Indirect mineral carbonation of titanium-bearing blast

Chinese Journal of Chemical Engineering 26 (2018) 583–592

Contents lists available at ScienceDirect

Chinese Journal of Chemical Engineering journal homepage: www.elsevier.com/locate/CJChE

Article

Indirect mineral carbonation of titanium-bearing blast furnace slag coupled with recovery of TiO2 and Al2O3☆ Lin Wang, Weizao Liu, Jingpeng Hu, Qiang Liu, Hairong Yue, Bin Liang, Guoquan Zhang, Dongmei Luo, Heping Xie, Chun Li ⁎ College of Chemical Engineering, Sichuan University, Chengdu 610065, China

a r t i c l e

i n f o

Article history: Received 27 March 2017 Received in revised form 19 June 2017 Accepted 19 June 2017 Available online 29 July 2017 Keywords: Mineral carbonation CO2 sequestration Ti-bearing blast furnace slag Ammonium sulphate Titanium dioxide Aluminium oxide

a b s t r a c t Large quantities of CO2 and blast furnace slag are discharged in the iron and steel industry. Mineral carbonation of blast furnace slag can offer substantial CO2 emission reduction and comprehensive utilization of the solid waste. This paper describes a novel route for indirect mineral carbonation of titanium-bearing blast furnace (TBBF) slag, in which the TBBF slag is roasted with recyclable (NH4)2SO4 (AS) at low temperatures and converted into the sulphates of various valuable metals, including calcium, magnesium, aluminium and titanium. High value added Ti- and Al-rich products can be obtained through stepwise precipitation of the leaching solution from the roasted slag. The NH3 produced during the roasting is used to capture CO2 from flue gases. The NH4HCO3 and (NH4)2CO3 thus obtained are used to carbonate the CaSO4-containing leaching residue and MgSO4-rich leaching solution, respectively. In this study, the process parameters and efficiency for the roasting, carbonation and Ti and Al recovery were investigated in detail. The results showed that the sulfation ratios of calcium, magnesium, titanium and aluminium reached 92.6%, 87% and 84.4%, respectively, after roasting at an AS-to-TBBF slag mass ratio of 2:1 and 350 °C for 2 h. The leaching solution was subjected to hydrolysis at 102 °C for 4 h with a Ti hydrolysis ratio of 95.7%and the purity of TiO2 in the calcined hydrolysate reached 98 wt%. 99.7% of aluminium in the Ti-depleted leaching solution was precipitated by using NH3. The carbonation products of Ca and Mg were CaCO3 and (NH4)2Mg(CO3)2·4H2O, respectively. The latter can be decomposed into MgCO3 at 100–200 °C with simultaneous recovery of the NH3 for reuse. In this process, approximately 82.1% of Ca and 84.2% of Mg in the TBBF slag were transformed into stable carbonates and the total CO2 sequestration capacity per ton of TBBF slag reached up to 239.7 kg. The TiO2 obtained can be used directly as an end product, while the Al-rich precipitate and the two carbonation products can act, respectively, as raw materials for electrolytic aluminium, cement and light magnesium carbonate production for the replacement of natural resources. © 2017 The Chemical Industry and Engineering Society of China, and Chemical Industry Press. All rights reserved.

1. Introduction Carbon dioxide capture (utilization) and storage (CCS/CCUS) is recognized as one of the options to tackle the increase in the atmospheric concentration of CO2 for climate change mitigation [1]. Although geological and oceanic sequestration of CO2 has a huge storage capacity, proper geological conditions are required, such as depleted oil and gas fields and geographical conditions, e.g., close to deep sea for the CO2 sources, as well as inevitable use of expensive monitoring equipment to ensure that CO2 does not escape into the atmosphere for a very long term [2,3]. Mineral carbonation is a potential CO2 sequestration method, in which calcium and magnesium oxides in various minerals,

☆ Supported by the National Key Projects for Fundamental Research and Development of China (2016YFB0600904). ⁎ Corresponding author. E-mail address: [email protected] (C. Li).

especially silicate minerals, react with CO2 and form thermodynamically stable carbonates. The silicate rocks are abundant in nature to the extent that in theory the potential for CO2 storage by mineral carbonation is larger than the potential for storage of CO2 by other methods [4]. Mineral carbonation consists of direct and indirect methods. The direct method mimics the weathering process in nature but proceeds at a much faster rate, enhanced by a variety of methods, including activation pretreatment of minerals [5], high temperature and pressure [6,7], and high gravity [8]. However, the products of direct mineral carbonation are generally a mixture, which is hard to separate for further utilization or selling as a valuable product. Indirect CO2 mineralization is ordinarily comprised of two successive steps: (1) extracting Ca and Mg from minerals with an acidic or weakly acidic additives and (2) carbonation of the Ca- and Mg-rich solutions or solids with CO2 in a basic or weakly basic environment. The method is now receiving widespread attention because of the relatively mild reaction conditions, high carbonation conversion of Ca and Mg, and purer, thus more valuable, by-products [9–11].

http://dx.doi.org/10.1016/j.cjche.2017.06.012 1004-9541/© 2017 The Chemical Industry and Engineering Society of China, and Chemical Industry Press. All rights reserved.

584

L. Wang et al. / Chinese Journal of Chemical Engineering 26 (2018) 583–592

Recently, the main raw materials for mineral carbonation are calcium- and magnesium-containing natural minerals [9,12,13] and industrial solid wastes [14–16]. Comparatively, the industrial solid wastes are easily available and cheap with high reactivity and usually appear where a large quantity of CO2 is emitted. Hence, there are many advantages for sequestration of CO2 with industrial solid wastes. Steelmaking slags are the main solid wastes in the iron and steel industry with total CaO + MgO content ≥40 wt%, including blast furnace slag and steel slag emitted in the iron refining and steelmaking process, respectively [17]. Currently, approximately 300–1000 kg blast furnace slag is discharged per ton of iron production, depending on the grade of the iron ores and process conditions employed [18]. Approximately 100–150 kg steel slag is produced per ton of steel production [19]. In 2015, the global output of steel was approximately 1.6 billion tons [20]. If all of the Ca and Mg elements in steelmaking slag worldwide were utilized to fix CO2, theoretically 200–680 million tons of CO2 could be safely sequestered annually. Although the amounts are small in comparison with the capacity of natural mineral resources, using steelmaking slags to store CO2 on site could provide an inexpensive way to reduce the CO2 emissions for individual iron and steel plants significantly. Ti-bearing blast furnace (TBBF) slag is a special kind of blast furnace slag, which is generated during the smelting of the vanadium titanomagnetite in a blast furnace for iron production. TBBF slag usually contains 10 wt%–25 wt% of TiO2 in addition to the conventional calcium, magnesium and aluminium silicates. Currently, approximately 20 million tons of the TBBF slag is discharged in China annually. The TBBF slag is difficult to use and thus must be stockpiled, resulting in serious environmental pollution as well as a waste of the precious titanium resources. Although numerous methods, including high temperature carbonization and low temperature chlorination [21], selective enrichment growth and separation process [22,23], acid processes [24,25], alkali processes [26,27] and iron-silicon-titanium alloy processes [28,29], have been proposed, those methods focus mainly on the recovery of titanium and pay almost no attention to the efficient utilization of the remaining components of the slag. Some of these methods may even produce secondary pollution. Recently, Xue et al. [30–32] proposed an ammonium sulphate roasting method for simultaneous recovery of Ti and Al, in which 91.7% of the titanium and 97% of the aluminium were extracted. However, large quantities of ammonium hydroxide and acidic leaching residue containing calcium sulphate and silicon dioxide were discharged and are difficult to utilize. Recently, we suggested a novel, facile pathway for simultaneous mineralization of CO2 with the Ca and Mg components in the TBBF slag and extraction of titanium and aluminium [33]. In the process, the TBBF slag was roasted with recyclable ammonium sulphate and

converted into sulphates of various metal elements in the slag. Highadded-value Ti- and Al-rich products can be achieved through stepwise precipitation of the leaching solution of roasted slag. The NH3 produced during the roasting was used to capture CO2 from flue gases. The solutions of NH4HCO3 and (NH4)2CO3 thus obtained are used, respectively, to carbonate the CaSO4-containing leaching residue and the MgSO4-rich leaching solution after the recovery of Ti and Al. The process flowsheet is shown in Fig. 1. Present study focuses on the process parameters and efficiency of the roasting, mineral carbonation and recovery of titanium and aluminium. 2. Experimental 2.1. Materials The TBBF slag used in this study was water-quenched slag, provided by Panzhihua Iron and Steel Group Company Limited (Sichuan, China). X-ray diffraction (XRD) analysis indicated that the slag was amorphous as shown in Fig. 2b and could be converted into crystalline state after annealing at 850 °C (Fig. 2a). The major crystal mineral constituents were perovskite (CaTiO3), titanium-bearing diopside (Ca(Ti,Mg,Al)(Si,Al)2O6), diopside (Ca(Mg,Fe,Al)(Si,Al)2O6), calcium

Fig. 2. The XRD patterns of the water-quenched slag before (b) and after (a) annealing in 850 °C.

Fig. 1. The process flowsheet of combined mineralization of CO2 and recovery of Ti and Al from the titanium-bearing blast furnace slag.


silicate (CaSiO3) and calcium aluminate (Ca3Al2O6). The chemical composition of the slag is listed in Table 1. From the literature [34], approximately 20%–30% of the titanium is low valence and distributed in the titanium-bearing diopside. The remaining titanium was mainly in the perovskite. The particle size distribution of water-quenched slag is shown in Fig. 3. The slag had a wide particle size distribution with a D50 of 38 μm and 99% of the particles were less than 100 μm. The morphology of the slag is presented in Fig. 4. Clearly, the slag consisted of smooth-faced irregular particles.

Table 1 Chemical composition of the TBBF slag used in this study (wt%) TiO2

CaO

MgO

SiO2

Al2O3

TFe

Others

21.19

26.84

7.25

23.31

13.69

2.38

3.22

585

must be determined. In each roasting experiment, the TBBF slag sample was thoroughly mixed with AS at a specific mass ratio. The mixture, placed in a crucible, was then heated at a rate of 10 °C·min−1 to the required temperature, ranging from 310 to 410 °C, and annealed for 120 min (except for the experiment for the effect of reaction time) in a muffle furnace (KSL-1200X, Heifei Kejing Materials Technology Co., Limited, Heifei, China). Then, the crucible was withdrawn and cooled naturally to room temperature. The roasted slags were leached with a 2.5 wt% H2SO4 at a liquid-tosolid mass ratio of 3:1 and 55 °C for 1 h and filtered. A leaching residue rich in CaSO4 and SiO2 and a leaching solution rich in TiOSO4, Al2(SO4)3, MgSO4 and AS was obtained. The concentrations of Ti, Al and Mg ions in the leaching solution were measured and the conversions calculated for evaluating the efficiency of the sulfation roasting. CaSO4 was slightly soluble in an aqueous solution and distributed between the leaching solution and the residue. Clearly, high conversion of Ti, Al and Mg components must correspond to high conversion of Ca since nearly all the phase constituents in the slag contained calcium, as shown in Fig. 2b. Besides, since an appreciable quantity of metal elements in the slag free of sulfation roasting with AS can be dissolved in a dilute acid solution, the conversion here actually reflected the contributions of both the roasting and acid leaching. The sulfation ratios (η1) of magnesium, titanium, or aluminium ions in the TBBF slag were calculated using Formula (1). η1 ¼ ½ðc1 V Þ=ðm α 1 Þ 100%

ð1Þ

where m (g) is the mass of the TBBF slag used in the roasting experiments, α1 (wt%) is the content of titanium, aluminium, or magnesium in the TBBF slag, V (L) is the volume of leaching solution, and c1 (g·L−1) is the concentration of titanium, aluminium, or magnesium ions in the leaching solution. The sulfation ratio (η2) of calcium in the TBBF slag was calculated using Formula (2). η2 ¼ ½ðc2 V þ α 3 m1 Þ=ðm α 2 Þ 100%

Fig. 3. The particle size distribution of the TBBF slag.

The chemicals, including ammonium sulphate (AS), sulfuric acid, ammonia, ammonium bicarbonate and ammonium carbonate were of analytical grade. 2.2. Experimental methods The purpose of roasting is to convert all the metal components in the slag into their sulphates, which are readily separated for further enrichment and carbonation. Thus, the optimum sulfation process parameters

ð2Þ

where c2 (g·L−1) is the concentration of calcium ion in the leaching solution, m1 (g) is the mass of leaching residue, α3 (wt%) is the content of calcium existing in the form of CaSO4 in the leaching residue, and α2 (wt%) is the content of calcium in the TBBF slag. The leaching solution was first subjected to thermal hydrolysis for recovery of titanium. In each test, 100 ml of the solution was placed in a 250-ml three-necked glass reactor fitted with magnetic stirring, a thermometer and a reflux condenser, and heated by using an oil bath with temperature fluctuation of ± 1 °C. The solution was heated at a rate of 5 °C·min−1 to 102 °C under stirring and annealed until a

Fig. 4. The SEM patterns of the TBBF slag.

586


precipitate occurred. This moment was considered the beginning of the hydrolysis. After a certain period of time, the reactor was taken out, and the slurry was cooled slowly and aged at 50 °C for 30–60 min. Then, the slurry was filtered. The concentration of titanium in the filtrate was measured and its hydrolysis ratio calculated. The hydrolysate was chemically analysed and characterized after calcining at 1000 °C for 2 h. The Ti-depleted leaching solution was further employed to recover aluminium. In each test, a certain amount of 13 wt% ammonia water was introduced to 50 ml of the leaching solution at room temperature, aged for 40 min and filtered. The pH and concentrations of Al and Mg in the filtrate were measured and the precipitation ratios calculated. The precipitate obtained at the optimal pH value was chemically analysed and characterized. The leaching residue was reacted with saturated NH4HCO3 solution at a molar ratio of NH4HCO3:CaSO4 of 3:1 and 50 °C for 120 min and filtered. The carbonation solid product was chemically analysed and characterized. The CO2 sequestration capacity M1 (kg CO2·t− 1) for the leaching residue per ton of TBBF slag was calculated using the following formula: M1 ¼ ½ðm2 α 4 Þ ð44=12Þ=m 103

ð3Þ

where m2 (g) is the mass of carbonation product from the leaching residue of m1 g, and α4 is the content (wt%) of carbon in the carbonation product. The Ti- and Al-depleted leaching solution was reacted with saturated (NH4)2CO3 solution. Since a small amount of CaSO4 dissolved in the leaching solution, this part of Ca would also participate in the carbonation reaction together with Mg. Thus, the reaction proceeded with a molar ratio of (NH4)2CO3 to (Ca2+ + Mg2+) of 3:1 at a temperature of 40 °C for 1 h. After filtration, a solid carbonate mixture and a filtrate rich in AS were obtained. The carbonation product was chemically analysed and characterized. The CO2 sequestration capacity M2 (kg CO2·t−1) for the Ti- and Aldepleted leaching solution per ton of TBBF slag was calculated using the following formula: M2 ¼ ½ðm3 α 5 Þ ð44=12Þ=m 103

respectively. The continuous scanning mode with a 0.03 s interval and 0.05 s set time was used to collect the XRD patterns. All elements in the TBBF slag were analysed using X-ray fluorescence spectrometry (XRF-1800, Shimadzu, Japan) with a Rh Kα radiation source. A thermogravimetric analyser (TGA/DSC, 2/1600, Mettler Toledo, Switzerland) and a mass spectrometer (GSD 320 T3, Pfeiffer Vacuum, Germany) were combined to investigate the thermal decomposition behaviour of the carbonation product of the Ti- and Al-depleted leaching solution under a nitrogen atmosphere at a flow rate of 30 ml·min− 1 and a heating rate of 10 °C·min−1. The CO2 amount fixed in the carbonation product in the form of CaCO3 and MgCO3 was analysed by a high frequency infrared-ray carbon and sulphur analyser (TL851–6, Lida, China). The surface morphology of the solid samples was observed using a scanning electron microscope (SEM, JSM-7500F, JEOL, Japan) at an accelerating voltage of 5 kV. The relative elemental content of samples was analysed with a combined energy-dispersive X-ray spectrometer (EDS, IS250, Oxford, Japan). The content of elements C, N, H and S in the Al-rich precipitation product was measured by an element analyser (EA) (Euro EA 3000, LEEMAN LABS, INC., USA) with their accuracy of C ≤ ± 0.1%, H ≤ ± 0.1%, N ≤ ± 0.05% and S ≤ ± 0.1%, respectively. The particle sizes of the TBBF slag were examined using an Omec LS-POP(9) Laser Particle Size Analyser (Zhuhai, China). 3. Results and Discussion 3.1. Sulfation roasting of TBBF slag 3.1.1. Effect of roasting temperature Fig. 5 shows the effect of roasting temperature on the sulfation of titanium, aluminium and magnesium elements in the TBBF slag.

ð4Þ

where m3 (g) is the mass of carbonation product from the Ti- and Aldepleted leaching solution, and α5 (wt%) is the content of carbon in the carbonation product. Therefore, the total CO2 sequestration capacity E (kg·t−1) per ton of BF slag can be calculated as follows: E ¼ M1 þ M2

ð5Þ

2.3. Analysis and characterization The concentrations of titanium, aluminium, calcium and magnesium ions in the leachates were determined by inductive coupled plasma atomic emission spectroscopy (ICP-AES) (Spectro ARCOS ICP, Germany). For determination of the chemical compositions of the leaching residue and the Al-rich precipitate, the solid sample was melted with sodium dioxide (Na2O2) and sodium hydroxide (NaOH) at 750 °C for 30 min and then leached with dilute hydrochloric acid. The concentrations of titanium, aluminium, calcium and magnesium ions in the solution thus obtained were analysed via ICP-AES. To determine the content of calcium sulphate in the leaching residue, 1 g of the residue was dissolved in 500 ml deionized water, and then the concentration of Ca in the solution was measured by ICP-AES. XRD analyses were performed using a DX-2007 X-ray diffraction spectrometer (Danton, China) operating with a Cu Kα radiation source that was filtered with a graphite monochromator at a frequency of λ = 1.54 nm. The voltage and anode current were 40 kV and 30 mA,

Fig. 5. Effect of the roasting temperature on the sulfation of titanium, aluminium and magnesium components (roasting: AS-to-slag mass ratio 2:1, reaction time 120 min; leaching: 2.5 wt% dilute sulfuric acid, liquid-to-solid mass ratio 3:1, 55 °C, 1 h).

The sulfation of the titanium component increased with the rising temperature. The sulfation was only 76% at 310 °C and almost levelled off with its equilibrium value of approximately 87% beyond 350 °C, while the sulfation of aluminium and magnesium components remained almost unchanged over the whole temperature range with their conversions of 85% and 92%, respectively. From the literature [35], the reaction of AS with silicate minerals can be divided into two stages, i.e., decomposition of AS into NH3 and NH4HSO4 and then the


digestion of silicate minerals by NH4HSO4. Therefore, the decomposition of TBBF slag by AS is essentially an acidolysis reaction. As a comparison, a blank experiment was conducted. The TBBF slag without roasting with AS was leached directly with 2.5 wt% dilute sulfuric acid at 55 °C for 1 h and the leaching of titanium, aluminium and magnesium was measured to be 10%, 29% and 54%, respectively. Clearly, the leaching of the three components resulted primarily from the diopside and titaniumbearing diopside since most of the magnesium and aluminium components and small amounts of titanium (20%–30%) in the TBBF slag were distributed in the diopside minerals, and CaTiO3 was much more acid resistant than the diopside minerals [24]. The results indicated that even at low temperatures, there were significant amounts of diopside minerals decomposed by dilute sulfuric acid, which could explain why the conversion of aluminium and magnesium became insensitive to the temperature employed during the roasting, which was much higher than the direct acid leaching temperature. Therefore, an increased roasting temperature mainly enhanced the conversion of CaTiO3 in the present study. Fig. 6 shows the XRD patterns of roasted slag at different temperatures. The major crystal products were CaSO4 and the compounded sulphates of aluminium combined with AS. The diffraction peaks of CaSO4 increased with the rise in temperature, probably due to the elevating CaTiO3 conversion, and beyond 350 °C, remained almost unchanged while the sulphate of aluminium at low temperatures was (NH4)3Al(SO4)3 and converted to be NH4Al(SO4)2 at high temperatures, indicating that the conversion of aluminium seems to change little. The results were in good agreement with the values implied by Fig. 5. The sulphates containing titanium and magnesium species were not observed, probably due to low crystallinity.

Fig. 7. Effect of the AS to TBBF slag mass ratio on the sulfation of titanium, aluminium and magnesium components (roasting: temperature 350 °C, reaction time 120 min; leaching: 2.5 wt% dilute sulfuric acid, liquid-to-solid mass ratio 3:1, 55 °C, 1 h).

sulfation ratios of titanium, aluminium and magnesium increased with the increasing mass ratio, the significant increase occurred at the mass ratio less than 2:1, which is close to the theoretical value. Further increasing the mass ratio, the sulfation of both magnesium and aluminium almost approached equilibrium while the sulfation of titanium increased slowly due to the more rapid reaction of diopside minerals with AS than the reaction of CaTiO3. Taking into consideration the energy consumption, mainly in the regeneration and recycle of AS, and economy of the whole process, the optimal mass ratio was chosen to be 2:1, where the titanium extraction was up to 87%, although a higher mass ratio would result in higher titanium recovery. Further, a comparison experiment for leaching of the slag roasted at the mass ratio of 2:1 with deionized water instead of 2.5 wt% H2SO4 demonstrated that the extractions of Ti, Mg and Al were 4%, 92% and 84%, respectively, and the slurry thus obtained was difficult to filter, indicating the hydrolysis of the titanium sulphate in the roasted slag might occur. To restrain the titanium hydrolysis during the water leaching, a minimum mass ratio of 6:1 was required due to the presence of some amounts of NH4HSO4 and (NH4)3H(SO4)2 in the roasted slag at high mass ratios (Fig. 8). Clearly, the leaching with a dilute acid solution

Fig. 6. XRD patterns of the roasted slag at different temperatures.

3.1.2. Effect of the mass ratio of AS to TBBF slag Assumption that the reactions of the metal oxides (abbreviated as MO or M2O3) in the TBBF slag with AS can be written as Eqs. (6) and (7). MO þ ðNH4 Þ2 SO4 ¼ MSO4 þ 2NH3 þ H2 O ðM ¼ Ca; Mg; …:Þ

ð6Þ

M2 O3 þ 3ðNH4 Þ2 SO4 ¼ M2 ðSO4 Þ3 þ 6NH3 þ 3H2 O ðM ¼ Al; Fe……Þ ð7Þ The theoretical mass ratio of AS to TBBF slag was calculated to be ~1.9 based on the chemical compositions listed in Table 1. Fig. 7 shows the effect of the mass ratio of AS to TBBF slag on the sulfation of titanium, aluminium and magnesium elements. The

587

Fig. 8. XRD patterns of the roasted slags at different AS to TBBF slag mass ratios.

588


can avoid the use of a high AS to TBBF slag mass ratio employed by other studies [30,32]. Fig. 8 shows the XRD patterns of the roasted slags at different mass ratios. The phase constituents of roasted products became more complicated with increasing mass ratio. At a mass ratio larger than 4:1, NH4HSO4 and (NH4)3H(SO4)2 appeared in the products, and the double salts formed by the metal sulphates and AS changed z+ towards higher molar ratio of NH+ (Mz+ = Ca2+, Al3+, Fe3+). 4 -to-M Obviously, the dosage of ammonium sulphate was then excessive. 3.1.3. Effect of the roasting time Fig. 9 shows the effect of roasting time on the sulfation of titanium, aluminium and magnesium components in the TBBF slag. The sulfation of aluminium and magnesium reached a plateau at times beyond 60 min while the sulfation of titanium increased monotonically, although its incremental range gradually decreased. A roasting time of 120 min was chosen to be the optimal parameter.

strong hygroscopicity of MgSO4, which would absorb ambient moisture during sample storage, preparation and testing. With the increasing reaction time it became amorphous, maybe forming a compound sulphate of magnesium and ammonium ions with low crystallinity. The diffraction peaks of CaSO4 increased with the rise of reaction time, due to the elevating CaTiO3 conversion, and beyond 120 min remained almost unchanged while the compound sulphate of aluminium and ammonium ions was (NH4)3Al(SO4)3 in the initial stage and it gradually changed to NH4Al(SO4)2 with the extension of reaction time. The conversion reaction can be expressed using Eq. (8): ðNH4 Þ3 AlðSO4 Þ3 ¼ NH4 AlðSO4 Þ2 þ NH4 HSO4 þ NH3

ð8Þ

The NH4HSO4 produced would further digest the TBBF slag. Based on the study above, the optimal roasting parameters were determined as follows: roasting temperature of 350 °C, AS to TBBF slag mass ratio of 2:1, and reaction time of 120 min. The roasted slag under the optimal condition was leached with a 2.5 wt% dilute sulfuric acid and the sulfation of titanium, aluminium, magnesium and calcium were measured as 87%, 84.4%, 92.6% and 85%, respectively. Fig. 11 shows the XRD pattern of the leaching residue. The primary crystal phases were calcium sulphate and unreacted perovskite. The existence of perovskite was responsible for the not too high titanium extraction. The content of calcium sulphate was measured to be 69.4 wt%. The chemical composition of the residue is presented in Table 2. A small amount of Al2O3 was detected, which was also responsible for the not too high aluminium extraction. We guess the Al2O3 present in the residue might be mainly in the form of Ca3Al2O6, which was even more resistant to acid digestion than perovskite in this study.

Fig. 9. The effect of the roasting time on the sulfation ratio of titanium, aluminium and magnesium (roasting: temperature 350 °C, AS-to-ore mass ratio 2:1; leaching: 2.5 wt% dilute sulfuric acid, liquid-to-solid mass ratio 3:1, 55 °C, 1 h).

Fig. 10 shows the XRD patterns of the roasted slag at different times. The sulphate of magnesium, MgSO4·7H2O, was observed in the initial stage. The occurrence of the crystal water was probably attributed to

Fig. 11. XRD pattern of the leaching residue.

Table 2 Chemical composition of the leaching residue (wt%) CaO

TiO2

SiO2

SO3

Al2O3

30.5

3.4

22.5

34.4

2.4

3.2. Recovery of titanium and aluminium components from the leaching solution

Fig. 10. XRD patterns of the roasted slag at different times.

The leaching solution obtained under the optimal roasting conditions consisted of 15.1 g·L− 1 Ti4 +, 8.4 g·L− 1 Al3 +, 5.6 g·L− 1 Mg2 + and 0.8 g·L−1 Ca2+. The solution was subjected to thermal hydrolysis


as described in the 2.2 section, Experimental methods, and the hydrolysis ratios of titanium at different times are shown in Fig. 12. The hydrolysis at 4 h reached 95.7%, close to the value in the commercial sulphate process for TiO2 production [36].

Fig. 12. Hydrolysis ratios of titanium at different times.

Fig. 13 shows the SEM photo of the 4 h hydrolysate. The product particles were uniform and spherical, with a size distribution of 300– 500 nm. Fig. 14 shows the XRD patterns of the hydrolysates before

589

and after calcining at 1000 °C. The primary crystalline phase in the hydrolysate was anatase, and was partially transformed into rutile after calcination. Chemical analysis indicated that the calcined product consisted of ~ 98 wt% TiO2 and ~ 2 wt% SiO2, demonstrating that a high-purity titanium dioxide was obtained. The mother liquor after titanium hydrolysis (the Ti-depleted leaching solution) was further subjected to ammonia neutralization for selective precipitation of Al3 +. The precipitation ratios of Al3 + and Mg2 + at different pH values are shown in Table 3. 99.7% of the Al3 + was precipitated while co-precipitated Mg2 + was only 0.89% at pH = 5, considered the optimal precipitation conditions. The XRD analysis showed that the Al-rich precipitate was amorphous. The chemical composition of the precipitate is listed in Table 4. The precipitate contained large quantities of Al2O3 (ca. 49.9 wt%) and a small amount of impurities such as titanium, iron and silicon. The LOI reached 40.25%, while the EA analysis didn't detect N element, indicating that the Al-rich precipitate contained a large amount of OH− and free water. Therefore, we inferred that the major Al-containing phase was basic aluminium sulphate. The Al-rich precipitate was calcined at 850 °C, and its XRD pattern is shown in Fig. 15. The main crystal phase was Al2O3. The purity of Al2O3 and the mass ratio of Al2O3 to SiO2 were measured to be 70.8 wt% and approximately 49, respectively, so this material could be used as a raw material for electrolytic aluminium production. Table 3 Variation of precipitation of Al3+ and Mg2+ with pH value pH

4

5

6

Precipitation ratio of Al3+/% Precipitation ratio of Mg2+/%

67 0.61

99.7 0.89

99.8 1.5

Table 4 Chemical composition of the Al-rich precipitate (wt%) Al2O3

SiO2

SO3

TiO2

Fe2O3

LOI

49.9

1.07

9.48

2.04

2.47

40.25

Fig. 13. SEM photo of the 4 h hydrolysate.

Fig. 15. XRD pattern of the precipitate after calcining at 850 °C.

3.3. Mineral carbonation of TBBF slag

Fig. 14. XRD patterns of the hydrolysates: (a) dried at 45 °C; (b) calcined at 1000 °C.

3.3.1. Carbonation of the leaching residue Recently, carbonation of various CaSO4-containing industrial solid wastes using CO2 under NH3 media or NH4HCO3 obtained by NH3

590


capture of CO2 in flue gases was reported [37–39]. In this study one of the main phase constitutions of the leaching residue was also CaSO4. Carbonation of the leaching residue was conducted with saturated NH4HCO3 solution at 50 °C with the molar ratio of NH4HCO3 to CaSO4 of 3:1. The conversion of calcium sulphate in 120 min was measured as 99.7%. Fig. 16 shows the XRD pattern of the carbonation product, and CaCO3 was the primary crystalline phase.

Mg2 + and Ca2 + reached 91% and ~ 100%, respectively. XRD analysis indicated that the main crystal product was (NH4)2Mg(CO3)2·4H2O as shown in Fig. 18. The SEM photos of the product are shown in Fig. 19. Combined with the EDS analysis, the majority of precipitates were compact, column-like (NH4)2Mg(CO3)2·4H2O crystals of dozens of microns while a small amount of minute particles of several microns were calcium carbonate. Based on the results, the carbonation reactions of the leaching solution can be expressed as follows: MgSO4 þ 2ðNH4 Þ2 CO3 þ 4H2 O ¼ ðNH4 Þ2 MgðCO3 Þ2 4H2 O↓ þ ðNH4 Þ2 SO4

ð9Þ CaSO4 þ ðNH4 Þ2 CO3 ¼ CaCO3 ↓ þ ðNH4 Þ2 SO4

ð10Þ

Fig. 16. XRD pattern of the carbonation product of the leaching residue.

Fig. 17 shows the SEM photos of the carbonation product. Combined with the EDS analysis, the irregular smooth lump particles were amorphous SiO2, while the regular minute particles with particle size of 2–4 μm were crystalline CaCO3. The content of carbon in the carbonation product was determined by a high frequency infrared-ray carbon and sulphur analyser and the CO2 sequestration capacity M1 of the leaching residue per ton of TBBF slag was calculated to be 170 kg·t−1 TBBF slag from Formula (3). Chemical analysis showed that the carbonation product consisted of 61.7 wt% of CaCO3, 31.3 wt% of SiO2 and 2.8 wt% of CaTiO3, and can be utilized as the raw material for cement production for the replacement of natural limestone and silica. 3.3.2. Carbonation of the Mg-rich leaching solution Carbonation of the Ti- and Al-depleted leaching solution with (NH4)2CO3 solution was carried out at 40 °C and the molar ratio of (NH4)2CO3 to Ca2++Mg2+ of 3:1 for 60 min. The conversion ratios of

Fig. 18. XRD pattern of the carbonation product of the magnesium-rich leaching solution.

TG-Mass was employed to analyse the thermal stability of the carbonation product, and the result is shown in Fig. 20. There are three obvious temperature ranges of mass loss: 30–200 °C, 200–500 °C and 500–750 °C. In the first temperature range, there was a large amount of NH3, CO2 and H2O. Obviously, the presence of this material is related to the decomposition of (NH4)2CO3 and removal of the adsorbed and/or crystal water. Therefore, low-temperature decomposition can be adopted to recover the ammonia for reuse from the carbonation product. The appearance of CO2 in the second temperature range was

CaCO3

SiO2

CaCO3

Fig. 17. SEM photos of the carbonation product of the leaching residue.


591

CaCO3 CaCO3

(NH4)2Mg(CO3)2·4H2O

Fig. 19. SEM photos of the carbonation product of the magnesium-rich leaching solution.

The sulfation of calcium, magnesium, titanium and aluminium components reached 83%, 92.6%, 87% and 84.4%, respectively. The leaching solution was subjected to hydrolysis at 102 °C for 4 h with a titanium hydrolysis ratio of 95.7% and the purity of titanium dioxide in the calcined hydrolysate reached up to 98 wt%. The Ti-depleted leaching solution was neutralized to pH = 5 with ammonia and approximately 99.7% of the aluminium was selectively precipitated with the coprecipitated magnesium less than 1%. The NH3 produced during the roasting is used to capture CO2 from the flue gases. The NH4HCO3 and (NH4)2CO3 thus obtained are used to carbonate the CaSO4-containing leaching residue and the MgSO4-rich leaching solution, respectively. In this process, approximately 82.1% of Ca and 84.2% of Mg in the TBBF slag were transformed into its carbonates with a total CO2 sequestration capacity of 239.7 kg per ton of TBBF slag.

Fig. 20. TG-Mass curves of the carbonation product of the magnesium-rich leaching solution.

attributed to the decomposition of MgCO3. Both the results in the first two temperature ranges were in good agreement with the literature [40]. In the last temperature range, a distinct CO2 peak, which was attributed to the decomposition of calcium carbonate [41], was observed. Based on the analysis of TG-Mass, the amount of CO2 fixed in the carbonation product of the Mg-rich leaching solution in the form of CaCO3 and MgCO3 was measured using a carbon and sulphur analyser following decomposition of the (NH4)2CO3 at 200 °C. The CO2 sequestration capacity M2 (kg·t−1) of the leaching solution per ton of the BF slag was calculated to be 69.7 kg CO2·(t slag)−1. The contents of MgCO3 and CaCO3 in the calcined product were measured to be approximately 95.6 wt% and 4.4 wt%, respectively. The product can be used as raw material in light magnesium carbonate production for the replacement of natural dolomite. Accordingly, approximately 82.1% of calcium and 84.2% of magnesium in the TBBF slag can be utilized to chemically sequester CO2 in this process, and their sequestration capacity reached 239.7 kg CO2 per ton of TBBF slag. 4. Conclusions In this paper, a novel, facile method for mineralization of CO2 coupled with recovery of TiO2 and Al2O3 was proposed, in which the TBBF slag was roasted with recyclable AS at a mass ratio of AS to TBBF slag of 2:1 and 350 °C for 2 h followed by leaching in a 2.5 wt% H2SO4.

References [1] Q.C. Chao, W.Y. Chen, The summary of carbon capture and storage technology and its impact on China, Adv. Earth Sci. 67 (2006) 1769–1774. [2] R.H. Worden, L.K. Smith, Geological sequestration of CO2 in the subsurface: lessons from CO2 injection enhanced oil recovery projects in oilfields, National Cathol. Rep. 233 (2004) 211–224. [3] E.E. Adams, D.S. Golomb, H.J. Herzog, Ocean disposal of CO2 at intermediate depths, Energy Convers. Manag. 36 (1995) 447–452. [4] K.S. Lackner, A guide to CO2 sequestration, Science 300 (2003) 1677–1678. [5] A. Santos, J.A. Toledo-Fernández, R. Mendoza-serna, L. Gago-Duport, N. Rosa-Fox, Chemically active silica aerogel− Wollastonite composites for CO2 fixation by carbonation reactions, Ind. Eng. Chem. Res. 46 (2006) 103–107. [6] W.J.J. Huijgen, G.J. Witkamp, R.N.J. Comans, Mechanisms of aqueous wollastonite carbonation as a possible CO2 sequestration process, Chem. Eng. Sci. 61 (2006) 4242–4251. [7] C.Y. Tai, W. Chen, S. Shin-Min, Factors affecting wollastonite carbonation under CO2 supercritical conditions, AIChE J. 52 (2006) 292–299. [8] E.E. Chang, S.Y. Pan, Y.H. Chen, C.S. Tan, P.C. Chiang, Accelerated carbonation of steelmaking slags in a high-gravity rotating packed bed, J. Hazard. Mater. 227-228 (2012) 97–106. [9] M. Ghoorah, R.D. Balucan, E.M. Kennedy, B.Z. Dlugogorski, Selection of acid for weak acid processing of Australian wollastonite for mineralisation of CO2, Fuel 122 (2011) 277–286. [10] S. Teir, H. Revitzer, S. Eloneva, C.J. Fogelholm, R. Zevenhoven, Dissolution of natural serpentinite in mineral and organic acids, Int. J. Miner. Process. 83 (2007) 36–46. [11] M.M. Maroto-Valer, D.J. Fauth, M.E. Kuchta, Y. Zhang, J.M. Andrésen, Activation of magnesium rich minerals as carbonation feedstock materials for CO2 sequestration, Fuel Process. Technol. 86 (2005) 1627–1645. [12] E.I. Nduagu, J. Highfield, J. Chen, R. Zevenhoven, Mechanisms of serpentineammonium sulfate reactions: towards higher efficiencies in flux recovery and Mg extraction for CO2 mineral sequestration, RSC Adv. 4 (2014) 64494–64505. [13] K. Jarvis, R.W. Carpenter, T. Windman, Y. Kim, R. Nunez, F. Alawneh, Reaction mechanisms for enhancing mineral sequestration of CO2, Environ. Sci. Technol. 43 (2009) 6314. [14] W.J. Huijgen, G.J. Witkamp, R.N. Comans, Mineral CO2 sequestration by steel slag carbonation, Environ. Sci. Technol. 39 (2005) 9676-6982. [15] H.Y. Sheng, L. Lv, B. Liang, C. Li, Aqueous carbonation of the potassium-depleted residue from potassium feldspar–CaCl2 calcination for CO2 fixation, Environ. Earth Sci. 73 (2015) 6871–6879.

592


[16] C. Cárdenas-Escudero, V. Morales-Flórez, R. Pérez-López, A. Santos, L. Esquivias, Procedure to use phosphogypsum industrial waste for mineral CO2 sequestration, J. Hazard. Mater. 196 (2011) 431–435. [17] S. Teir, S. Eloneva, C.J. Fogelholm, R. Zevenhoven, Dissolution of steelmaking slags in acetic acid for precipitated calcium carbonate production, Energy 32 (2007) 528–539. [18] X. Yue, Theoretical and experimental study on waste heat recovery of Blast furnace slag pyrolysising biomass, Master thesis, Qingdao Techological University, Qingdao, China, 2013. (in Chinese). [19] A. Said, H.P. Mattila, M. Järvinen, R. Zevenhoven, Production of precipitated calcium carbonate (PCC) from steelmaking slag for fixation of CO2, Appl. Energy 112 (2013) 765–771. [20] The illustration global crude steel production in 2015, http://finance.sina.com.cn/ roll/2016-02-25/doc-ifxpvutf3363699.shtml 2016–02-25. [21] Y. Peng, Development of technologies for recovering titanium from Pangang BF slag, Titan. Ind. Prog. 22 (2005) 44–48. (in Chinese). [22] Z.T. Sui, Z.Z. Guo, L. Zhang, L.N. Zhang, M.Y. Wang, T.P. Tai, G.Q. Li, Green separation technique of Ti component from Ti-bearing blast furnace slag, J. Mate. MetaII. 5 (2006) 93–97. (in Chinese). [23] P.M. Guo, P. Zhao, Technical analysis on selective and enrichment of Ti-bearing blast furnace slag based on phase diagrams, Iron Steel Vanadium Titanium 26 (2005) 5–10. (in Chinese). [24] Y. Xiong, C. Li, B. Liang, J. Xie, Leaching behavior of air cooled Ti-bearing blastfurnace slag in hydrochloric acid, Chin. J. Nonferrous Met. 18 (2008) 557–563. (in Chinese). [25] F. Yan, C. Li, B. Liang, A two-step sulfuric acid leaching process of Ti-bearing blast furnace slag, Chin. J. Process. Eng. 6 (2006) 413–417. (in Chinese). [26] Y. Wang, T. Qi, J. Chu, W. Zhao, Production of TiO2 from CaTiO3 by alkaline roasting method, Rare Metals 29 (2010) 162–167. [27] Y.C. Gong, K.H. Qiu, J.H. Li, J.F. Li, Preparation of titanium-rich raw material from titanium-bearing blast furnace slag by low temperature acid-base method, Min. Metall. Eng. 30 (2010) 79–82. (in Chinese). [28] S.F. Medina, J.E. Mancilla, Influence of alloying elements in solution on static recrystallization kinetics of hot deformed steels, ISIJ Int. 36 (1996) 1063–1069. [29] X.L. Zou, X.G. Lu, Preparation of alloy by direct reduction of Ti-bearing blast furnace slag, Chin. J. Nonferrous Met. 20 (2010) 1829–1835. (in Chinese).

[30] S.J. Wang, Y. Zhang, X.X. Xue, H. Yang, Recovery of titanium from titanium-bearing blast furnace slag by ammonium sulfate melting method, J. Chem. Eng. 63 (2012) 991–995. (in Chinese). [31] X.X. Xue, Y. Zhang, H. Yang, L.L. Bai, A Method of Comprehensive Utilization of Titanium-Bearing Blast Furnace Slag, CN. Patent 101988158A. (in Chinese). [32] M.T. Li, Reserch on process of titanium leaching and extracting from titanium-bearing blast furnace slag, Master thesis, Northeastern University, Shenyang, Liaoning, China, 2013. (in Chinese). [33] C. Li, L. Wang, W.Z. Lui, B. Liang, H.R. Yue, J.P. Hu, L. Lv, H.P. Xie, A method of comprehensive utilization of titanium-bearing blast furnace slag, CN. Patent 106082322A. (2016) (in Chinese). [34] Y.Q. Li, C.M. Ke, S.X. Hou, B.Q. Han, N li, study on melting reduction of induction furnace for titanium-bearing blast furnace slag of Panzhihua, Refractories 41 (2007) 244–247. (in Chinese). [35] J. Highfield, H. Lim, J. Fagerlund, R. Zevenhoven, Activation of serpentine for CO2 mineralization by flux extraction of soluble magnesium salts using ammonium sulfate, RSC Adv. 2 (2012) 6535–6541. [36] L.J. Zheng, B. Liang, L. Lv, L.N. Jia, C. Li, Effect of impurities on the hydrolysis of lowconcentration titanyl sulfate solutions, Res. Chem. Intermed. 41 (2014) 1–16. [37] S.M. Pérez-Moreno, M.J. Gázquez, J.P. Bolívar, CO2 sequestration by indirect carbonation of artificial gypsum generated in the manufacture of titanium dioxide pigments, Chem. Eng. J. 262 (2015) 737–746. [38] H. Zhao, H. Li, W. Bao, S. Li, W. Lin, Experimental study of enhanced phosphogypsum carbonation with ammonia under increased CO2 pressure, J. CO2 Util. 11 (2014) 10–19. [39] M.G. Lee, Y.N. Jang, K.W. Ryu, W. Kim, J.H. Bang, Mineral carbonation of flue gas desulfurization gypsum for CO2 sequestration, Energy 47 (2012) 370–377. [40] R.M. Dell, S.W. Weller, The thermal decomposition of nesquehonite MgCO3·3H2O and magnesium ammonium carbonate MgCO3·(NH4)2CO3·4H2O, Trans. Faraday Soc. 55 (1959) 2203–2220. [41] L. Ye, H. Yue, Y. Wang, H. Sheng, B. Yuan, CO2 mineralization of activated K-feldspar + CaCl2 slag to fix carbon and produce soluble potash salt, Ind. Eng. Chem. Res. 53 (2014) 10557–10565.