Experimental Study on the Dissolution Behavior of Calcium Fluoride

: The presence of halogens has an adverse e ﬀ ect on the zinc extraction process through electrowinning, the last phase of the RLE (Roasting, Leaching and Electrowinning) zinc extraction route. Fluoride (F − ) may be present as calcium ﬂuoride (CaF 2 ) and this is, for example, the case in double leached Waelz oxide (DLWO). E ﬃ cient removal of F − from primary and secondary raw materials for zinc extraction results in a simpliﬁed process and increases ﬂexibility in the selection of raw materials. Understanding of the solubility behavior of pure CaF 2 can give valuable information on treatment for maximized halogen removal. Dissolution of CaF 2 was studied with the addition of sodium carbonate (Na 2 CO 3 ) and sodium bicarbonate (NaHCO 3 ). Dissolution studies were combined with thermodynamic calculations to understand the solubility behavior of CaF 2 under di ﬀ erent conditions. Results from the experiments and the thermodynamic calculations show that Na 2 CO 3 and NaHCO 3 have similar behavior if the pH is controlled at the same value. The available carbonate (CO 32 − ) ion in the system limits the concentration of calcium (Ca 2 + ) ion by precipitation of CaCO 3 , which enhances the dissolution of CaF 2 . At higher temperatures and pH, calcite, vaterite, and aragonite were formed and co-precipitation of CaF 2 along with calcium carbonate (CaCO 3 ) was observed. At lower temperatures and lower pH levels, only calcite and vaterite were formed and a coating by CaCO 3 on CaF 2 was found to hinder complete dissolution reaction. The results of this study indicate that the temperature along with the reagents used for the dissolution tests have a signiﬁcant impact on the CaCO 3 polymorph mixture (calcite, vaterite and aragonite) formation.

During electrowinning, F − attacks the aluminum (Al) cathode, leading to formation of cavities where Zn deposits. The deposited Zn may form an alloy with Al cathode, which leads to difficulties in stripping the Zn sheet [8][9][10][11].
Of the few methods of removal of halogens from WO, the most common is through lixiviation with sodium carbonate (Na 2 CO 3 ) [3] at pH 8-10 (Refer Appendix A.1 about self-generated pH of WO slurry) and temperatures of 50-90 • C [3,5]. Autoclave pressure leaching of WO by using Na 2 CO 3 is also an industrially established process for removal of the halogens [12] WO can, alternatively, be washed in a single stage leaching with 100 g/L Na 2 CO 3 , at 90 • C with pulp density of 30% for 3 h [5]. Characterization studies conducted on DLWO confirmed the presence of calcium fluoride (CaF 2 ) [4,7]. Furthermore, F − content determined with fluoride ion selective electrode (FISE), 19 F Liquid State Nuclear Magnetic Resonance (LS NMR) and calcium content determined by inductively coupled plasma-optical emission spectroscopy (ICP-OES) showed that the content of CaF 2 was~0.2% [7] in analyzed samples. Though literature on leaching of WO [3][4][5] with Na 2 CO 3 gives a general explanation on removal of halides, detailed knowledge of the dissolution behavior of CaF 2 during applied alkaline leaching conditions is not available.
The alkaline leaching [5] can be described by the following reaction: MeF 2 (s) + Na 2 CO 3 (aq) → MeCO 3 (s) + 2 NaF (aq) where Me = Zn, Pb, Ca. It is known from studies conducted on natural ground water at alkaline pH of 7.5-8.5 due to naturally occurring sodium bicarbonate, dissociation of natural CaF 2 is enhanced, precipitating calcium carbonate (CaCO 3 ) as calcite and increasing F − concentration in ground water [13][14][15]. Based on this knowledge the solubility of fluorite was studied by Rao et al. [15] in the laboratory at 28 • C with 1M Na 2 CO 3 and 2M NaHCO 3 as leaching reagent precipitating calcite after 30 days of leaching. The dissociation of CaF 2 is expressed as: CaCO 3 is known to exhibit polymorphism i.e., the ability of crystalline system to exist in more than one crystal structure. A review article by Meldrum, F.C., & Cölfen, H. on polymorphism of CaCO 3 on biological and synthetic system indicates three crystalline polymorphs of anhydrous CaCO 3 namely vaterite, aragonite, and calcite, listed here in order of increasing thermodynamic stability [16]. In nature, aragonite and calcite are commonly seen. The least dense amorphous CaCO 3 is precipitated first, which rapidly changes to hollow metastable vaterite, which further changes to aragonite or calcite depending on the free energies of activation of nucleation in different environments. Several experimental studies were conducted to understand CaCO 3 polymorphism using soluble salts of calcium where the sources of carbonate ions include soluble carbonates and bicarbonates [16,17]. However, literature on polymorphism exhibited by precipitated CaCO 3 from sparingly soluble CaF 2 is not available. Ichikuni (1979) showed that due to ion exchange reaction, one carbonate ion in aragonite is replaced by two fluoride ions, leading to coprecipitation of fluoride in CaCl 2 -Na 2 CO 3 -NaF system [18] Y. Kitano and M. Okumura concluded that the coprecipitation of fluoride with aragonite is larger than calcite [19].
Therefore, in order to develop an efficient removal method of CaF 2 from secondary zinc-containing dust material such as WO and DLWO, understanding the dissolution chemistry of CaF 2 with Na 2 CO 3 /NaHCO 3 and the crystallization process of CaCO 3 in a more detailed manner is required. For this purpose, a dissolution study of pure CaF 2 was conducted under different conditions with either Na 2 CO 3 or NaHCO 3 as reagent. Results from laboratory dissolution tests were evaluated with support from thermodynamic calculation using FactSage 7.2. Design of CaF 2 dissolution tests was based on reported research work and patents [3,5,12] and on the fact that the dissolution of zinc oxide is lowest at pH around 9 [4].

Particle Size
Particle size distribution of CaF 2 was conducted through laser-diffraction measurements using a CILAS 1064 unit (CILAS, Orleans, France) and Fraunhofer approximation was used for the evaluation of the collected data.

Dissolution Experiments
The dissolution experiments were conducted batch-wise in a four-necked round-bottomed 1-L glass reactor fitted with a dropping funnel, mechanical paddle stirrer, pH electrode, thermometer, and heating mantle. The suspensions were stirred at a speed of 300 rpm/min and the temperature was kept at set value by using a temperature controller. The experimental dissolution conditions were: reagents (150 mM Na 2 CO 3 and 150 mM NaHCO 3 ); temperatures (70 • C and 95 • C) and pH (10.6 and 9.0). The natural pH of 150 mM (16 g/L) Na 2 CO 3 is 10.6 and of 150 mM (12.68 g/L) NaHCO 3 is 8.7. To meet the need to decrease or increase the pH, concentrated HNO 3 or 10 M NaOH, respectively, was added. One liter of water with selected dissolution reagent was taken in a glass reactor, heated to desired temperature, pH was adjusted and solid 300 mg/L CaF 2 added. To understand the reaction kinetics, experiments were performed for 480 min and, at selected time intervals, 20 mL samples of the solution were collected for the analysis of dissolved F − in the leach liquor. To minimize the error due to large sampling volume for each dissolution condition, three identical experimental set ups were assembled. From each experimental set up, only three samples were collected but at different pre-determined time intervals (after 5 min, 15 min, and 30 min from set up one; after 60 min, 120 min, and 180 min from set up two; after 240 min, 360 min, and 480 min from set up three). On completion of the experiment, the solid residue from set three was filtered and dried for mineralogical characterization by XRD-Rietveld refinement and SEM-EDS.

Fluoride Analyses
Fluoride analyses were carried out using Fluoride Ion Selective Electrode (FISE Metrohm 6.0502.150, Herisau, Switzerland, Titrando 888 with Tiamo software) and the standard methods for the examination of water and wastewater were applied [20]. A calibration curve with measured potentials in mV as function of concentration in mg/L was plotted for standards of, 2, 20, and 200 mg/L. The content of F − was deduced from the calibration curve using the measured potentials in mV for each sample.

Thermodynamic Calculations
The Equilibrium module of FactSage 7.2 (Thermfact/CRCT, Montreal, QC, Canada and GTT-Technologies, Aachen, Germany) [21,22] was used to carry out equilibrium calculations in the pH range of 7 to 11 by changing the concentration of H + ion and using the database FactPS (Fact Pure substances database 2018). Calculations involving aqueous species such as Na + , HCO 3 − , CO 3 2− , NaHCO 3 , F − , Ca 2+ and gaseous species CO 2 were conducted for the experimentally used dissolution conditions. This also included the effect from temperature, type of reagent, the added amount of reagent and CaF 2 .

Inductively Coupled Plasma-Optical Emission Spectroscopy
Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES, Model: Thermo Scientific ICAP 7200 DUO, Thermo Fisher Scientific, Waltham, MA, USA) was used to analyze the calcium and sodium content in the solution.

Quantitative X-ray Powder Diffraction with Rietveld Analysis
To determine and quantify mineralogical components present in the precipitated residue from the dissolution experiments X-ray diffraction (XRD) analyses were performed. A PANalytical Empyrean X-ray diffractometer (Malvern Panalytical, Almelo, The Netherlands), equipped with copper K α radiation of 45 kV and 40 mA was used for XRD analyses. XRD patterns were recorded in the 2-theta range 10-90 • with a step of 0.026 • . Evaluation of data was carried out using the software HighScore Plus and the Inorganic Crystal Structure Database (ICSD). For Rietveld refinement, background correction, scale factors, unit cells, preferred orientations and profile variables were included.

Scanning Electron Microscopy (SEM) Coupled with Energy Dispersive Spectroscopy (EDS)
SEM-EDS was conducted on the polished and carbon-coated epoxy sample using Zeiss Gemini Merlin equipped with an energy dispersive spectrometer (EDS-Xmax 80 mm, Zeiss, Oberkochen, Germany). The acceleration voltage was set to 10 kV and the emission current was 250 pA.

Particle Size
The particle size distribution of CaF 2 feed showed that the mean diameter was 3.37 µm, 50% particles were <3.00 µm, whereas the majority (90%) of the particles were <5.93 µm. All the particles are less than 13 µm in size.

Dissolution at pH 10.6
Dissolution of CaF 2 for 480 min at 70 • C and pH 10.6 with Na 2 CO 3 showed 99.3% dissolution of fluoride whereas at 95 • C and pH 10.6 83.3% dissolution was achieved (Figure 1). At higher temperature (95 • C) the dissolution reaction is initially rapid and F − content in the solution reaches maximum after approximately 100 min. The dissolution rate is retarded after 100 min. As can be seen from Figure 1, similar result was observed when NaHCO 3 was used and the pH was adjusted to 10.6 using 10 M NaOH. However, an equilibrium calculation based on experimental data indicates that complete dissolution of added amount of CaF 2 is theoretically possible under the given experimental condition ( Figure 1). Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES, Model: Thermo Scientific ICAP 7200 DUO, Thermo Fisher Scientific, Waltham, MA, USA) was used to analyze the calcium and sodium content in the solution.

Quantitative X-Ray Powder Diffraction with Rietveld Analysis
To determine and quantify mineralogical components present in the precipitated residue from the dissolution experiments X-ray diffraction (XRD) analyses were performed. A PANalytical Empyrean X-ray diffractometer (Malvern Panalytical, Almelo, The Netherlands), equipped with copper Kα radiation of 45 kV and 40 mA was used for XRD analyses. XRD patterns were recorded in the 2-theta range 10-90° with a step of 0.026°. Evaluation of data was carried out using the software HighScore Plus and the Inorganic Crystal Structure Database (ICSD). For Rietveld refinement, background correction, scale factors, unit cells, preferred orientations and profile variables were included.

Scanning Electron Microscopy (SEM) Coupled with Energy Dispersive Spectroscopy (EDS)
SEM-EDS was conducted on the polished and carbon-coated epoxy sample using Zeiss Gemini Merlin equipped with an energy dispersive spectrometer (EDS-Xmax 80 mm, Zeiss, Oberkochen, Germany). The acceleration voltage was set to 10 kV and the emission current was 250 pA.

Particle Size
The particle size distribution of CaF2 feed showed that the mean diameter was 3.37 μm, 50% particles were <3.00 μm, whereas the majority (90%) of the particles were <5.93 μm. All the particles are less than 13 μm in size.

Dissolution at pH 10.6
Dissolution of CaF2 for 480 min at 70 °C and pH 10.6 with Na2CO3 showed 99.3% dissolution of fluoride whereas at 95 °C and pH 10.6 83.3% dissolution was achieved (Figure 1). At higher temperature (95 °C ) the dissolution reaction is initially rapid and F − content in the solution reaches maximum after approximately 100 min. The dissolution rate is retarded after 100 min. As can be seen from Figure 1, similar result was observed when NaHCO3 was used and the pH was adjusted to 10.6 using 10 M NaOH. However, an equilibrium calculation based on experimental data indicates that complete dissolution of added amount of CaF2 is theoretically possible under the given experimental condition (Figure 1).  3.2.2. Dissolution at pH 9.0 Figure 2 shows the development of dissolution of CaF 2 over time at pH 9.0. With NaHCO 3 at 70 • C, the CaF 2 dissolution was 88.3% after 480 min whereas at 95 • C it was 96.7%. Similar results were obtained with Na 2 CO 3 at 95 • C when the pH was adjusted to pH 9.0 using HNO 3 ; at 70 • C the dissolution was slightly lower compared to that of NaHCO 3 . At low temperature and low alkaline pH the rate of the reaction is slower. At 70 • C, the % dissolution of fluoride increases with increase in time. Compared to at pH 10.6 the increase in initial % dissolution is less at pH 9.0, but the final % dissolution reached at pH 9 is 13% higher at 95 • C. Again, equilibrium calculation based on experimental data indicates that complete dissolution of added amount of CaF 2 is theoretically possible under the given experimental conditions (Figure 2).  Figure 2 shows the development of dissolution of CaF2 over time at pH 9.0. With NaHCO3 at 70 °C , the CaF2 dissolution was 88.3% after 480 min whereas at 95 °C it was 96.7%. Similar results were obtained with Na2CO3 at 95°C when the pH was adjusted to pH 9.0 using HNO3; at 70°C the dissolution was slightly lower compared to that of NaHCO3. At low temperature and low alkaline pH the rate of the reaction is slower. At 70 °C , the % dissolution of fluoride increases with increase in time. Compared to at pH 10.6 the increase in initial % dissolution is less at pH 9.0, but the final % dissolution reached at pH 9 is 13% higher at 95 °C . Again, equilibrium calculation based on experimental data indicates that complete dissolution of added amount of CaF2 is theoretically possible under the given experimental conditions (Figure 2).

Thermodynamic Calculations
In Table 1, thermodynamically calculated and experimentally found ionic concentrations (mM) for existing species, at conditions corresponding to the experimental ones, possibly influencing dissolution are compared. The major specie in the carbonate system at pH 9.0 is HCO3 − and at pH 10.6 it is CO3 2− . Thermodynamically calculated and experimentally determined concentrations of Na + and F − are comparable but Ca 2+ concentration differs three orders in magnitude (in the case of NaHCO3) and four orders in magnitude (in the case of Na2CO3). According to the literature [23,24], the solubility product of CaF2 does not differ significantly between 70 and 95 °C. Due to kinetic reasons, complete dissolution as predicted by equilibrium calculation is being hindered.

Mineral Phases in Residue by XRD
The diffractograms of the residues collected after 480 min dissolution are presented in Figure 3. Phases detected were different polymorphs of CaCO3 such as calcite, vaterite and aragonite in different proportions. When dissolution of CaF2 was done with Na2CO3 at 70 °C , calcite and vaterite were detected in the residue along with 1.5% fluorite (CaF2) ( Table 2). When the reaction temperature is higher (95°) 15% aragonite is also present along with 15.2% fluorite in the residue. According to the literature, aragonite formation is favored when the precipitation reaction is rapid [18] and at higher

Thermodynamic Calculations
In Table 1, thermodynamically calculated and experimentally found ionic concentrations (mM) for existing species, at conditions corresponding to the experimental ones, possibly influencing dissolution are compared. The major specie in the carbonate system at pH 9.0 is HCO 3 − and at pH 10.6 it is CO 3 2− . Thermodynamically calculated (Refer Figures A1 and A2) and experimentally determined concentrations of Na + and F − are comparable but Ca 2+ concentration differs three orders in magnitude (in the case of NaHCO 3 ) and four orders in magnitude (in the case of Na 2 CO 3 ). According to the literature [23,24], the solubility product of CaF 2 does not differ significantly between 70 and 95 • C. Due to kinetic reasons, complete dissolution as predicted by equilibrium calculation is being hindered.

Mineral Phases in Residue by XRD
The diffractograms of the residues collected after 480 min dissolution are presented in Figure 3. Phases detected were different polymorphs of CaCO 3 such as calcite, vaterite and aragonite in different proportions. When dissolution of CaF 2 was done with Na 2 CO 3 at 70 • C, calcite and vaterite were detected in the residue along with 1.5% fluorite (CaF 2 ) ( Table 2). When the reaction temperature is higher (95 • ) 15% aragonite is also present along with 15.2% fluorite in the residue. According to the literature, aragonite formation is favored when the precipitation reaction is rapid [18] and at higher pH [25]. At a temperature of 95 • , the dissolution of CaF 2 is initially rapid (Figure 1), indicating that precipitation of CaCO 3 may also be rapid. Aragonite formation at 95 • leads to more coprecipitation of fluoride compared to calcite [18], inhibiting complete dissolution of fluoride.
In the diffractograms of residues from dissolution with NaHCO 3 at pH 9.0 and temperature 70 and 95 • , calcite was found to be the dominating phase but aragonite ( Table 2) was also observed in the residue from dissolution at 95 • (Figure 3). Also, for this sample, the initial rate of the dissolution is rapid (Figure 2) and this leads to rapid precipitation of CaCO 3 . In the residue obtained from dissolution experiment with NaHCO 3 at pH 9.0 and 70 • C, no aragonite was found but 9.1% of CaF 2 was found.
Metals 2019, 9, x FOR PEER REVIEW 6 of 12 pH [25]. At a temperature of 95°, the dissolution of CaF2 is initially rapid (Figure 1), indicating that precipitation of CaCO3 may also be rapid. Aragonite formation at 95° leads to more coprecipitation of fluoride compared to calcite [18], inhibiting complete dissolution of fluoride.
In the diffractograms of residues from dissolution with NaHCO3 at pH 9.0 and temperature 70 and 95°, calcite was found to be the dominating phase but aragonite ( Table 2) was also observed in the residue from dissolution at 95° (Figure 3). Also, for this sample, the initial rate of the dissolution is rapid (Figure 2) and this leads to rapid precipitation of CaCO3. In the residue obtained from dissolution experiment with NaHCO3 at pH 9.0 and 70 °C, no aragonite was found but 9.1% of CaF2 was found. As can be seen in Table 2, the highest total amount of CaCO3 is present in the solid residue after dissolution with Na2CO3 at 70 °C and pH 10.6, CaCO3 is present as calcite and vaterite. Almost as much CaCO3, but mainly calcite, is formed with NaHCO3 dissolution done at 95 °C and pH 9.0. According to Jun Kawano et al. [26], the formation of different polymorphs or mixture of polymorphs is dependent on the precipitation temperature. Researchers A. Sarkar and S. Mahapatra [17] reported transformation of metastable vaterite to thermodynamically more stable aragonite or even more stable calcite depending on the conditions. Thermodynamically, the dissolution of vaterite leading to the formation of aragonite is possible, but this kind of transformation is rarely observed. There is a kinetic barrier for this transformation that may take place at elevated temperature when the activation energy is supplied. The highest amount of vaterite (Table 2) was formed when dissolution was done at pH 10.6 with Na2CO3 at 95 °C and in the same dissolution experiment the highest amounts of aragonite and fluorite were found. Initial rapid dissolution kinetics leads to rapid precipitation and, hence, the least thermodynamically stable vaterite was formed. Less thermodynamically stable crystals are formed when crystallization process is kinetically controlled. As can be seen in Table 2, the highest total amount of CaCO 3 is present in the solid residue after dissolution with Na 2 CO 3 at 70 • C and pH 10.6, CaCO 3 is present as calcite and vaterite. Almost as much CaCO 3 , but mainly calcite, is formed with NaHCO 3 dissolution done at 95 • C and pH 9.0. Table 2. Amount of present minerals based on residue analysis by XRD-Rietveld refinement and calculated considering total amount of residue generated. According to Jun Kawano et al. [26], the formation of different polymorphs or mixture of polymorphs is dependent on the precipitation temperature. Researchers A. Sarkar and S. Mahapatra [17] reported transformation of metastable vaterite to thermodynamically more stable aragonite or even more stable calcite depending on the conditions. Thermodynamically, the dissolution of vaterite leading to the formation of aragonite is possible, but this kind of transformation is rarely observed. There is a kinetic barrier for this transformation that may take place at elevated temperature when the activation energy is supplied. The highest amount of vaterite (Table 2) was formed when dissolution was done at pH 10.6 with Na 2 CO 3 at 95 • C and in the same dissolution experiment the highest amounts of aragonite and fluorite were found. Initial rapid dissolution kinetics leads to rapid precipitation and, hence, the least thermodynamically stable vaterite was formed. Less thermodynamically stable crystals are formed when crystallization process is kinetically controlled.

Residue (Dissolution Condition
Relative to the initial amount of CaF 2 added (300 mg/L), %CaF 2 dissolved was estimated based on F − content in the solution and %CaF 2 in the residue estimated via the Rietveld method applied on XRD data, both estimations are stated in Table 3. After 8 h of reaction the highest dissolution (Table 3) of CaF 2 was found at 70 • C and pH 10.6 with Na 2 CO 3 . Table 3. %CaF 2 dissolved based on %F in solution and present in residue, both calculated considering the feed weight (300 mg/L CaF 2 ).

SEM-EDS
SEM-EDS study on the residue obtained from the experiments with Na 2 CO 3 at 95 • C, pH 10.6 ( Figure 4) and with NaHCO 3 at 70 • C, pH 9.0 ( Figure 5) was performed, as the residues from these tests contain comparably high amount of CaF 2 and lesser amount of CaCO 3 . Both residues have different ratios of calcite, vaterite and aragonite in the total amount of CaCO 3 (Refer Table 2). Relative to the initial amount of CaF2 added (300 mg/L), %CaF2 dissolved was estimated based on F − content in the solution and %CaF2 in the residue estimated via the Rietveld method applied on XRD data, both estimations are stated in Table 3. After 8 h of reaction the highest dissolution (Table  3) of CaF2 was found at 70 °C and pH 10.6 with Na2CO3. Table 3. %CaF2 dissolved based on %F in solution and present in residue, both calculated considering the feed weight (300 mg/L CaF2).

SEM-EDS
SEM-EDS study on the residue obtained from the experiments with Na2CO3 at 95 °C , pH 10.6 ( Figure 4) and with NaHCO3 at 70 °C , pH 9.0 ( Figure 5) was performed, as the residues from these tests contain comparably high amount of CaF2 and lesser amount of CaCO3. Both residues have different ratios of calcite, vaterite and aragonite in the total amount of CaCO3 (Refer Table 2).  In Figure 4a, the residue sample mainly contains ~1 µ m particles of CaF2, significantly smaller than 3.37 µ m that was the mean diameter of the CaF2 feed. SEM-EDS analyses detected white particles  Relative to the initial amount of CaF2 added (300 mg/L), %CaF2 dissolved was estimated based on F − content in the solution and %CaF2 in the residue estimated via the Rietveld method applied on XRD data, both estimations are stated in Table 3. After 8 h of reaction the highest dissolution (Table  3) of CaF2 was found at 70 °C and pH 10.6 with Na2CO3.

SEM-EDS
SEM-EDS study on the residue obtained from the experiments with Na2CO3 at 95 °C , pH 10.6 ( Figure 4) and with NaHCO3 at 70 °C , pH 9.0 ( Figure 5) was performed, as the residues from these tests contain comparably high amount of CaF2 and lesser amount of CaCO3. Both residues have different ratios of calcite, vaterite and aragonite in the total amount of CaCO3 (Refer Table 2).  In Figure 4a, the residue sample mainly contains~1 µm particles of CaF 2 , significantly smaller than 3.37 µm that was the mean diameter of the CaF 2 feed. SEM-EDS analyses detected white particles and light gray particles as CaF 2 , while darker gray particles mainly consist of CaCO 3 . Spheroidal shaped grains [27] that dominate the residue are vaterite, cf. Figure 4a. Larger particles with sharp edges with a rhombohedral shape are calcite [27], c.f. Figure 4b. Needle-shaped crystals (Figure 4b) were also observed, which correspond to aragonite [27]. As previously mentioned, the dissolution result for this sample shows initially rapid dissolution and, hence, indicates rapid CaCO 3 precipitation, until a breakpoint when the fluoride content is unchanged, cf. Figure 1. Figure 5. shows images of the residue obtained from the experiment with NaHCO 3 at 70 • C, pH 9.0. The sample contains slightly larger particles, roughly 2-5 µm and, also, particles with sharp edges and a rhombohedral shape, similar to calcite [27], were quite commonly observed. Large particles of CaF 2 were found, indicating unreacted CaF 2 due to incomplete dissolution, and some of these seem to be coated with CaCO 3 , leading to passivation of CaF 2 during dissolution reaction. The results of the SEM investigation together with XRD findings indicate that it is not only the temperature and the reagents used but also the polymorphism property exhibited by CaCO 3 which has a significant effect on the dissolution behavior of fluoride.

Dissolution Chemistry of CaF 2 at Alkaline pH in Carbonate/Bicarbonate Media
In alkaline condition, CaF 2 will be dissolved, as the content of Ca 2+ in the solution is restricted by precipitation of CaCO 3 in the presence of CO 3 2− ions [28]. The content of Na + in the solution remains unchanged during the whole process, as Na + salts of F − , CO 3 2− /HCO 3 − are water soluble.
As dissolved Ca 2+ is removed, the further dissolution of CaF 2 is enhanced. In the aqueous system studied, co-existing equilibriums are: The temperature dependent solubility products for these equilibriums are: When the Ca 2+ content is lowered below the value corresponding to the solubility product of CaF 2 , dissolution can continue. Thus, dissolutions of CaCO 3 and CaF 2 are interdependent, as the equilibria determining their solubility involve concentration of Ca 2+ in both cases. By keeping the CO 3 2− content in the solution high enough, the solubility product for CaCO 3 will be exceeded before the one for CaF 2 is reached. As CO 3 2 ions are consumed, these can be generated again through the equilibrium with present HCO 3 2− ions in the solution.
Theoretically, the removal of Ca 2+ can be enhanced by achieving a higher CO 3 2− ion concentration at a higher level of pH; that is, the solubility of CaCO 3 [29] decreases with increased temperature as well as pH. Experimentally, it was found that for the dissolution of fluoride, a temperature of 70 • C is preferable at pH 10.6, while 95 • C is preferable at pH 9.0 (Figures 1 and 2). At pH 10.6 the whole carbonate system remains mainly as CO 3 2− , which facilitates rapid CaCO 3 formation. Furthermore, CaCO 3 precipitates rapidly at high temperature. The authors Kitano and Okumura showed that a high precipitation rate of CaCO 3 favors co-precipitation of CaF 2 [19]. At pH 10.6 and 95 • C, it is possible that rapid kinetics due to high temperature, high content of CO 3 2− and low solubility of CaCO 3 contributes to higher coprecipitation of CaF 2 [30]. While at pH 9 the CO 3 2− concentration is low and, hence, formation of CaCO 3 is also low. So at a higher temperature, less CaCO 3 precipitation happens and, hence, less coprecipitation of CaF 2 is found. During precipitation of CaCO 3 encapsulation of CaF 2 particles is also possible, which leads to passivation of the reaction, which was observed in SEM studies. In such conditions, total fluoride dissolution will be lower. According to Norio Wada et al., in systems of soluble calcium nitrate-sodium carbonate (Ca(NO 3 ) 2 -Na 2 CO 3 ) and soluble calcium chloride-sodium carbonate (CaCl 2 -Na 2 CO 3 ), aragonite precipitation is favored at high temperature and in the presence of divalent cations [31] such as Zn 2+ , Fe 2+ , Mg 2+ . Hence, during leaching of WO, DLWO and other zinc-containing dust material with Na 2 CO 3 /NaHCO 3 at higher temperatures, aragonite formation and co-precipitation of CaF 2 may occur. Complex interplay of both thermodynamic and kinetic factors affects CaCO 3 /CaF 2 equilibrium; to elucidate, it is influenced by factors such as temperature, retention time, pH, reagents used, stirring conditions, etc. This leads to polymorphism of CaCO 3 [16,17,26], which influences dissolution behavior of CaF 2 by either passivating it or co-precipitating it.

•
This experimental and thermodynamic dissolution study on CaF 2 showed that the dissolution chemistry is similar irrespective of the use of 150 mM of either Na 2 CO 3 or Na 2 HCO 3 as reagent.
The following conclusions are drawn from this work: o Highest dissolution of CaF2 at pH 9 is achieved at 95 • C. o Highest dissolution of CaF2 at pH 10.6 is achieved at 70 • C. o An initially strong positive effect on the dissolution of CaF2 is observed at a higher temperature. o According to the experimental results, the complete dissolution of CaF2 was not achieved in any of the dissolution conditions studied. o According to the thermodynamic calculations, complete dissolution of CaF2 is possible in all the dissolution conditions studied.
• XRD-Rietveld refinement and SEM-EDS investigations showed that a mixture of polymorphs of CaCO 3 , namely calcite, vaterite and aragonite, are formed in different ratios for different reaction conditions. o The type of polymorphs formed has an impact on the fluoride dissolution. o Higher temperature and higher alkaline pH cause rapid CaCO 3 precipitation, forming aragonite and higher coprecipitation of CaF 2 .
• In dissolution of CaF 2 to Ca 2+ and F − , control of aqueous Ca 2+ concentration by CO 3 2− concentration plays an important role. Consumption of Ca 2+ by CaCO 3 precipitation is affected by multiple factors including temperature and pH, which that influence the CO 3