1. Introduction
Nickel is widely used for the manufacturing of special steels, special alloys, batteries, and catalysts. Attempts have been made to extract nickel from ores and nickel-containing wastes using the chlorination method in which nickel oxide is converted into soluble or gaseous nickel chloride when roasted in the presence of chlorinating agents [
1,
2,
3,
4,
5,
6,
7,
8,
9,
10,
11,
12,
13,
14,
15].
Selective chlorination was carried out to recover valuable metals from spent catalysts in a series of studies by Gaballah et al. [
1,
2,
3]. The spent catalysts were roasted with Cl
2-N
2 [
1,
2,
3], Cl
2-O
2 [
1], Cl
2-CO [
1], Cl
2-air [
1,
2,
3], Cl
2-CO-N
2 [
2,
3] at a temperature between 473 K and 1173 K. Up to 98% of nickel could be recovered after roasting under the optimum conditions and following a subsequent water leaching.
Alvarez and Bohé [
4] investigated the direct chlorination of nickel-containing materials under an Ar-Cl
2 atmosphere. They found that the beginning temperatures of the chlorination reaction were 998 K, 746 K, and 701 K for NiO, the mixture of NiO and Al
2O
3, and the NiO-Al
2O
3 catalyst, respectively. The recovery percentages were 85% and 96% from the mixture and catalyst when roasted at 1073 K, respectively.
The beginning temperatures of nickel oxide chlorination were 573 K and 1023 K for chlorine and calcium chloride, respectively [
5]. The presence of active additives (C, BaS, S) shifted the equilibrium of the reaction toward the formation of nickel chloride. Similar trends were also obtained for the chlorination of nickel ferrite [
6] and nickel silicate [
7].
Selective chlorination was also adopted to extract nickel from reduced or unreduced laterite, followed by water or hot-acidulous water leaching [
8,
9,
10]. The chlorination agents included the mixtures of NaCl and MgCl
2·6H
2O (mass ratio 0.4) [
8], AlCl
3·6H
2O [
9], and gaseous HCl [
10]. The leaching percentage was 87% from laterite after roasting at 1173 K for 90 min with about 19 wt% of mixtures of NaCl and MgCl
2·6H
2O (mass ratio 0.4) as an addition [
8]. The contact chance between reactants and ores increases because the eutectic mixtures of NaCl and MgCl
2·6H
2O easily penetrate into tiny pore. A total of 91% of Ni was extracted from reduced limonitic laterite (goethite FeOOH and hematite Fe
2O
3) using water leaching at 353 K, after roasting at 733 K for 120 min with 40 wt% of AlCl
3·6H
2O in addition [
9]. To further understand the process, selective chlorination of pre-reduced limonitic laterite was conducted in a HCl-O
2-H
2O-N
2 atmosphere to investigate the kinetics and the effects of temperature, partial pressure of hydrogen chloride, oxygen, water vapor, and total gas-flow rate [
10].
Nickel oxide is converted into gaseous chloride at higher temperatures. A thermogravimetric analysis (TGA) technique was used to investigate the chlorination behaviors of nickel oxide in chlorine and hydrogen chloride [
11]. TGA curves showed that nickel oxide started to chloridize at 773 K and 673 K in a chlorine atmosphere and a hydrogen chloride atmosphere, respectively, and nickel chloride started to volatilize at 1073 K in both atmospheres. Fruehan and Martonik measured the rate of chlorination of NiO and NiFe
2O
4 with Cl
2 at 1073 to 1473 K and with HCl at 1073 to 1273 K diluted with He or Ar [
12]. The rates were affected by mass transfer for all cases. A high-temperature chlorination method was also used for metal recovery from roasted printed-circuit-board waste [
13]. The nickel’s volatilization percentage increased with the temperature in chlorine gas; they were both about 85% at 1073 K and 1173 K, respectively.
However, the use of gaseous chlorinating agents is limited due to their toxicity. This study aims to investigate the high-temperature chlorination of nickel oxide using calcium chloride as a chloridizing agent. Effects of different variables on nickel-volatilization percentage and kinetics are investigated, where variables include flow rate, temperature, molar ratio of CaCl2 to NiO, and oxygen partial pressure.
2. Materials and Methods
Reagent-grade NiO, SiO
2, Fe
2O
3, and anhydrous CaCl
2 were used in this study. Samples before roasting were prepared according to the composition shown in
Table 1. Effects of different variables on nickel-volatilization percentage were investigated with the compositions of Group A to Group C. The composition of Group D was used to study the high-temperature chlorination kinetics of nickel oxide using calcium chloride. The molar ratios of CaCl
2 to NiO for Group A, B, C, and D were 10, 7.5, 5, and 7.5, respectively, and those of SiO
2 to CaCl
2 were 4, 4, 4, and 2, respectively. After sufficient mixing, the mixtures were pressed into briquettes.
The experiment of the effects of different variables were carried out in a horizontal tube furnace. Carrier gas consisted of high-purity nitrogen and oxygen, whose flow rates were controlled with mass-flow controllers. An alumina boat (length: 60 mm, width: 30 mm) with about 10 g of briquettes was located on the center of quartz tube at about 873 K and heated to the desired temperature with a rate of 25 K/min. Samples were cooled inside the furnace to about 873 K after holding them at desired temperatures for 1 h and then cooling to room temperature with the protection of high-purity nitrogen.
The experiments of kinetics were conducted in a muffle furnace under a controlled atmosphere. High-purity nitrogen was injected into the hearth through the inlet below the thermocouple and escaped through the outlet at the top near the door. An alumina boat with about 10 g of briquettes was put at the center of the furnace hearth at desired temperatures. The sample was taken out immediately after holding for a certain time and then cooled to room temperature with the protection of high-purity nitrogen. Exhaust gas was discharged into the atmosphere after alkaline-solution treatment for all the experiments.
Nickel concentrations of samples were determined using an inductively coupled plasma optical emission spectrometer (ICP-OES, Optima 8000, PerkinElmer, Waltham, MA, USA) after dissolution with acid. The nickel-volatilization percentage was calculated using Equation (1).
where
wi and
wf are the sample weights before and after roasting, respectively.
ci and
cf are nickel concentrations before and after roasting, respectively. The collected volatile matters that condensed near the gas outlet were analyzed with X-ray diffraction (XRD, X’PertPro, PANalytical, Almelo, The Netherlands).
3. Results and Discussions
3.1. Chlorination Product
As shown in
Figure 1, NiCl
2 is the main phase of condensed volatile matter. The nickel concentration of condensed volatile matter is 45.12 wt%, which is very close to the nickel’s theoretical concentration of NiCl
2 (45.29 wt%). It means that NiCl
2 is the main volatilization product during the high-temperature chlorination of nickel oxide using calcium chloride.
3.2. Effect of Gas-Flow Rate
The change of nickel-volatilization percentage with carrier gas-flow rate is shown in
Figure 2. Nickel-volatilization percentage becomes larger as gas-flow rate increases from 0 mL/min to 200 mL/min, which could be ascribed to the diffusion strengthening of gaseous reaction products. And, it keeps at a near constant with a larger gas-flow rate when the molar ratios of CaCl
2 to NiO are 7.5 and 10, respectively. It decreases as gas-flow rate increases from 400 mL/min to 600 mL/min with the molar ratio of five CaCl
2 to NiO, which could be attributed to the loss of chloridizing agents (gaseous CaCl
2 or Cl
2 generated from the decomposition of CaCl
2). Therefore, it is suitable for all subsequent experiments with the gas flow rate of 400 mL/min to obtain their maximum volatilization percentages.
3.3. Effect of CaCl2 Dosage
Figure 3 shows the change of nickel-volatilization percentage with different molar ratios of CaCl
2 to NiO. It increases significantly from 53.4% to 90.9% as the molar ratio of CaCl
2 to NiO is strengthened from 5 to 10 at 1273 K. A similar trend is observed at 1373 K. It increases from 58.1% to 93.8% as the molar ratio of CaCl
2 to NiO increases from 5 to 10 at 1373 K.
The direct chlorination of nickel oxide in the presence of SiO
2 could be written as
While, the indirect chlorination of nickel oxide in the presence of SiO
2 could be written as
The chlorination of nickel oxide is promoted with greater CaCl2 dosage which accelerates the balance of high-temperature chlorination turning it to right in the case of direct chlorination. Meanwhile, a greater CaCl2 dosage means more Cl2 is released with the decomposition of CaCl2, and more nickel oxides are subsequently chlorinated in the case of indirect chlorination. Therefore, it is beneficial to obtain a larger nickel-volatilization percentage with a higher CaCl2 dosage.
3.4. Effect of Roasting Temperature
Figure 4 shows the change of nickel-volatilization percentage with roasting temperature. The volatilization percentages are enhanced with higher temperature. They increase from 31.4% to 58.1%, from 39.4% to 84.8%, and from 40.4% to 93.7% as the roasting temperature increases from 1173 K to 1373 K when the molar ratios of CaCl
2 to NiO are 5, 7.5 and 10, respectively.
Figure 5 shows the XRD results of calcines from Group A after roasting with the oxygen partial pressure of 0.2. There are newly generated CaSiO
3 and CaFe
2O
4 and unreacted SiO
2 and Fe
2O
3 in the NiO-SiO
2-Fe
2O
3-CaCl
2 system at the investigated temperature range with a molar ratio of 10 CaCl
2 to NiO and a molar ratio of 4 SiO
2 to CaCl
2.
It is inevitable that part of the CaCl
2 evaporates from the NiO-SiO
2-Fe
2O
3-CaCl
2 system at temperatures between 1173 K and 1373 K which are above the melting point of CaCl
2 [
16]. The reaction could be written as
Solid or liquid NiCl
2 (melting point 1304 K) [
17] is converted into a gaseous state at a higher temperature, which could be written as
The standard Gibbs free-energy changes of Equations (2)–(6) are calculated from the thermochemical data [
17] and are shown in
Figure 6.
decreases significantly as the temperature increases and becomes closer to zero producing liquid CaCl
2. Reaction (2) cannot happen in the standard state due to the positive value of
. However, Δ
G2 could decrease and turn from positive to negative by decreasing the partial pressures of gaseous NiCl
2 according to Equation (7). In this case, Reaction (2) can possibly occur. It is almost unchanged and significantly below zero in the investigated temperature range producing gaseous CaCl
2. This indicates that a higher temperature could promote the direct chlorination of nickel oxide.
decreases slightly, and all the values are less than 21.5 kJ/mol producing liquid CaCl
2. Similarly, Reaction (3) can possibly occur when Δ
G3 turns from positive to negative by decreasing the partial pressures of Cl
2 according to Equation (8). Although it becomes significantly larger, all the values are much lower than zero producing gaseous CaCl
2. Therefore, Reaction (3) could happen in the investigated temperature range.
turns from positive to negative as temperature increases, which suggests that the chlorination of nickel oxide with Cl
2 generated from the decomposition of CaCl
2 is enhanced. Therefore, a higher temperature could improve the indirect chlorination of nickel oxide.
On the other hand,
decreases significantly and
turns from positive to negative as the temperature increases. This suggests that CaCl
2 and NiCl
2 tend to evaporate from the NiO-SiO
2-Fe
2O
3-CaCl
2 system at a higher temperature. Compared with liquid CaCl
2, Reactions (2) and (3) are more achievable with gaseous CaCl
2, as shown in
Figure 6. Reactions (2) and (4) are promoted with the continuous gaseous NiCl
2 volatilization from the system which breaks the reaction balances. The saturated vapor pressures of CaCl
2 and NiCl
2 are calculated and shown in
Figure 7. They increase significantly with a rise of temperature, which leads to the enhancement of volatilization percentage as it increases the roasting temperature.
3.5. Effect of Oxygen Partial Pressure of Carrier Gas
Nickel-volatilization percentage is significantly influenced by oxygen partial pressure. As shown in
Figure 8, nickel-volatilization percentages decrease from 84.3% to 53.4% as increasing the oxygen partial pressure from near zero to 0.2 at 1273 K with the molar ratio of five CaCl
2 to NiO. And, the percentages decrease from 96.9% to 76.1% and from 99.7% to 90.9% at 1273 K with the molar ratios of CaCl
2 to NiO of 7.5 and 10, respectively. The magnitude of volatilization percentage reduction caused by increasing the oxygen partial pressure becomes narrower with a larger molar ratio of CaCl
2 to NiO.
Assuming that the activities of CaCl
2, NiO, SiO
2, and CaSiO
3 are taken together and Cl
2 is equal to 10
−6, the Gibbs free-energy changes of reactions with different oxygen partial pressure at 1273 K are calculated according to Equations (7)–(9) and shown in
Figure 9. It should be noted that only the case of liquid CaCl
2 is considered. It could be seen that the direct high-temperature chlorination of nickel oxide is not affected by oxygen partial pressure, but influenced by the NiCl
2 partial pressure. Reaction (3) is promoted as increasing oxygen partial pressure, and Reaction (4) is inhibited with both greater oxygen and NiCl
2 partial pressure.
The equilibrated chlorine partial pressure of Reactions (3) and (4) at 1273 K could be calculated using Equations (10) and (11), respectively, and shown in
Figure 10, assuming that activities of CaCl
2, NiO, SiO
2, and CaSiO
3 are taken together. It could be seen that the equilibrated chlorine partial pressure calculated using Equation (10) is larger than that using Equation (11) at the same oxygen partial pressure with a smaller NiCl
2 partial pressure only considering liquid CaCl
2. This indicates that Cl
2 generated with Reaction (3) could meet the chlorine-partial-pressure requirement of Reaction (4). Therefore, the indirect chlorination happens. The chlorine-partial-pressure demand of Reaction (4) enhances as NiCl
2 partial pressure increases from 10
−6 to 10
−1 and gets close to the corresponding equilibrated chlorine partial pressure calculated with Equation (10). Indirect chlorination is inhibited in this case no matter the oxygen partial pressure.
To sum up, when oxygen partial pressure is close to zero, Reaction (3) does not happen and there is only direct high-temperature chlorination. As the oxygen partial pressure increases, Reaction (3) starts to occur. However, NiCl2 partial pressure keeps at a high level as a result of the direct high-temperature chlorination, which leads to the inhibition of Reaction (4). Cl2 generated using Reaction (3) could not meet the chlorine partial pressure requirement of Reaction (4), and part of CaCl2 is consumed and the volatilization percentage decreases. A further increase of oxygen partial pressure leads to a rising chlorine partial pressure which meets the demand of Reaction (4). The indirect high-temperature chlorination happens. When oxygen partial pressure is large enough, Reaction (4) is inhibited. In this condition, the generation rate of Cl2 using Reaction (3) increases gradually, and becomes greater than the chlorination rate of nickel oxide. Excess Cl2 escapes from the system leading to the lower utilization efficiency of CaCl2. Therefore, nickel-volatilization percentage is negatively correlated to oxygen partial pressure. With a larger molar ratio of CaCl2 to NiO, there is still enough CaCl2 used for the direct high-temperature chlorination, leading to a narrower magnitude of volatilization-percentage reduction caused by oxygen partial pressure.
3.6. High-Temperature-Chlorination Kinetics
The reaction process of the high-temperature chlorination of nickel oxide using calcium chloride is extremely complicated. In the case of direct chlorination, the process mainly consists of the phase transformation of CaCl2 from liquid to gas phase, the chlorination reaction between nickel oxide and liquid or gaseous CaCl2, and the volatilization of NiCl2. Indirect chlorination is unavoidable due to the existence of trace amounts of oxygen. The side-reaction process mainly includes the phase transformation of CaCl2 from liquid to gas phase, the reaction between trace amounts of oxygen and liquid or gaseous CaCl2 to give off Cl2, the chlorination reaction between nickel oxide and Cl2, and the volatilization of NiCl2. The macro-kinetics of high-temperature chlorination of nickel oxide using calcium chloride is established according to the unreacted shrinking-core model.
For the chemical-reaction control model, the relationship between nickel-volatilization percentage and time could be expressed as [
18]
For the diffusion control model, the relationship could be written as [
18]
where X is the nickel-volatilization percentage; k is the apparent reaction rate constant; and t is time.
Figure 11 shows the change of nickel-volatilization percentage with time. According to Equation (12), a change of [1 − (1 − X)
1/3] with time is obtained and shown in
Figure 12. Similarly, a change of [1 − 3(1 − X)
2/3+2(1 − X)] with time is also obtained according to Equation (13) and shown in
Figure 13. The apparent reaction-rate constants at different temperatures are calculated to the straight slopes in
Figure 12 and
Figure 13 for the chemical-reaction control model and diffusion control model, respectively, and are shown in
Figure 14. According to the Arrhenius law, the apparent activation energies are calculated as 142.91 kJ/mol and 250.22 kJ/mol for the chemical-reaction control model and diffusion control model, respectively.
The activation energies from 423 to 473 K and from 473 to 673 K were 46.1 and 12.4 kJ/mol for the chlorination of nickel oxide with gaseous hydrogen chloride, respectively [
14]. And, it was 12.4 kJ/mol for the chlorination of nickel-containing lateritic iron ore with gaseous hydrogen chloride [
15]. The values suggest that the chlorination of nickel oxide with gaseous hydrogen chloride was controlled by the diffusion of gaseous reactant [
14,
15]. The chlorination of hematite with Cl
2 is probably diffusion controlled between 873 K and 1148 K with the activation energy of 74 kJ/mol [
19].
The activation energy was determined to be 119 kJ/mol, which confirmed the assumption that the chlorination of NiFe
2O
4 with Cl
2 between 973 K and 1173 K was chemically controlled [
6]. The apparent activation energies of the PbSO
4 chlorination and carbochlorination were about 174 kJ/mol and 114 kJ/mol, respectively, which suggested that both the rates of chlorination and carbochlorination were probably controlled by the chemical reaction [
20].
Based on the above analysis, the activation energy of diffusion control is lower than that of chemical-reaction control. Therefore, such values of apparent activation energies suggest that it is more likely controlled by the chemical reaction of high-temperature chlorination of nickel oxide using calcium chloride.
4. Conclusions
The conclusions obtained from this study are listed as follows:
(1) Nickel-volatilization percentage increases as gas-flow rates increase from 0 to 200 mL/min, and then is kept almost at a constant with a higher gas-flow rate and greater CaCl2 dosage.
(2) Nickel-volatilization percentage is positively correlated to temperature from 1173 K to 1373 K and to CaCl2 dosage.
(3) Nickel-volatilization percentage is negatively correlated to oxygen partial pressure.
(4) The high-temperature chlorination of nickel oxide using calcium chloride is controlled by a chemical reaction. The apparent activation energy is 142.91 kJ/mol and between 1173 K and 1323 K.