Dry mechanochemical conversion of SrSO4 to SrCO3

Article abstract of DOI:10.1016/j.jallcom.2009.11.114

The effect of milling time on the dry conversion of celestite, SrSO₄, to SrCO₃ using sodium carbonate in a planetary type ball mill was investigated. After milling, the powders were leached in distilled water at room temperature and the solids then washed with 1 N HCl to separate the products. X-Ray diffraction analysis showed that SrSO₄ started to transform to SrCO₃ within 10 min of milling and the extent of conversion increased with milling time reaching >90% after 30 min. A <99% conversion of celestite to SrCO₃ was obtained in 30 min by increasing the molar ratio of Na₂CO₃:SrSO₄ to 1.3:1.

Full text of DOI:10.1016/j.jallcom.2009.11.114

Journal of Alloys and Compounds 492 (2010) 389–391
Contents lists available at ScienceDirect
Journal of Alloys and Compounds
journal homepage: www.elsevier.com/locate/jallcom
Dry mechanochemical conversion of SrSO to SrCO
4
3
N. Setoudeh , N.J. Welham , S.M. Azamic
a,∗
b
a
Materials Engineering Department, Yasouj University, Yasouj, Iran
Extractive Metallurgy, University of Ballarat, Ballarat, Victoria, Australia
Department of Chemistry, Yasouj University, Yasouj, Iran
b
c
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 14 June 2009
Received in revised form
4 3
The effect of milling time on the dry conversion of celestite, SrSO , to SrCO using sodium carbonate in
a planetary type ball mill was investigated. After milling, the powders were leached in distilled water at
room temperature and the solids then washed with 1 N HCl to separate the products. X-Ray diffraction
analysis showed that SrSO4 started to transform to SrCO3 within 10 min of milling and the extent of
conversion increased with milling time reaching >90% after 30 min. A <99% conversion of celestite to
SrCO3 was obtained in 30 min by increasing the molar ratio of Na2CO3:SrSO4 to 1.3:1.
1
4 November 2009
Accepted 17 November 2009
Available online 20 November 2009
©
2009 Elsevier B.V. All rights reserved.
Keywords:
Celestite
Dry milling
Mechanochemistry
SrCO3
1
. Introduction
During the last few years, applications of mechanical milling
to induce solid-state reactions [12,13], reduction of metal oxide to
metallic materials [14,15], synthesis of nanoparticles and compos-
ite materials [16–18] have been widely studied. Zhang and Saito
[19] prepared SrCO3 indirectly by carbonation of Sr(OH)2 which
was produced by dry high-energy milling of celestite and NaOH
in a planetary mill. Obut et al. [20] prepared SrCO3 from celestite
concentrate using both solid and aqueous sodium carbonate in
a planetary ball mill and showed that 93% conversion of SrCO3
could be obtained after 40 min activation. Further experiments
showed that the best conversion of celestite to strontianite was
obtained during dry milling [20]. Erdemoglu et al. [21] also stud-
ied the kinetics of the conversion of celestite to SrCO3 using wet
mechanochemical process and reported results similar to Obut et
al. [20]. The results of Obut at al. are very interesting since they
Of the naturally occurring strontium compounds, only the min-
erals strontianite (SrCO ) and celestite (SrSO ) are of economic
3
4
importance. Celestite is commercially mined and is the major start-
ing material for the production of strontium compounds. Although
strontianite occurs naturally the majority of demand is supplied by
production from celestite [1,2].
There are two commercial processes to manufacture SrCO from
3
celestite [1]. In the black ash process, SrSO4 is reduced with coke
◦
(
or coal) at temperatures generally over than 1000 C to produce
water soluble SrS [3,4]. The sulfide is then leached in hot water
and strontium carbonate is precipitated using soda ash, ammo-
nium bicarbonate and/or CO [3,5,6]. In the second process, finely
2
powdered celestite is reacted with hot sodium carbonate solution
◦
(
>90 C) to obtain SrCO3 and by-product sodium sulfate by a dou-
demonstrated a direct solid-state conversion of SrSO to SrCO and
4 3
ble decomposition reaction [7–9]. Although the black ash method
a further examination of dry milling is warranted.
produces a higher purity SrCO product it is more energy intensive
As noted previously [8,9,20,21], reaction (1) proceeds in water
3
than the double decomposition process.
Other than these two above commercial methods and their
modifications, there are several reports of the conversion of
due to the difference between the solubility products of SrSO
4
−
7
−10
(Ksp = 2.8 × 10 ) and SrCO3 (Ksp = 9.4 × 10
). This equates to a
at 298 K which is why
−
1
net free energy change of −22.3 kJ mol
celestite to SrCO by other methods including the high temperature
dry milling will also induce the reaction.
3
and high-pressure reaction between celestite and either sodium or
potassium carbonate [10] and direct leaching of celestite with Na2S
SrSO4 + Na2CO3 = SrCO3 + Na2SO4
(1)
to produce SrCO [11].
3
This article examines the conversion of a purified commercial
celestite concentrate to strontium carbonate in a dry planetary ball
mill. The aim is to demonstrate the effect of milling time on the
∗ Corresponding author. Tel.: +98 741 2221711; fax: +98 741 2221711.
E-mail address: nsetoudeh@mail.yu.ac.ir (N. Setoudeh).
changes during milling which leads to the formation of SrCO3.
0
925-8388/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.jallcom.2009.11.114

3
90
N. Setoudeh et al. / Journal of Alloys and Compounds 492 (2010) 389–391
2.
Experimental
Purified celestite concentrate (Likak Mine, Yasouj, Iran), reagent grade anhy-
drous Na2CO3 and distilled water were used in this study. The concentrate was
crushed, ground and screened to give a starting particle size range of 53–106 ␮m.
The anhydrous Na2CO3 was also screened to give the same particle size range. The
ground concentrate was washed with dilute HCl (1 N) solution to remove any car-
bonate impurity present. After acid washing and filtration, the solid residue was
◦
washed several times with distilled water then, the solid parts dried at 110 C for
2
h in a laboratory oven. X-ray fluorescence indicated the concentrate contained
about 92% SrSO4 and after washing in HCl 98% SrSO4, the remaining 2% was gypsum,
CaSO4·2H2O.
A planetary type ball mill (Farapazhouhesh, Iran, FP2 model) was used in the
milling experiments. Preliminary experiments indicated the following conditions
were appropriate: 600 rpm, five 20 mm diameter high Cr-steel balls in a high Cr-
steel milling chamber gave a ball-to-sample weight ratio of 40:1. The mixture
of raw materials (purified celestite concentrate and sodium carbonate) was pre-
pared in accordance with the stoichiometry given by reaction (1). The molar ratio
of SrSO4:Na2CO3 was also examined between 1:1 and 1:2 to assess the effect of
changing this parameter.
After each milling run, the product powder was washed with water to remove
any remaining Na2CO3 and the Na2SO4 produces during reaction. The solid residue
◦
was dried at 105 C for 1 h in a laboratory oven, weighed then leached with 1 N
Fig. 2. Mass loss after leaching in 1 N HCl, for a SrSO4:Na2CO3 molar ratio of 1:1.
hydrochloric acid to dissolve all phases except the SrSO4. All runs were performed
in triplicate and the average results reported. The variation between runs was <0.5%.
XRD analyses (XRD, BRUKER D8-Advance, 40 kV, 40 mA, with EVA software
◦
their intensities are comparatively weak. This data indicates that
the conversion of celestite into strontium carbonate starts within
8
2
,0,0,2 version 8) were performed using Cu K␣ radiation over a range of 15–70
◦
ꢀ, using a step size of 0.05 and 3 s count time per step.
1
0 min of dry milling in a planetary ball mill.
The weakening of the peaks cannot be due to amorphisation
3
. Results and discussion
of the celestite. There are two mechanisms by which the celestite
peaks will disappear from the XRD, consumption by reaction and
amorphisation by extended milling. If amorphisation was tak-
Fig. 1 shows the XRD pattern of the acid washed concentrate,
only peaks for celestite (JCPDS card No. 005-0593) were observed.
Fig. 1 also shows XRD patterns of the as-milled samples milled
for up to 90 min. Even after as little as 10 min the major peaks for
◦
ing place then a broad peak would become evident around 30
2
ꢀ, which is clearly not the case. It is well known that during
the milling of binary mixtures the softer phase is preferentially
amorphised. In the system under investigation the softer phase
celestite are clearly present. There are some peaks for Na SO but
2
4
is the Na CO3 therefore the peaks for this phase would weaken
2
and broaden more rapidly than those for SrSO , again, this is not
4
the case. Amorphisation would not result in increased dissolu-
tion in acid as the solubility will only change if there are phase
changes. There are clearly peaks present for new, positively iden-
tified, phases after 10 min and these peaks increase in intensity
as those for SrSO4 weaken. Clearly, a gradual chemical reaction is
occurring.
Increasing the milling time beyond 20 min has only a minor
effect on the trace. Indeed, all peaks become less intense and
broader as would be expected as the crystallite size becomes
smaller with increasing milling time. The major peaks for celestite
decrease in intensity but do not disappear completely implying
the reaction reached a certain extent and did not proceed further
despite the additional milling time.
It has been reported that the greatest extent of conversion of
celestite to strontianite is obtained during dry milling [20]. It is
known that wet milling and/or the use of small balls is more favor-
able for increasing the surface area whilst dry milling and/or the
use of larger balls favors solid-state reactions due to the increased
breakage–rewelding of particles [22]. Therefore, the results of Fig. 1
verify the role of milling in progressing the solid-state reaction.
Fig. 2 shows the mass loss in 1 N HCl. The starting mixture
shows 36.6% mass loss, i.e. only Na CO was soluble. After 10 min
2
3
milling there is a significant increase in solubility due to reaction
1) which indicates the formation of acid soluble SrCO and Na SO .
(
3
2
4
The extent of mass loss beyond 30 min is only slightly higher than
at 30 min, evidently the reaction was substantially slower beyond
3
0 min, and presumably this was due to the decreased chance of
unreacted solids sharing a common interface. No increase in mass
loss was evident beyond 60 min of milling implying the reaction
was at its greatest extent.
The maximum extent of reaction is clearly less than 100% as both
products of reaction (1) are soluble in the acid used. Obut et al. [20]
Fig. 1. XRD pattern of celestite concentrate and samples after dry milling for up to
0 min.
9

N. Setoudeh et al. / Journal of Alloys and Compounds 492 (2010) 389–391
391
Previous works [7–9] have showed that a superstoichiomet-
ric amount of sodium carbonate led to enhanced conversion of
celestite to strontium carbonate. Obut et al. [20] also showed that
beyond a 20% excess of Na CO no significant increase in the con-
2
3
version occurred.
. Conclusions
It has been shown that the conversion of a celestite concen-
4
trate to strontium carbonate can be obtained by high-energy dry
milling in the presence of Na CO . Increasing milling time leads to
2
3
increase in the rate of conversion due to creation of fresh reaction
interfaces due to breakage and rewelding of particles. The conver-
sion of celestite to strontium carbonate started within 10 min and
was essentially complete within 30 min. The results showed that
>
90% conversion could be obtained after 30 min milling of a 1:1
molar ratio of SrSO :Na CO . Increasing the SrSO :Na CO molar
4
2
3
4
2
3
Fig. 3. Mass loss in 1 N HCl as a function of stoichiometry.
ratio to 1:1.3, increased the extent of conversion to 99% after 30 min
milling.
have shown that the dissolution of wet milled solids (SrSO –SrCO3
4
Acknowledgement
mixture) was 93.0% after 20 min wet milling in a planetary ball
mill, whereas in a dry planetary ball mill the percent of dissolu-
tion in HCl reached to about 99% after 20 min. They also showed
that the increase in dissolution between 20 and 40 min was smaller
than the increase between 10 and 20 min for wet milling. It should
be noted that Obut et al. [20] used a SrSO :Na CO mole ratio
The authors gratefully acknowledge Deputy of Research and
Technology of Yasouj University for financial support for the project
“Preparation of strontium carbonate via a mechanochemical pro-
cess” (Grant No.1316).
4
2
3
of 1:1.2 and did their milling runs with a ball-to-sample weight
ratio 20:1.
References
The effect of the molar ratio SrSO :Na CO on the dissolution in
HCl is shown in Fig. 3. Clearly, increasing the ratio from 1:1 to 1:1.3
gave an increase in the extent of reaction; however, any greater
4
2
3
[1] F. Habashi, Handbook of Extractive Metallurgy, IV, Wiley VCH, 1997, pp. 2329-
2336.
[
2] J.A. Ober, Minerals Yearbook—Strontium, U.S. Geological Survey, 2006, 73.1-
73.5.
excess of Na CO had no major effect. In all cases there remained
2
3
a small insoluble fraction (∼0.4%) suggesting the gypsum in the
[3] W. Chen, Y. Zhu, Trans. Inst. Min. Metall. (Sect. C: Miner. Process Extr. Metall.)
09 (2000) C1–C5.
4] R.S. Sonawane, B.B. Kale, S.K. Apte, M.K. Dongare, Metall. Mater. Trans. 31B (1)
2000) 35–41.
[5] M. Erdemoglu, M. Canbazoglu, Hydrometallurgy 49 (1998) 135–150.
1
starting strontianite was insoluble in the acid.
[
On the basis of the above results it seems that the greatest
conversion is achieved by milling a molar ratio of 1:1.2–1:1.3
for 30 min. There are several possible reasons for the conversion
not achieving 100%, the most probable of which are consumption
of the SrSO4 and nearing thermodynamic equilibrium. The start-
ing ore contained only ∼98% SrSO4 as natural minerals inevitably
contain impurities and the calculation method did not take the
insoluble fraction into account. There was a consistent acid insol-
uble mass of 0.4% regardless of the milling time and molar ratio
(
[
[
[
6] G. Owusu, J.E. Litz, Hydrometallurgy 57 (2000) 23–29.
7] M. Iwai, M. Toguri, Hydrometallurgy 22 (1989) 87–100.
8] F.R. Carrillo-Pedroza, A. Uribe-Salas, A.H. Castillejos-Escobar, Miner. Eng. 8 (4/5)
(1995) 495–509.
[
9] A.H. Castillejos-Escobar, F.P. De La Cruz-De Bosque, A. Uribe-Salas, Hydromet-
allurgy 40 (1996) 207–222.
10] R. Suárez-Ordu n˜ a, J.C. Rendón-Angeles, K. Yanagisawa, Int. J. Miner. Process. 83
(2007) 12–18.
[
[
[
11] M. Erdemoglu, M. Sarıkaya, M. Canbazoglu, J. Dispers. Sci. Technol. 27 (2006)
439–442.
of Na CO :SrSO . The equilibrium constant for reaction (1) was
2
3
4
◦
12] L. Takacs, Hyperfine Interact. 111 (1–4) (1998) 245–250.
[13] C.-O. Li, M.-F. Zhang, H.-B. Zho, J.-C. Han, S.-H. Meng, J. Inorg. Mater. 23 (6)
(2007) 1131–1134.
14] X. Shengqi, Q. Xiaoyan, M. Mingliang, Z. Jingen, Z. Xiulin, W. Xiaotian, J. Alloys
Compd. 268 (1–2) (1998) 211–214.
calculated at 25 C using the above solubility products giving a
value of 7931. From the reaction stoichiometry we can calculate
the fractional conversion which will lead to this equilibrium con-
[
[
SrCO ][Na SO ]
3 2 4
stant using K =
. If we start with 1 mol of [SrSO ] and
[15] J. Kanoa, E. Kobayashia, W. Tongampb, S. Miyagia, F. Saitoa, J. Alloys Compd.
80 (2) (2009) 666–669.
16] J.-L. Bobet, F.J. Castro, B. Chevalier, J. Alloys Compd. 376 (1–2) (2004) 205–210.
4
[SrSO ][Na2CO3]
4
4
[
Na CO ] and converted a fraction, x, of the starting amount, then
2 3
[
at equilibrium [SrCO ] = [Na CO ] = x, and [SrSO ] = [Na CO ] = 1–x.
3
2
3
4
2
3
[17] D.P. Xiang, Y. Liu, M.J. Tu, Y.Y. Li, S. Xia, B.Q. Chen, J. Alloys Compd. 473 (1–2)
(2009) 453–457.
[18] S. Palaniandy, N. Hidayu Jamil, J. Alloys Compd. 476 (1–2) (2009) 894–902.
[19] Q. Zhang, F. Saito, Chemical Eng. J. 66 (1997) 79–82.
20] A. Obut, P. Balaz, I. Girgin, Miner. Eng. 19 (2006) 1185–1190.
21] M. Erdemoglu, S. Aydogan, M. Canbazoglu, Hydrometallurgy 86 (2007) 1–5.
22] P. Balaz, Mechanochemistry in Nanoscience and Minerals Engineering,
Springer-Verlag, Berlin, Heidelberg, 2008.
2
x
x.x
1−x)(1−x)
Substituting these gives K = (
ranged into a quadratic soluble for x. For K = 7931, x = 0.9889, i.e. a
8.9% conversion is expected for equilibrium thus we should not
=
, which can be rear-
2
(
1−x)
[
[
[
9
expect to have a 100% conversion. On this basis, a maximum acid
solubility of 99.30% can be calculated; this value is within experi-
mental error of the dissolution determined earlier.