Waste Management xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Waste Management
journal homepage: www.elsevier.com/locate/wasman

Synthesis of high-purity precipitated calcium carbonate during the

process of recovery of elemental sulphur from gypsum waste
M. de Beer a,d,⇑, F.J. Doucet b,⇑, J.P. Maree c, L. Liebenberg d
a
DST/CSIR National Centre for Nanostructured Materials, Council for Scientific and Industrial Research, PO Box 395, Pretoria 0001, South Africa
b
Council for Geoscience, Private Bag X112, Pretoria 0001, South Africa
c
Department of Environmental, Water and Earth Science, Faculty of Science, Tshwane University of Technology, Private Bag X680, Pretoria 0001, South Africa
d
Centre for Research and Continued Engineering Development (Pretoria), North-West University, South Africa

a r t i c l e i n f o a b s t r a c t

Article history: We recently showed that the production of elemental sulphur and calcium carbonate (CaCO3) from gyp-
Received 18 May 2015 sum waste by thermally reducing the waste into calcium sulphide (CaS) followed by its direct aqueous
Revised 13 August 2015 carbonation yielded low-grade carbonate products (i.e. <90 mass% as CaCO3). In this study, we used
Accepted 17 August 2015
the insight gained from our previous work and developed an indirect aqueous CaS carbonation process
Available online xxxx
for the production of high-grade CaCO3 (i.e. >99 mass% as CaCO3) or precipitated calcium carbonate
(PCC). The process used an acid gas (H2S) to improve the aqueous dissolution of CaS, which is otherwise
Keywords:
poorly soluble. The carbonate product was primarily calcite (99.5%) with traces of quartz (0.5%). Calcite
Gypsum waste
Indirect carbonation
was the only CaCO3 polymorph obtained; no vaterite or aragonite was detected. The product was made
Precipitated calcium carbonate up of micron-size particles, which were further characterised by XRD, TGA, SEM, BET and true density.
Calcium sulphide Results showed that about 0.37 ton of high-grade PCC can be produced from 1.0 ton of gypsum waste,
Valorisation and generates about 0.19 ton of residue, a reduction of 80% from original waste gypsum mass to mass
of residue that needs to be discarded off. The use of gypsum waste as primary material in replacement
of mined limestone for the production of PPC could alleviate waste disposal problems, along with con-
verting significant volumes of waste materials into marketable commodities.
Ó 2015 Published by Elsevier Ltd.

1. Introduction
Disposal of industrial by-products is becoming an increasingly
serious constraint on many industries. Gypsum waste
(CaSO4
2H2O) is a calcium-rich solid residue generated in the min-
ing industry as a result of the treatment of acid mine drainage
(AMD). Acidic water is naturally generated by the weathering of
sulphide minerals that are exposed to atmospheric conditions dur-
ing the mining of valuable ores (Smith et al., 2013). Most of the
AMD treatment systems use limestone or lime as acid neutralizing
agents, which results in the production of gypsum sludge (Kalin
et al., 2006; Johnson and Hallberg, 2005). Long-term storage and
maintenance of gypsum waste present economic as well as poten-
tial environmental concerns. The main environmental concerns
associated with waste gypsum dumps and disposal sites are the
⇑ Corresponding authors at: DST/CSIR National Centre for Nanostructured Mate-
rials, Council for Scientific and Industrial Research, PO Box 395, Pretoria 0001, South
Africa (M. de Beer), Council for Geoscience, Private Bag X112, Pretoria 0001, South
Africa (F.J. Doucet).
E-mail addresses: mdebeer@csir.co.za (M. de Beer), fdoucet@geoscience.org.za
(F.J. Doucet).
leaching of saline water into surface and underground water and
the generation of airborne dust (Mbhele et al., 2009). Not only
are these gypsum stacks unsightly, but they also occupy large areas
of land and require long-term expenditures for maintenance and
monitoring (Tayibi et al., 2009). In the context of reducing the envi-
ronmental burden and enhancing economic benefit, technologies
for converting waste materials into products of commercial value
are in great demand. To achieve a sustainable and environment-
friendly society, two potential approaches for the remediation of
gypsum waste are being investigated: material recycling and
chemical recycling. Research with regards to material recycling
has mostly been directed toward the utilization of gypsum waste
in the conditioning of alkaline soil (Gupta and Abrol, 1991) and
the preparation of hardened gypsum products (Singh and Garg,
2000).
In line with chemical recycling, the Council for Scientific and
Industrial Research (CSIR, Pretoria, SA) has developed a novel tech-
nological process for the conversion of gypsum waste into high-
value elemental sulphur, while a low-grade CaCO3 is generated
as a by-product (De Beer et al., 2014; Mbhele et al., 2009;
Nengovhela et al., 2007; Maree, 2005). Sulphur is a key raw mate-
rial for many manufacturing industries such as fertilizer, acids,

http://dx.doi.org/10.1016/j.wasman.2015.08.023
0956-053X/Ó 2015 Published by Elsevier Ltd.

Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
2 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx
steel, petroleum, insecticides, and explosives (Jin et al., 2010). The
CSIR technology involves three steps, namely, the thermal reduc-
tion of calcium sulphate to calcium sulphide (Eq. (1)) at 1050 °C,
the production of hydrogen sulphide from the calcium sulphide
(Eq. (2)) at ambient conditions, and subsequently, the conversion
of hydrogen sulphide to elemental sulphur (Eq. (3)) via the com-
mercially available chemical catalytic Clauss process (Mark et al.,
1978).
Thermal reduction of solid gypsum to produce a solid calcine
product:
CaSO4
2H2 O ðsÞ þ 2C ðsÞ ! CaS ðsÞ þ 2CO2 ðgÞ þ 2H2 O ðlÞ
Carbonation of the calcine product to produce hydrogen sul-
phide and calcium carbonate:
CaS ðsÞ þ H2 O ðlÞ þ CO2 ðgÞ ! H2 S ðgÞ þ CaCO3 ðsÞ
Recovery of elemental sulphur from the hydrogen sulphide gas:
2H2 S ðgÞ þ O2 ðgÞ ! 2S ðsÞ þ 2H2 O ðlÞ
Although the overall process successfully produced high-purity
elemental sulphur (Nengovhela et al., 2007), the study by de Beer
et al. (2014) showed that the direct carbonation of calcium sul-
phide (Eq. (2)) only yielded low-purity CaCO3 products (i.e.
86 mass% as CaCO3). The insight gained on the solution chemistry
and the characteristics of the carbonate products from the direct
aqueous CaS carbonation process served as a basis for the develop-
ment of a two-step or indirect carbonation process route for the
production of high-purity CaCO3 (i.e. >99 mass% as CaCO3). An out-
line of the indirect CaS carbonation process step is schematically
illustrated in Fig. 1 (the highlighted area represents the focus of
this study).
The feasibility of the gypsum technology can be enhanced if car-
bonate products of higher value can be recovered. In general,
CaCO3 is a versatile material with a variety of uses. However, the
application and uses of the low-purity, low-value CaCO3 is limited
to acid mine water neutralisation, cement manufacturing and met-
allurgical applications such as fluxing agent in steel making. High-
purity CaCO3 or precipitated calcium carbonate (PCC) has long
been recognized as a versatile additive and its importance is well
known due to its large range of industrial applications: e.g. as poly-
mer composite; filler for rubber and ceramics; additives for plas-
tics, paper, ink, cosmetics, toothpastes, detergents, a component
of pharmaceuticals and foodstuffs as well as numerous uses in
electronics and catalysis (Donate-Robles and Martín-Martínez,
2011; Shen et al., 2010, 2008; Gao et al., 2009; Gorna et al.,
Waste
Coal
gypsum
Thermal reducon CaS dissoluon
CaSO4.2H2O + 2C → CaS CaS solids CaS + H2O + H2S → H2O
+ 2CO2 + H2O + Ca(HS)2
Fig. 1. Process flow diagram for the two-step, indirect CaS carbonation process using H2S gas for CaS dissolution and CO2 gas for CaCO3 precipitation.
Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
2008; Lazzeri et al., 2005; Li et al., 2002). The use of high-purity
PCC can also benefit flue gas desulphurization technologies.
Valuable, high-purity gypsum will precipitate from the reaction
of pure CaCO3 with SO2, which is required for the production of
gypsum boards. Because of the higher market price of pure
CaCO3, the gypsum waste treatment process with the production
of high-purity CaCO3 could be more economically feasible than
the simple one-step or direct aqueous CaS carbonation step.
Simultaneously, major benefits in terms of waste reduction are
realised.
This study focused on the synthesis and characterisation of
ð1Þ high-purity CaCO3 from gypsum waste in a laboratory-scale exper-
imental set-up.
2. Experimental section
ð2Þ
2.1. Feedstock
ð3Þ A calcium sulphide-rich calcine powder (hereafter denominated
as CaSr) was produced from gypsum waste generated at an acid
mine water neutralisation plant in South Africa. Thermal reduction
of gypsum waste to CaSr using duff coal as the reducing agent was
performed in a pilot-scale electrically-heated kiln using an existing
process (Ruto et al., 2011). At completion of the reduction process,
CaSr was stored in a nitrogen environment to avoid spontaneous
oxidation to CaSO4 (García-Calzada et al., 2000). The procedures
used for the thermal reduction of gypsum waste and the character-
isation (i.e. XRD and XRF) of gypsum waste and formed CaSr were
described elsewhere (de Beer et al., 2014).
Gaseous hydrogen sulphide (100% H2S; Pure Gas, South Africa)
was used in all experiments to induce the dissolution of CaSr.
Gaseous carbon dioxide (100% CO2; Air Liquide, South Africa) was
used to strip H2S gas from solution and to stimulate the carbona-
tion of solubilized calcium derived from CaSr dissolution to form
calcium carbonate.
2.2. Experimental procedures
All experiments were performed in a 3-L (for CaSr dissolution)
and 1-L (for carbonation of Ca-rich leachate solution) Perspex stir-
red tank, batch reactors. The two reactors were equipped with four
equally spaced baffles and a sparger with small diameter (<1 mm)
openings for the introduction of H2S or CO2 gas. A mechanical over-
head stirrer (RW 20 digital from IKAÒ-Werke GmbH & Co. KG,
Germany) and a Rushton turbine impeller (manufactured by
Sulphur recovery
2H2S + O2 → 2S + 2H2O Elemental Sulphur
Aqueous stream
H 2S
H2S stripping /
Separaon Calcium carbonaon Separaon
Ca(HS)2 + CO2 + H2O →
CaCO3 ↓ + 2H2S ↑
CO2
Residue High-grade CaCO3
 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx
Manten Engineering, South Africa) were used for mixing. The
Rushton turbine impeller had six vertical blades fixed onto a disk.
A rotameter (Fisher & Porter, different models for H2S and CO2 gas)
was used to control the gas flow-rate. All experiments were con-
ducted at atmospheric pressure and room temperature.
Calcine (CaSr) was dispersed in distilled water (CaSr/water ratio
of 5%) to obtain a pre-determined slurry concentration. After
30 min of continuous mixing at 300 min1, H2S gas was introduced
at a constant flow-rate into the slurry to induce CaSr dissolution.
The pH, electrical conductivity and temperature of the suspension
were logged at 5 s intervals using a Hanna HI 2829 multi-
parameter logger (Hanna Instruments, South Africa) to collect
time-series data on the dynamics of the system. The reaction was
terminated by shutting off the H2S gas stream when the pH and
electrical conductivity remained unchanged for 15–20 min. Upon
completion, the suspension was removed from the reactor by vac-
uum filtration using Whatman grade no. 1 filter paper. The filter
cake was thoroughly rinsed with water and dried at mild temper-
ature (i.e. 60 °C; to prevent thermal conversion of gypsum
(Ballirano and Melis, 2009)) for 24 h. The mineralogical composi-
tion of the undissolved residues was analysed using XRD and their
particle size distribution was determined using the laser diffrac-
tion method (Horiba LA-950). The filtrate (Ca2+-rich Ca(HS)2 solu-
tion) was transferred into the 1-L Perspex reactor for further
reaction with CO2.
After 10 min of continuous mixing of the filtrates (Ca(HS)2 solu-
tions containing 450 mmol/L, 900 mmol/L and 1800 mmol/L Ca2+
respectively) from the CaS dissolution step in the 1-L reactor,
CO2 gas was introduced at a constant flow-rate via a sparger in
order to precipitate high-grade CaCO3. The pH, electrical conduc-
tivity and temperature of the suspension in the reactor were
recorded as described earlier. Samples of the Ca-rich solutions sub-
jected to carbonation were also collected from the reactor at regu-
lar intervals and were filtered using 0.45 lm PALL acrodisc PSF
GxF/GHP membranes (Microsep (Pty) Ltd, South Africa). The fil-
trates (undiluted, acidified samples) were analysed for their cal-
cium contents by inductively coupled plasma mass spectrometry
(ICP-MS) at an accredited laboratory (Consulting and Analytical
Services, CSIR, Pretoria, South Africa). The reaction was terminated
when the pH and electrical conductivity remained unchanged for
10–15 min, indicating the completion of the reaction.
Immediately upon completion of the experimental run, the final
suspension was removed from the reactor by vacuum filtration
using 0.45 lm Millipore HA membranes (Microsep (Pty) Ltd,
South Africa) and washed twice with distilled water. The filter
cakes were dried at 60 °C for 24 h.
2.3. Characterisation of solid materials
The mineralogical composition of the gypsum waste, calcine
(CaSr) and the carbonate products were analysed by XRD
(PANalytical X’Pert Pro powder diffractometer). The phases were
identified using the X’Pert Highscore plus software. Relative phase
amounts (wt%) were estimated using the Rietveld method.
High-vacuum scanning electron micrographs were collected
using a JEOL JSM7500 microscope to obtain information on mor-
phologies and size distribution of all the solid materials. Samples
were dispersed on carbon tape and sputter-coated with a thin, con-
ductive layer of gold using a Emitech K950X sputter coater.
Thermogravimetric analysis of the dried carbonate powders
was performed using the TGA Q500 (TA Instruments) in a nitrogen
atmosphere. Samples were heated from room temperature to 900 °
C at a heating rate of 10 °C/min.
Infrared spectra were recorded at room temperature with a
Perkin–Elmer Spectrum 100 Fourier Transform Infrared
Spectrometer (FTIR) equipped with a Perkin–Elmer Precisely
Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
3
Universal Attenuated Total Reflectance (ATR) sampling accessory
and a diamond crystal. The spectral study was extended over the
range 400–4000 cm1, with a resolution of 4 cm1 and 32 accumu-
lations. The spectra were analysed using the Spectrum 100 soft-
ware (Perkin Elmer).
The specific surface area of the products was measured by the
Brunauer–Emmett–Teller (BET) method of N2 adsorption–desorp-
tion at 196 °C with an ASAP 2010 surface area analyzer
(Micrometrics BET, USA). Prior to analysis, the samples were pre-
conditioned at 120 °C in vacuum for 4 h to remove the humidity
and contaminants adsorbed to the particle surfaces.
A helium gas displacement method with a pycnometer
(AccuPyc II 1340, Micromeritics, USA) was used to measure the
true density (qs) of the carbonate products.
3. Results and discussion
3.1. Characterisation of gypsum waste and calcine material
The untreated gypsum waste used in this study was charac-
terised and described elsewhere (de Beer et al., 2014). In brief, it
featured a mean diameter D (m, 0.5) of 43.9 lm with 10% of the par-
ticles smaller than 4.0 lm and 90% of the particles smaller than
176.7 lm. Its bulk mineralogical composition consisted of gypsum
(CaSO4
2H2O; 96.9%), which co-existed with magnesite (MgCO3;
3.1%). In contrast to our previous study (de Beer et al., 2014) where
gypsum waste had only been partly converted to a calcium
sulphide-rich calcine powder (CaSr) containing 50.7% of the min-
eral oldhamite (i.e. CaS), this study benefited from an optimised
thermal reduction process. The latter generated a CaSr sample of
higher purity (83.0% oldhamite). Less abundant mineral phases
included hydroxyapatite (Ca5(PO4)3(OH); 8.2%), quartz (SiO2;
4.3%), anhydrite (CaSO4; 4.2%), and lime (CaO; 0.3%). The level of
purity of the CaSr sample was further confirmed by a wet analytical
method (Procedure 4500-S2/Iodometric method described in
Standard Methods; Clesceri et al., 1989), which reported a CaS con-
tent of 81.7 ± 0.8 mass%.
3.2. Aqueous dissolution of CaSr
As conveyed in our previous paper (de Beer et al., 2014), CaS is a
sparingly soluble salt in water, as indicated by its low solubility
ranging from 0.125 g/L to 1.0 g/L (Dean, 1992; Perry and Green,
1984; Weast, 1972). However, the extent of its dissolution can be
substantially improved by sparging H2S gas to aqueous CaS sus-
pensions. In order to confirm this effect, CaSr was initially sus-
pended in water (without the addition of H2S gas) and
continuously mixed for 30 min. The solution pH rapidly increased
and stabilized at values above 10 within the first 2 min of stirring.
This was accompanied by an increase in solution conductivity
(1.7 mS/cm). This confirmed our previous observations made with
CaSr of lower purity (i.e. 50.7% CaS; de Beer et al., 2014). In addi-
tion, the high sulphur content (395 mmol/L as S) of CaSr dispersed
in water and the low amount (4 mmol/L as S) of dissolved sulphur
measured in solution after the 30 min of mixing in the absence of
H2S gas confirmed the low solubility of CaS in water.
H2S gas was then bubbled through CaSr suspensions in order to
promote the aqueous dissolution of CaS particles. It was antici-
pated that H2S gas would first dissolve in water to produce aque-
ous H2S, which subsequently would react with CaS to form
water-soluble Ca(HS)2 (248.7 g/L; Seidell and Linke, 1940). In the
presence of H2S, CaS dissolution would therefore proceed accord-
ing to Eq. (4).
CaS ðsÞ þ H2 O ðlÞ þ H2 S ðaqÞ ! CaðHSÞ2 ðaqÞ þ H2 O ðlÞ
DH25 C ¼ 44:8 kJ ð4Þ
4 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx

Following the addition of H2S gas (time = 0 min) to the aqueous the total amount of calcium added as CaS (575 mmol/L or 69.1 g
CaS suspension, immediate temporal changes (i.e. increase/ as Ca), 91.7% dissolved upon H2S addition to form a Ca(HS)2 solu-
decrease) in solution pH, electrical conductivity, dissolved calcium tion containing 530 mmol/L (or 63.7 g) as Ca 45 mmol/L (or 5.4 g
concentration and temperature were observed. The conductivity, as Ca) of the added Ca was lost to the residue.
temperature and dissolved calcium exhibited similarly-shaped The dissolution of CaS using H2S gas generated a calcium-rich
profiles, with all three parameters having reached their maxima solution (to be subsequently carbonated), and a dark-grey solid
after about 20 min of reaction (Fig. 2). The solution conductivity residue (waste). Approximately 0.38 kg of this residue was gener-
(Fig. 2a), temperature (Fig. 2b) and dissolved calcium (Fig. 2c) ated for every 1.0 kg of CaSr, or for every 2 kg of waste gypsum pro-
increased over the first 20 min before levelling off and stabilizing cessed. This represented a reduction of ca. 80% in mass of waste to
at about 54 mS/cm, 28 °C and 530 mmol/L respectively. The forma- be discarded, compared to the starting gypsum waste mass. The
tion of the plateau was much more rapid (ca. 20 min; Fig. 2) than residue was characterised by a mean diameter D (m, 0.5) of
that observed previously for direct carbonation in the presence of 14.3 lm with 10% of the particles smaller than 6.34 lm and 90%
CO2 and the absence of H2S (200–210 min; de Beer et al., 2014). of the particles smaller than 64.39 lm. The primary mineral phases
The change in temperature from 21.7 to 28.3 °C (DT = 6.6 °C) con- identified in the residue were anhydrite (CaSO4; 35.7%), apatite
firmed the exothermic nature of CaS dissolution. The increase in (Ca5(PO4)3OH; 27.9%), quartz (SiO2; 13.6%) and two carbonate
conductivity could be ascribed to the absorption of H2S gas and polymorphs, calcite and vaterite (CaCO3; 12.3%). Less abundant
the subsequent formation of aqueous H2S, as well as to the disso- mineral phases included fluorite (CaF2; 5.7%) and bassanite
lution of CaS present in CaSr. The gradual increase in conductivity (CaSO4
½H2O; 4.4%). Apart from the carbonate phases, the other
values was therefore a first indirect indication of the formation of primary minerals were present in the starting CaSr powder and
Ca(HS)2 in solution. Over the pH range 7.6–11.8, HS is the domi- most probably did not react during the H2S-facilitated dissolution
nant sulphide species (Suleimenov and Krupp, 1994), which step.
implies that during the dissolution stage, little, if any S is stripped
from the reactor system. During the same period, the solution pH
3.3. Carbonation of calcium-rich solutions
(Fig. 2d) initially dropped sharply from 11.9 to 9.6 within the first
minute, after which it continued decreasing more gradually down
Experiments of aqueous dissolution of CaSr were repeated using
to ca. 7.6 at 20 min, levelled off and remained constant from 20 to
distinct CaSr/water ratios of 4%, 8% and 16% (or 97.2 g, 194.4 g and
40 min of reaction time. The assumption was made that complete
388.8 g CaSr in 3 L of water) in order to produce Ca(HS)2 solutions
dissolution of CaS particles would be indicated by the levelling-off
containing 450 mmol/L, 900 mmol/L and 1800 mmol/L Ca2+
of all the salient parameters; this assumption was supported by the
respectively. At completion of these dissolution experiments, the
levelling off of dissolved calcium (Fig. 2c) and the observed reduc-
aqueous Ca2+-rich filtrates were subjected to a carbonation reac-
tion in solution temperature after 20 min of reaction (Fig. 2b). Of
tion using flow rates of 1.05 (for 4% and 8% CaSr/water ratios)

60 30
(a) (b)
50
Conductivity (mS/cm)

28
Temperature (°C)

40
26
30
24
20

22
10

0 20
0 10 20 30 40 0 10 20 30 40
Time (min) Time (min)

600 100
(c) 12 (d)
Dissolved calcium (mmol/L)

500 80
Dissolved calcium (%)

% dissolved calcium 11
400
60
10
300
pH

40
9
200

20 8
100

0 0 7
0 10 20 30 40 0 10 20 30 40
Time (min) Time (min)

Fig. 2. Solution conductivity (a), temperature (b), dissolved calcium (mmol/L on Y1 and % on Y2 axis) (c) and pH (d) profiles of the calcine slurry in equilibrium with distilled
water upon H2S addition (CaS slurry containing 575 mmol/L as S; initial pH: 11.9; gas flow: 8.4 L/min/kg calcine; stirring rate: 700 min1; a time of 0 min on the x-axis
corresponds to the start of the addition of H2S to solution).

Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx
and 1.75 (for 16% CaSr/water ratio) L/min/mol Ca2+ to ensure that
CO2 was present in excess in relation to the concentration of dis-
solved Ca2+.
Upon introduction of CO2 gas, CaCO3 was expected to precipi-
tate from solution according to Eq. (5):
CaðHSÞ2 ðaqÞ þ CO2 ðgÞ þ H2 O ðlÞ
! CaCO3 ðsÞ þ 2H2 S ðgÞ DH25 C ¼ 7:0 kJ
3.3.1. Analysis of CaCO3 polymorphism
The phase identification of the carbonate products was carried
out by XRD. XRD patterns of the samples produced at increasing
initial dissolved Ca2+ concentrations are shown in Fig. 3.
Diffraction patterns of the products exhibited a significant peak
at 29.3° (2h), which is characteristic of crystalline calcite (Chen
et al., 2012). In comparison with the standard JCPDS file (calcite:
83-0586), the diffraction peaks were indexed as rhombohedral cal-
cite. Based on the Scherrer equation, the crystallite size is reversely
proportional to the width of the peak at half maximum of the
diffraction line on the 2h scale (Langford and Wilson, 1978).
According to the equation on the 104 peak located at 2h = 29.3°,
the average grain size of the calcite crystallite of all the products
were identical at 834 Å or 83.47 nm. The sharp and strong peaks
further confirmed that the products were well-crystallized
(Fig. 3a–c). The bulk mineral composition, as determined by
Rietveld refinement of XRD diffraction data, was primarily calcite
(99.5%) with traces of quartz (0.5%). According to XRD, calcite
was therefore the only CaCO3 polymorph obtained at the end of
each reaction; no vaterite or aragonite was detected under these
experimental conditions.
The formation of a single polymorph of CaCO3 was further con-
firmed by FT-IR analysis (Fig. 4). Infrared spectroscopy detects the
vibration characteristics of chemical functional groups in samples.
When an infrared light interacts with the sample, chemical bonds
will stretch, contract and bend. As a result, a chemical functional
group tends to adsorb infrared radiation in a specific wavenumber
range regardless of the structure of the rest of the molecule. FT-IR
is therefore a useful tool for the identification of calcium carbonate
polymorphs due to the differences in their carbonate ions, CO2 3 900 mmol/L and 1800 mmol/L resulted in actual mass loss of
(Islam et al., 2011). The spectral data obtained for the samples
revealed the characteristic transmittance peaks of calcite, which
8000 (a)
6000
4000
2000
0
10
Intensity (counts)
8000 (b)
6000
4000
2000
0
10
8000 (c)
6000
4000
2000
0
10
Fig. 3. XRD diffraction patterns of high-grade CaCO3 produced at initial Ca2+ concentrations of (a) 450 mmol/L, (b) 900 mmol/L and (c) 1800 mmol/L.
Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
5
are the in-plane (t4 mode) and the out-of plane (t2 mode) vibra-
tional bands at 713 cm1 and 877 cm1, respectively, and the
anti-symmetry stretch (t3 mode) at around 1400 cm1 and sym-
metric stretching (t1 mode) at around 1800 cm1 (Xu and
Poduska, 2014; Menahem and Mastai, 2008). The IR spectra con-
firmed the presence of the single calcite phase by the identification
of the characteristic t4 band of calcite at 713 cm1 and the absence
ð5Þ of the characteristic t4 bands of vaterite at 745 cm1 and aragonite
at 700 cm1.
3.3.2. Purity analysis
The thermogravimetric analysis curves (Fig. 5) illustrate the
weight loss of the carbonate products during thermal decomposi-
tion over the temperature range 25–900 °C. All three products
were stable below 100 °C. The first step was observed at about
100 °C; it was characterised by a small weight loss (<0.6%), which
was attributed to loss of moisture from the samples. A second step
was identified in the temperature range 600 to 730–760 °C; this
weight loss was rapid, pronounced, and corresponded to the
decomposition of carbonate products with the accompanying
release of CO2 (Wang et al., 1999). This weight loss equated to
approximately 43.6% of the total mass. The sample weights
remained unchanged upon heating above 760 °C. Calcium carbon-
ate thermally decomposes according to the following reaction:
CaCO3 ðsÞ ! CaO ðsÞ þ CO2 ðgÞ ðloss of carbon dioxideÞ
1
DH25 C ¼ 182:1 kJ mol ð6Þ
Additional TGA experiments were performed on these three
samples to obtain additional quantification between the weight
fractions. For this purpose, during heating, the samples were kept
isothermally at 105 °C, 400 °C and 850 °C for 15 min. The weight
fraction (Dm400850°C) based on dry weight (m105°C) was used to
calculate the calcium carbonate content, which was expressed as
wt% of CO2, using equation Eq. (7).
CO2 ðwt%Þ ¼ Dm400—850 C =m105 C  100 ð7Þ
Analysis of the thermal decomposition of the carbonate prod-
ucts synthesised at initial Ca2+ concentrations of 450 mmol/L,
43.5%, 43.7% and 43.6%, respectively, which is very similar to the
theoretical mass loss expected for 100% pure CaCO3 (i.e. 43.9%;
20 30 40 50 60 70
20 30 40 50 60 70
20 30 40 50 60 70
2 Theta (°)
6 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx

100
90 (a)
80
70
60
2000 1800 1600 1400 1200 1000 800 600
Transmittance (%)
100
90 (b)
80
70
60
2000 1800 1600 1400 1200 1000 800 600
100
90 (c)
80
70
60
2000 1800 1600 1400 1200 1000 800 600
-1
Wavenumber (cm )

Fig. 4. FT-IR transmission spectra of high-grade CaCO3 produced at initial Ca2+ concentrations of (a) 450 mmol/L, (b) 900 mmol/L and (c) 1800 mmol/L.

100 Increasing the initial Ca2+ concentrations from 450 to

(a) Adsorbed water: 0.39% 1800 mmol/L Ca2+ in the filtrates had therefore little effect on the
90 mineral composition, purity and percentage yield of the formed
Mass loss:
80 43.54% carbonate products (Table 1). The nearly identical CaCO3 yield
obtained confirmed that CO2 had been present in excess in relation
70
to dissolved Ca2+ concentration.
60 759 °C The product generated from a 1800 mmol/L (as Ca) solution
50 decomposed slightly more rapid (Fig. 5c; 600–730 °C) than the
100 200 300 400 500 600 700 800 900 products obtained from solutions containing 900 or 450 mmol/L
Weight (% of initial mass)

100 (as Ca) (Fig. 5a and b; 600–760 °C). The slight difference could
(b) Adsorbed water: 0.54% not be explained by the polymorph composition, crystallite size
90
Mass loss:
and purity of the products. It was also required to characterise
80 43.71% the products in terms of their physical properties.
70

60 756 °C 3.3.3. Structural and morphological analysis

The effect of the initial dissolved Ca2+ concentration in solution
50
100 200 300 400 500 600 700 800 900 on the surface characteristics and morphology of CaCO3 crystals
was also determined. SEM observations indicated precipitates of
100
micron-sized, interpenetrated rhombohedral cubes of calcite
90
(c) Adsorbed water: 0.57%
(Fig. 6), with a notable decrease in particle size with increasing ini-
80 Mass loss: tial Ca2+ concentration. A laser diffraction technique was used to
43.62% compare the particle size characteristics of the precipitated solids.
70
Fig. 7a shows the cumulative volume fractions and Fig. 7b the par-
60 732 °C ticle size frequency distributions. Data in Table 2 show the mea-
50 sured true density, surface area and particle sizes of the
100 200 300 400 500 600 700 800 900 products. With increasing initial Ca2+ concentration, the specific
Temperature (°C) surface area of the precipitates increased as the mean particle size
diameter became smaller. The particles of smallest size
Fig. 5. TGA curves for high-grade CaCO3 produced at initial Ca2+ concentrations of
(a) 450 mmol/L, (b) 900 mmol/L and (c) 1800 mmol/L.
Table 1
Quantitative XRD analysis and yields of high-grade CaCO3 produced at different initial
Frost et al., 2009). The purity of CaCO3 in the products was calcu- Ca2+-concentrations during the carbonation step.

lated using Eq. (8), where MW denotes molecular weights and Initial Ca2+ concentration mmol/L 450 900 1800
CO2 (wt%) represents the weight loss of the sample. Mineral composition
Calcite mass% 99.5 99.5 99.4
PCaCO3 ð%Þ ¼ CO2 ðwt%Þ  MWCaCO3 =MWCO2 ð8Þ Quartz mass% 0.5 0.5 0.6
The purity of the CaCO3 in the formed products was consistently CaCO3 yield
higher than 99.0% and confirmed the high purity of the synthesized Actual mass g/100 g CaSr 72.2 ± 0.7 71.8 ± 1.0 70.2 ± 0.2
Yield % 91.6 ± 0.8 90.8 ± 1.2 88.8 ± 0.2
carbonate products suggested by XRD (Table 1).

Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx 7

Fig. 6. SEM images (top row 1000 and bottom row 5000 magnification) of high-grade CaCO3 produced at initial Ca2+ concentrations of (a) 450 mmol/L, (b) 900 mmol/L and
(c) 1800 mmol/L.

(a) These findings were in agreement with the theoretical models of

Cumulative volume

100
precipitation by which an increase in initial Ca2+ concentration
fraction (%)

80 (i) 450 mmol/L gives rise to nucleation of a larger number of smaller crystals
(ii) 900 mmol/L
60 (iii) 1800 mmol/L (Mumtaz and Hounslow, 2000).
40
20
iii ii i 3.3.4. Solution chemistry
Fig. 8 illustrates temporal changes in solution pH and dissolved
0
0.1 1 10 100 1000 Ca2+ concentration during the carbonation of the filtrates charac-
terised by varying initial dissolved Ca2+ concentrations. It must
Diameter (µm)
be noted that the pH at the end of the dissolution stage was around
(% particles by volume)

25 7.5 while H2S gas was introduced. As soon as the gas was shut off
(b) and the solid precipitate removed, the pH increased rapidly to
20 ii
about 10.5, presumably due to the volatilisation of excess H2S.
Frequency

15 (i) 450 mmol/L When adding CO2 (time = 0 min), the pH (Fig. 8, bottom row)
(ii) 900 mmol/L
10 (iii) 1800 mmol/L dropped sharply from >10.5 to approximately 8.5 within a very
iii
i short period of time (<2 min), after which it continued decreasing
5
more gradually before stabilizing at approximately 6.5 at the end
0 of the CaCO3 precipitation reaction.
0.1 1 10 100 1000
Plateauing of the dissolved calcium concentration indicated the
Diameter (µm)
end of the CaCO3 formation. The descending parts of dissolved Ca
Fig. 7. Particle size distribution curves of cumulative volume fraction (a) and concentration vs. time curves correspond to the consumption of
frequency (b) for high-grade CaCO3 produced at various initial Ca2+ concentrations. dissolved Ca and the concurrent formation of CaCO3. It was found
that all three reactions had been completed within a similar
(Figs. 6c, 7aiii, biii and 8) decomposed faster (Fig. 5c) than larger amount of time (50–55 min). For this reason and because the initial
particles (Figs. 6a and b, 7ai and bi, 8 and 5a and b), most probably dissolved Ca concentrations had been dissimilar, the slope of the
because smaller-sized particles have a greater surface area per unit lines did not represent the rate of CaCO3 precipitation, but was
mass than larger particles (Table 2). The products generated at dif- rather representative of the differences in the amount of Ca being
ferent initial dissolved Ca2+ concentrations exhibited densities of converted to carbonates. Fig. 8 depicts calculated values obtained
2.7 g/cm3, which is characteristic of calcite (Plummer and for the slopes of the calcium vs time profiles. The higher values
Busenberg, 1982). obtained for starting filtrates containing 1800 mmol/L Ca com-
The degree of aggregation of the precipitated particles also cor- pared to those containing less Ca indicated that much more Ca
related with the increase in initial Ca2+ concentration, as shown had been consumed during the carbonation stage of the former
from the SEM images of the corresponding samples (Fig. 6). solutions.

Table 2
Physical properties of high-grade CaCO3 produced at increasing levels of Ca2+ concentrations in the starting filtrates.

Initial [Ca2+] True density (g/cm3) BET surface area (m2/g) Particle size distribution
D (v, 0.1) (lm) D (v, 0.5) (lm) D (v, 0.9) (lm)
(a) 450 mmol/L 2.710 ± 0.003 0.30 ± 0.01 27.73 44.32 82.30
(b) 900 mmol/L 2.710 ± 0.005 0.40 ± 0.02 23.31 32.36 46.15
(c) 1800 mmol/L 2.720 ± 0.007 0.79 ± 0.03 13.16 24.65 42.10

Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
8 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx

2000 (a) 2000 (b) 2000 (c)

-36.2 ± 0.1 mmol/L/min

Calcium (mmol/L)
1500 1500 1500 2
(R = 0.9977)

1000 1000 -15.7 ± 0.1 mmol/L/min 1000

2
(R = 0.9988)
-9.5 ± 0.1 mmol/L/min
2
(R = 0.9993)
500 500 500

0 0 0
0 20 40 60 80 0 20 40 60 80 0 20 40 60 80
Time (min) Time (min) Time (min)

11 11 11
(d) (e) (f)
10 10 10

9 9 9
pH

8 8 8

7 7 7

6 6 6

0 20 40 60 80 0 20 40 60 80 0 20 40 60 80
Time (min) Time (min) Time (min)

Fig. 8. Dissolved calcium (a) and pH (d) profiles of the carbonation reaction at the initial Ca2+ of 450 mmol/L, initial pH 10.8 and gas flow rate of 1.05 L/min/mol Ca2+.
Dissolved calcium (b) and pH (e) profiles of the carbonation reaction at the initial Ca2+ of 900 mmol/L, initial pH 11.5 and gas flow rate of 1.05 L/min/mol Ca2+. Dissolved
calcium (c) and pH (f) profiles of the carbonation reaction at the initial Ca2+ of 1800 mmol/L, initial pH 10.5 and gas flow rate of 1.75 L/min/mol Ca2+.

4. Conclusions References

The potential for the synthesis of high-purity CaCO3 from indus-
trial gypsum waste was demonstrated in this study. High-grade
CaCO3 (i.e. >99.5 mass% as CaCO3) or PCC was obtained during
the indirect aqueous carbonation of CaS, an intermediate product
in the process of the recovery of elemental sulphur from gypsum
waste. We are now optimising the process further to generate
sub-micron, high-purity CaCO3 particles. The effect of initial Ca2+
concentration in solution, which is a function of the solid–liquid
ratio (CaSr/water ratio during the dissolution step) on the formed
carbonate products, was investigated. The effect on the final car-
bonate products in terms of mineral composition, purity, density
and percentage yield were similar. The results on the mineralogical
analyses also showed that calcite was the only CaCO3 polymorph
obtained. Rhombohedral structured calcite can find application in
for instance the paper and pulp industry where sheet bulk, porosity
or some other property is of paramount importance. This process
offers the significant advantage of producing two commercial-
grade products (elemental sulphur, and PCC) from gypsum waste.
Acknowledgments
The authors thank the following organisations for financial sup-
port and/or technical/scientific input: THRIP (Technology and
Human Resource for Industry Programme of the National
Research Foundation (NRF)), Tshwane University of Technology
(TUT), Northwest University (NWU), the Council for Scientific and
Industrial Research (CSIR) and the Council for Geoscience (CGS).
The authors are grateful to the DST/CSIR National Centre for
Nanostructured Materials characterisation facility staff, specifically
K Selatile and Z Ndaba for their help with analyses.
Ballirano, P., Melis, E., 2009. Thermal behaviour and kinetics of dehydration of
gypsum in air from in situ real-time laboratory parallel-beam X-ray powder
diffraction. Phys. Chem. Miner. 36 (7), 391–402.
Chen, Z., Yang, B., Nan, Z., 2012. Modification of calcium carbonate crystals growth
by dibenzoic acids. Mater. Chem. Phys. 132 (2–3), 601–609.
Clesceri, L.S., Greenberg, A.E., Trussel, R.R. (Eds.), 1989. Standard Methods for the
Examination of Water and Wastewater, 17th ed. American Public Health
Association, Washington.
De Beer, M., Maree, J.P., Liebenberg, L., Doucet, F.J., 2014. Conversion of calcium
sulphide to calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Manage. 34, 2373–2381.
Dean, J.A. (Ed.), 1992. Lange’s Handbook of Chemistry, 14th ed. McGraw-Hill, New
York.
Donate-Robles, J., Martín-Martínez, J.M., 2011. Addition of precipitated calcium
carbonate filler to thermoplastic polyurethane adhesives. Int. J. Adhes. Adhes.
31 (8), 795–804.
Frost, R.L., Hales, M.C., Martens, W.N., 2009. Thermogravimetric analysis of selected
group (II) carbonate minerals – Implication for the geosequestration of
greenhouse gases. J. Therm. Anal. Calorim. 95 (3), 999–1005.
Gao, Y., Liu, L., Zhang, Z., 2009. Mechanical performance of nano-CaCO3 filled
polystyrene composites. Acta Mech. Solida Sin. 22 (6), 555–562.
García-Calzada, M., Marbán, G., Fuertes, A.B., 2000. Decomposition of CaS particles
at ambient conditions. Chem. Eng. Sci. 55 (9), 1661–1674.
Gorna, K., Hund, M., Vučak, M., Gröhn, F., Wegner, G., 2008. Amorphous calcium
carbonate in form of spherical nanosized particles and its application as fillers
for polymers. Mater. Sci. Eng. A 477 (1–2), 217–225.
Gupta, R.K., Abrol, I.P. (Eds.), 1991. Salt-affected soils: their Reclamation and
Management for Crop Production. Springer, New York.
Islam, K.N., Bakar, M.Z.B.A., Noordin, M.M., Hussein, M.Z.B., Rahman, N.S.B.A., Ali, M.
E., 2011. Characterisation of calcium carbonate and its polymorphs from cockle
shells (Anadara granosa). Powder Technol. 213 (1–3), 188–191.
Jin, S., Shuichang, Z., Guangyou, Z., Zhi, Y., Bin, Z., Anguo, F., Debin, Y., 2010.
Geological reserves of sulfur in China’s sour gas fields and the strategy of sulfur
markets. Petrol. Explor. Dev. 37 (3), 369–377.
Johnson, D.B., Hallberg, K.B., 2005. Acid mine drainage remediation options: a
review. Sci. Total Environ. 338 (1–2), 3–14.
Kalin, M., Fyson, A., Wheeler, W.N., 2006. The chemistry of conventional and
alternative treatment systems for the neutralization of acid mine drainage. Sci.
Total Environ. 366 (2–3), 395–408.

Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
 M. de Beer et al. / Waste Management xxx (2015) xxx–xxx
Langford, J.I., Wilson, A.J.C., 1978. Scherrer after sixty years: a survey and some new
results in the determination of crystallite size. J. Appl. Crystallogr. 11 (2), 102–
113.
Lazzeri, A., Zebarjad, S.M., Pracella, M., Cavalier, K., Rosa, R., 2005. Filler toughening
of plastics. Part 1 – The effect of surface interactions on physico-mechanical
properties and rheological behaviour of ultrafine CaCO3/HDPE nanocomposites.
Polymer 46 (3), 827–844.
Li, L., Collis, A., Pelton, R., 2002. A new analysis of filler effects on paper strength. J.
Pulp Pap. Sci. 28 (8), 267–273.
Maree, J.P., 2005. Conversion of Sulphur-Containing Waste Material into a Sulphur-
Containing Product. Australia: AU 2005201759 patent http://ip.com/patfam/en/
35455786.
Mark, H.F., Othmer, D.F., Overberger, C.G., Seaborg, G.T. (Eds.), 1978. Kirk-Othmer:
Encyclopedia of Chemical Technology (22, 276), third ed. Wiley-Interscience,
New York.
Mbhele, N.R., Van der Merwe, W., Maree, J.P., Theron, D., 2009. Recovery of Sulphur
from Waste Gypsum. In: Abstracts of the International Mine Water Conference,
19–23 October 2009, pp. 622–630.
Menahem, T., Mastai, Y., 2008. Controlled crystallization of calcium carbonate
superstructures in macroemulsions. J. Cryst. Growth 310 (15), 3552–3556.
Mumtaz, H.S., Hounslow, M.J., 2000. Aggregation during precipitation from
solution: an experimental investigation using Poiseuille flow. Chem. Eng. Sci.
55 (23), 5671–5681.
Nengovhela, N.R., Strydom, C.A., Maree, J.P., Oosthuizen, S., Theron, D.J., 2007.
Recovery of sulphur and calcium carbonate from waste gypsum. Water SA 33
(5), 741–747.
Perry, R.H., Green, D. (Eds.), 1984. Perry’s Chemical Engineers’ Handbook, sixth ed.
McGraw-Hill Book Company, New York.
Plummer, L.N., Busenberg, E., 1982. The solubilities of calcite, aragonite and vaterite
in CO2–H2O solutions between 0 and 90 °C, and an evaluation of the aqueous
model for the system CO3-CO2-H2O. Geochim. Cosmochim. Acta 46 (6), 1011–
1040.
Ruto, S., Maree, J.P., Zvinowanda, C.M., Louw, W.J., Kolesnikov, A.V., 2011. Thermal
studies on gypsum in a pilot-scale, rotary kiln. In: Water in the South African
Please cite this article in press as: de Beer, M., et al. Synthesis of high-purity precipitated calcium carbonate during the process of recovery of elemental
sulphur from gypsum waste. Waste Management (2015), http://dx.doi.org/10.1016/j.wasman.2015.08.023
9
Minerals Industry Conference, 15–17 February 2011, The Southern African
Institute of Mining and Metallurgy, The Southern African Institute of Mining
and Metallurgy.
Seidell, A., Linke, W.F. (Eds.), 1940. Solubilities of Inorganic and Metal Organic
Compounds; a Compilation of Quantitative Solubility Data from the Periodical
Literature. D. Van Nostrand Company Inc., New York.
Shen, J., Song, Z.Q., Qian, X.R., Song, C.J., 2008. Preparation of chitosan-coated
papermaking grade PCC filler and its application. Chung-kuo Tsao Chih/China
Pulp Paper 27 (1), 21–25.
Shen, J., Song, Z., Qian, X., Yang, F., 2010. Carboxymethyl cellulose/alum modified
precipitated calcium carbonate fillers: preparation and their use in
papermaking. Carbohydr. Polym. 81 (3), 545–553.
Singh, M., Garg, M., 2000. Making of anhydrite cement from waste gypsum. Cem.
Concr. Res. 30 (4), 571–577.
Smith, L.J.D., Moncur, M.C., Neuner, M., Gupton, M., Blowes, D.W., Smith, L., Sego, D.
C., 2013. The Diavik waste rock project: design, construction, and
instrumentation of field-scale experimental waste-rock piles. Appl. Geochem.
36, 187–199.
Suleimenov, O.M., Krupp, R.E., 1994. Solubility of hydrogen sulfide in pure water
and in NaCl solutions, from 20 to 320 °C and at saturation pressures. Geochim.
Cosmochim. Acta 58 (11), 2433–2444.
Tayibi, H., Choura, M., López, F.A., Alguacil, F.J., López-Delgado, A., 2009.
Environmental impact and management of phospho-gypsum. J. Environ.
Manage. 90 (8), 2377–2386.
Wang, L., Sondi, I., Matijevic, E., 1999. Preparation of uniform needle-like aragonite
particles by homogeneous precipitation. J. Colloid Interface Sci. 218 (2), 545–
553.
Weast, R.C. (Ed.), 1972. CRC Handbook of Chemistry and Physics, 53rd ed. The
Chemical Rubber Co., Cleveland, Ohio.
Xu, B., Poduska, K.M., 2014. Linking crystal structure with temperature-sensitive
vibrational modes in calcium carbonate minerals. Phys. Chem. Chem. Phys. 16
(33), 17634–17639.