Magnesite, Sulphuric acid, Taguchi method
IPC Code: C01B6/21
Magnesite (MgCO3) is the primary source for
production of magnesium and its compounds. Natural
magnesite contains theoretically 47.6% MgO and
some impurities such as silisium, iron and calcium. It
is the basic raw material for the manufacture of
alkaline refractory and is used in iron-steel, cement,
glass, sugar, lime and paper industries as well as paint
and ink industry, pharmaceutical industry as an antiacid, and in the production of many magnesium
chemicals1,2.
Basic data on leaching of magnesite with sulphuric
acid are of interest from the point of view of the
industrial process for obtaining pure MgSO4.
Magnesium sulphate does not dissociate completely
in water. The anhydrous sulphate cannot be obtained
from solution, but only by dehydration of one of its
hydrates. Hydrolytic decomposition may take place at
relatively low temperatures (250°C), but if the heating
is carried out in the presence of a small amount of
concentrated sulphuric acid, a relatively stable
anhydrous product is obtained which can be heated
without further decomposition to about 800°C (ref. 2).
Many studies have been carried on the dissolution
of magnesite minerals to produce magnesium
compounds3-8. Ekmekyapar et al.9 investigated the
dissolution kinetics of magnesite ore with sulphuric
acid and it was found that the dissolution rate is
controlled by the surface reaction. The activation
energy for the reaction was calculated to be 55 kJ.mol-1.
Shcherbakova et al.10 manufactured MgSO4.7H2O by
treating magnesite with sulphuric acid in the presence
of a magnesium sulphate main liquor in which the
corrosiveness of the medium was decreased and
productivity of the process was increased. Abali
et al.11 reported, that, the activation energy for
dissolution kinetics of magnesite mineral in water
saturated by SO2 gas was 81 kJ.mol-1 and the reaction
was controlled by the surface chemical reaction.
Özbek et al.12 investigated the dissolution kinetics of
magnesite with Cl2 gas in water. They found that the
reaction was controlled by diffusion through the fluid
layer. Chou et al.13, investigated the dissolution of
various carbonates (including calcite, magnesite and
dolomite) in HCl solutions at 25°C by using a
continuous fluidized bed reactor and samples of
relatively coarse particle size. Kennedy and Harris14
studied a chlorination study of magnesium carbonate
in a stirred tank reactor. Chlorination rates were
measured over a range of temperatures of 740 to
910°C. Activation energy of process was calculated as
80 kJ mol-1 over the temperature range of 740 to
825°C and the fastest chlorination was reached at a
860°C. Demir et al.15 found that the leaching kinetics
of magnesite in citric acid solutions was controlled by
392
INDIAN J. CHEM. TECHNOL., JULY 2006
chemical reaction in developing semi-empirical
model. The activation energy of the process was
determined as 61.35 kJ mol-1. Also, Harris et al.16
studied the production of magnesium from
concentrated
magnesium
chloride
solutions.
Optimisation for the dissolution of ores in different
acidic media has been investigated by a number of
authors17-23.
Taguchi's Orthogonal Array (OA) analysis is used
to produce the best factors for the optimum design
process, with the least number of experiments. In
recent years, Taguchi method has been used to
determine optimum factors because of its
advantages21,22. The main advantages of this method
over other statistical experimental design methods are,
that the factors affecting an experiment can be
investigated as controlling and not controlling, and
that the method can be applied to experimental design
involving a large number of design factors.
In this study, the magnesite was dissolved in H2SO4
solutions, and Taguchi experimental design method
was employed to determine optimum leaching
conditions.
Experimental Procedure
Magnesite mineral used in the experiments was
obtained from Hasankale Region in Erzurum, Turkey.
The sample was sieved by using a -315+140 μm
ASTM standard sieve. The chemical composition of
the ore was determined by volumetric and gravimetric
methods26 and the result is given in Table 1. In
addition, X-ray analysis showed that the main mineral
is magnesite (Fig. 1). The average density of
magnesite has been determined as 2.947 g.cm-3.
The dissolution experiments were carried out in a
250 mL glass reactor equipped with a mechanical
stirrer having a digital controller unit and timer, a
thermostat and a cooler. The temperature of the
reaction medium could be controlled within ±0.5°C.
First 100 mL H2SO4 of known concentration was
introduced into the reactor. After the desired reaction
temperature was reached, a predetermined amount of
the magnesite was added into the solution while the
content of the vessel was stirred at a certain speed. At
the end of the experiment, the contents of the vessel
were filtered by a filter paper of Filtrak 391 and the
filtrate solution was analyzed volumetrically for Mg26.
The use of the quantity design in the Taguchi
method to optimize a process with multiple
performance characteristics includes the following
steps: (a) to identify the performance characteristics
and select process factors to be evaluated; (b) to
determine the number of quantity levels for the
process and possible interaction between the process
factors; (c) to select the appropriate orthogonal array
and assignment of process factors to the orthogonal
array; (d) to conduct the experiments based on the
arrangement of the orthogonal array; (e) calculate the
performance characteristics; (f) to analyze the
experimental result using the performance
characteristic and ANOVA; (h) to select the optimal
levels of process factors; and (i) to verify the optimal
process
factors
through
the
confirmation
experiment23,27. Experimental factors and their levels,
determined in the light of preliminary tests, are given
in Table 2.
The orthogonal array (OA) experimental design
was chosen as the most suitable method to determine
experimental plan, L25 (55) (Table 3), five factors each
with five values27. In order to observe the effects of
noise sources on the dissolution process, each
experiment was repeated twice under the same
conditions at different times. The performance
characteristics were chosen as the optimization
criteria. There are three categories of performance
characteristics, the larger-the-better, the smaller-thebetter and the nominal-the-better. The two
performance characteristics were evaluated by using
Eqs (1) and (2)28,29.
Fig. 1—X-ray diffractogram of magnesite ore
Table 1—The chemical composition of magnesite mineral used in
the study
Component
wt%
MgO
CaO
Fe2O3
SiO2
Ignition loss
46.36
1.06
0.41
0.71
51.46
ABALI et al: STUDIES ON DISSOLUTION OF MAGNESITE WITH H2SO4 SOLUTIONS
393
Table 2—Parameters and their values corresponding to their levels studied in the experiment
Parameters
A
B
C
D
E
o
Reaction temperature, C
Solid-to-liquid ratio, g/mL
Acid concentration, M
Reaction time, min
Stirring speed, rpm
1
2
Parameter levels
3
4
5
20
0.5/100
0.2
5
150
35
1/100
0.5
10
300
50
2/100
1.0
20
450
65
5/100
2.0
30
600
75
10/100
5.0
60
750
Larger-the-better SNL
1 ⎞
⎛1
SNL = −10 log⎜ ∑ 2 ⎟
⎝n Y ⎠
…(1)
Smaller-the-better SNS
⎛ Y2
SNS = −10 log⎜⎜ ∑
n
⎝
⎞
⎟⎟
⎠
…(2)
where smaller-the-better SNL and Smaller-the-better
SNS are performance characteristics, n number of
repetition done for an experimental combination, and
Yi performance value of ith experiment.
In Taguchi method, the experiment corresponding
to optimum working conditions might not been done
during the whole period of the experimental stage. In
such cases the performance value corresponding to
optimum working conditions can be predicted by
utilizing the balanced characteristic of OA. For this
the following additive model may be used27;
Y = μ + ∑ X i + εi
…(3)
where μ is the overall mean of performance value, Xi
the fixed effect of the quantity level combination used
in ith experiment, and ei the random error in ith
experiment.
If experimental results are in percentage (%),
before evaluating Eq. (3) transformation of percentage
values should be applied first using Eq. (4) by which
values of interest are also determined later by carrying
out reverse transformation by using the same
equation30:
⎛ P ⎞
Ωi = −10 ⋅ log ⎜ i ⎟
⎝ 1 − Pi ⎠
…(4)
where Ω (dB) is the decibel value of percentage value
subject to omega transformation and the percentage of
the product obtained experimentally. Because Eq. (3)
is a point estimation, which is calculated by using
experimental data in order to determine, whether the
additive model is adequate or not, the confidence
limits for the prediction error must be evaluated27.
The prediction error is the difference between the
observed Yi and the predicted Yi. The confidence
limits for the prediction error, Se, is sum of squares
due to error degrees of freedom for error
Se = ±2
1 2 1 2
σe + σe
n0
nr
1 1 ⎡ 1 1⎤ ⎡ 1 1⎤
= +⎢
− ⎥ + ⎢ − ⎥ + ...
n0 n ⎣ nAi n ⎦ ⎣ nBi n ⎦
σe2 =
sum of squares due to error
degrees of freedom for error
…(5)
…(6)
…(7)
where Se is the two-standard-deviation confidence
limit, n the number of rows in the matrix experiment,
nr the number of repetition in confirmation
experiment and nAi, nBi, nCi, … are the replication
number for variables quantity level Ai, Bi, Ci,… If the
prediction error is outside these limits, it should be
suspected of the possibility that the additive model is
not adequate. Otherwise, additive model can be
considered to be adequate. A verification experiment
is a powerful tool for detecting the presence of
interactions among the control quantities. If the
predicted response under the optimum conditions
does not match the observed response, then it implies
that the interactions are important. If the predicted
response matches the observed response, then it
implies that the interactions are probably not
important and that the additive model is a good
approximation27.
INDIAN J. CHEM. TECHNOL., JULY 2006
394
The order of the experiments was obtained by
inserting factors into columns of OA, L25 (55), chosen
as the experimental plan given in Table 3. But the
order of experiments was made random in order to
avoid noise sources which had not been considered
initially and which could take place during an
experiment and affect results in a negative way.
Equation (8) shows the particles dissolve and
become progressively smaller in size, and there is
no solid product layer formed during the leaching
reaction. Hence, the possibility of ash layer
diffusion is not present. On the other hand, if the
rate is very sensitive to the temperature variations,
chemical reaction may be considered to be the ratecontrolling factor31.
Results and Discussion
Statistical analysis
Dissolution reactions
So as to see effective parameters and their
confidence levels on dissolution process, the analysis
of variance was performed. A statistical analysis of
variance (ANOVA) was performed to see whether
process parameters are statistically significant or not.
F-test is a tool to see which process factors have a
significant effect on the dissolution value. The F value
for each process parameter is simply a ratio of mean
of the squared deviations to the mean of squared
error. Usually, the larger the F value, the greater the
effect on the dissolution value due to the change of
the process parameter. With the performance
To determine the optimum conditions for the
dissolution of magnesite in sulphuric acid solutions,
the effect of reaction temperature, solid-to-liquid
ratio, acid concentration, reaction period and stirring
speed were investigated. The experimental results are
given in Table 3.
The dissolution reaction of magnesite in sulphuric
acid solutions can be described by the following
equation:
MgCO3(s)+H2SO4(aq)→MgSO4(aq)+CO2(g)+H2O(l)
…(8)
Table 3—L25 (55) Experimental reaction conditions
Experiment
No
A
B
C
D
E
XMg
(a)
XMg
(b)
XMg
(average)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
1
1
1
1
1
2
2
2
2
2
3
3
3
3
3
4
4
4
4
4
5
5
5
5
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
1
2
3
4
5
2
3
4
5
1
3
4
5
1
2
4
5
1
2
3
5
1
2
3
4
1
2
3
4
5
3
4
5
1
2
5
5
2
3
4
2
3
4
5
1
4
5
1
2
3
1
2
3
4
5
4
5
1
2
3
2
3
4
5
1
5
1
2
3
4
3
4
5
1
2
0.0819
0.0745
0.09226
0.1339
0.1370
0.4270
0.3024
0.4637
0.2174
0.0963
0.9935
0.3564
0.4547
0.2626
0.3365
0.9765
0.9638
0.8961
0.9568
0.2870
0.9865
0.9560
0.3258
0.5625
0.9436
0.0894
0.0772
0.1108
0.1448
0.1929
0.2868
0.2854
0.3688
0.2338
0.0841
0.9322
0.2248
0.4201
0.2458
0.3320
0.8177
0.9388
0.8617
0.8692
0.4272
0.9938
0.9773
0.4498
0.7961
0.9360
0.0856
0.0758
0.1015
0.1395
0.1649
0.3569
0.2939
0.4162
0.2256
0.0902
0.9628
0.2906
0.4374
0.2542
0.3342
0.8971
0.9515
0.8789
0.9030
0.3571
0.9901
0.9666
0.3878
0.6793
0.9398
ABALI et al: STUDIES ON DISSOLUTION OF MAGNESITE WITH H2SO4 SOLUTIONS
characteristics and ANOVA analyses, the optimal
combination of process parameters can be predicted20.
The results of variance analysis were given in
Table 4.
To obtain optimal dissolution performance, the
higher-the-better performance characteristic [Eq. (1)]
has been taken for dissolution of magnesite ore. The
order of graphs in Fig. 2 is according to the degrees of
the influences of factors on the performance
characteristics. The optimal level of a process
quantity is the level with the highest SN value
calculated by Eq. (1).
For the dissolution of MgCO3 within the chosen
experimental range (Table 2) the reaction
temperature, dissolution time, solid-to-liquid, and the
stirring speed have significant effects on the
dissolution process while the acid concentration has
no effect.
Under the light of the above consideration for a
heterogeneous reaction system, the mechanism
controlling reaction rate can be determined by
considering some factors which affect the reaction
rate. Accordingly, it can be deduced that for the
processes whose temperature is more effective, the
rate is controlled by chemical reaction while for the
processes whose stirring speed is more effective, the
rate is controlled by diffusion. The activation energy
value of 48.15 kJ.mol–1 for this dissolution process29
confirms this conclusion. Therefore, for the present
work it can be stated that the dissolution rate of
magnesite is controlled by chemical reaction33.
The effectiveness of the parameter on optimization
criteria and the selected optimum reaction conditions
are shown in Table 5. The conditions which provided
maximum amount in minimum cost are selected as
optimum reaction conditions for MgSO4 production.
As it is seen in the tables, the optimum process
395
conditions for dissolving magnesite mineral and the
formation of magnesium sulphate are chosen as A4,
B1, C5, D5, E2. Based on these conditions 96.32%
magnesium carbonate dissolution was achieved,
• Reaction temperature = 65°C
• Solid-to-liquid ratio =0.5/100 g.mL-1.
• Acid concentration = 2.0 M
• Reaction time = 60 min
• Stirring speed = 300 rpm
• Predicted Value = 100 %
• Observed Value = 96.32 %
• Confidence limits for the prediction error= 84.2100
The fact that the dissolution percentages from
confirmation experiments are within the calculated
confidence intervals calculated from Eq. (7), and the
experimental results are within ±5% in error
(significance level 95%), this case states that there is a
good agreement between the predicted values and
experimental values, and the interactive effects of the
factors are indeed negligible. It may be concluded that
the additive model is adequate for describing the
dependence of this dissolution process on various
factors27.
Fig. 2—The effect of each parameter on the performance statistics
for magnesium carbonate
Table 4—The results of variance analysis
Parameters
DF
(Degrees of freedom)
SS
(Sum of squares)
MS
(Mean of squares)
F
(Test statistic)
A
Reaction temperature, °C
4
37589.7
9397.4
162.68
B
Solid-to-liquid ratio, g/mL
4
4610.1
1152.5
19.95
C
Acid concentration, M
4
1340.8
335.2
5.80
D
Reaction time, min
4
9183.8
2296.0
39.75
E
Stirring speed, rpm
4
2583.5
645.9
11.18
Error
29
1675.2
57.8
Total
49
56983.1
1162.92
INDIAN J. CHEM. TECHNOL., JULY 2006
396
Table 5—Determination of optimum parameter values
Parameters
A
B
C
D
E
Reaction temperature (oC)
Solid-to-liquid ratio (g/mL)
Acid concentration (M)
Reaction time (min)
Stirring speed (rpm)
Parameter
values
20
35
50
65
75
0.5/100
1/100
2/100
5/100
10/100
0.2
0.5
1.0
2.0
5.0
5
10
20
30
60
150
300
450
600
750
Conclusions
The major conclusions drawn from the present
work include:
(i) the most important factor affecting the solubility
is dissolution temperature. The dissolution of
magnesite increases with increasing temperature, acid
concentration, dissolution time and stirring speed, and
decreasing dissolution mass concentration.
(ii) the optimum conditions are 65°C for reaction
temperature, 0.05 for solid-to-liquid ratio, 5 M for
acid concentration, 60 min for reaction time, and 750
rpm for stirring speed. Under these conditions, the
dissolution is 100% for MgCO3 and confidence level
84-100.
(iii) the predicted and observed dissolution values
are close to each other; it may be concluded that the
additive model is adequate for describing the
dependence of dissolution process on various factors.
(iv) since optimum conditions determined by
Taguchi method in laboratory environment are
reproducible in real production environments as well,
the findings of the present study may be very useful
for processing in industrial scale.
(v) dissolution rates of magnesite are controlled by
chemical reaction
Conversion
fractions (XMg)
0.1134
0.2766
0.4558
0.7975
0.7927
0.6585
0.5157
0.4444
0.4403
0.3772
0.4551
0.4115
0.4789
0.5366
0.5539
0.2693
0.4360
0.5208
0.5273
0.6827
0.4934
0.6166
0.4751
0.4515
0.3996
Cost
Selection
Min.
A4
Max.
Max.
B1
Min.
Min.
C4
Max.
Min.
D5
Max.
Min.
E2
Max.
References
1 Kraus E H, Hunt W F & Raksdell L S, Minerology, 4th edn
(McGraw Hill Book Company), 1951 .
2 Kirk R E & Othmer D F, Encyclopedia of Chemical
Technology, Vol. 8 (The Interscience Encyclopedia, Inc.,
New York), 1952, 229.
3 Filippou D, Katiforis N, Papassiopi N & Adam K, J Chem
Technol Biotechnol, 74 (1999) 322.
4 Ekmekyapar A, Erşahan H & Dönmez B, Doğa-Tr J Eng
Environ Sci, 17 (1993) 197.
5 Bengtson K B, TMS Annual Meeting, Warrendale, USA,
1999, 1151.
6 Erşahan H, Ekmekyapar A & Sevim F, Int J Miner Proces,
42 (1994) 121.
7 Abalı Y, Energy, Educ, Sci Technol, 1(2) (1998) 67.
8 Ainscow W S, Refract J, 59(3) (1984) 6.
9 Ekmekyapar A, Kocakerim M M & Çolak S, Erciyes Univ
Fen Bil Derg, 7(1) (1991) 1132.
10 Scherbakova V V, Polyakov E V & Fairushin L, USSR Pat.
827 395, 1981.
11 Abalı Y, Çolak S & Ekmekyapar A, Doğa-Tr J Eng Environ
Sci, 16 (1992) 319 (Turkish).
12 Özbek H, Abalı Y, Çolak S, Ceyhun İ & Karagölge Z,
Hydrometallugy, 51 (1999) 173.
13 Chou L, Garrels R M & Wollast R, Chem Geology, 78
(1989) 269.
14 Kennedy M & Harris R, Can Metal Q, 39(3) (2000) 269.
15 Demir F, Dönmez B & Colak S, J Chem Eng Jpn, 36 (6)
(2003) 683.
ABALI et al: STUDIES ON DISSOLUTION OF MAGNESITE WITH H2SO4 SOLUTIONS
16 Harris G B, Peacey J G & Monette S, Chem Abstr, 109
(1988) 24855c.
17 Abalı Y, Çolak S & Yapıcı S, Hydrometallurgy, 46 (1997)
27.
18 Demirbaş A, Abalı Y & Mert E, Res Conservation Recycling,
26 (1999) 251.
19 Syrcos G P, J Mater Proces Technol, 135 (2003) 68.
20 Ata O N, Çolak S, Ekinci Z & Çopur M, Chem Biochem Eng
Q, 24 (2001) 409.
21 Çopur M, Chem Biochem Eng Q, 15(4) (2002) 191.
22 Yartaşı A, Çopur M, Özmetin C, Kocakerim M M & Temur
H, Energy Educ Sci Technol, 3(2) (1999) 77.
23 Çopur M, Pekdemir T, Çelik C & Çolak S, Ind Eng Chem
Res, 36 (1997) 682.
24 Nian C Y, Yang W H & Tarng Y S, J Maters Proces
Technol, 95 (1999) 90.
397
25 Barrado E, Vega M, Pardo R, Grande P & Del Valle J L,
Water Res, 30(10) (1996) 2309.
26 Furman N H, Standard Method of Chemical Analysis (DVan
Nostrand Comp, Inc. New York), 1963.
27 Phadke M S, Quality Engineering Using Robust Design
(Prentice Hall: New Jersey), 1989, 61.
28 Phadke M S, Kackar R N, Speeney D V & Grieco M J, Bell
System Tech J, 62(5) (1983) 1273.
29 Pignatiello & Joseph J Jr, IIE Transactions, 20(3) (1988)
247.
30 Taguchi G, System of Experimental Design (Quality
Resources, New York), 1987, 108.
31 Levenspiel O, Chem Reaction Eng, 2nd edn (John Wiley and
Sons, New York), 1972, 372.
32 Çopur M, Chem Biochem Eng Q, 15(4) (2001) 181.
33 Jackson E, Hydrometallurgical Ext Reclamation (Ellis
Horwood Limited: Chichester), 1986, 42.