One-step extraction of bismuth from bismuthinite in sodium carbonate–sodium chloride molten salt using ferric oxide as sulfur-fixing agent

Abstract

In view of serious shortcomings existing in the current bismuthinite metallurgical process, such as high energy consumption, low recovery rate of bismuth and low-concentrate SO₂ fume pollution, a clean process was proposed, which involves molten salt smelting of bismuthinite concentrate. Using Na₂CO₃–NaCl molten salt as reaction medium and Fe₂O₃ as sulfur-fixing agent, crude bismuth could be produced directly with high recovery—more than 97%—and high bismuth grade—around 90%. To identify the reaction mechanism, thermodynamic calculations and mechanistic experiments were performed and the results indicated that the reducing and sulfur-fixing reactions of Bi₂S₃ occurred easily and that sulfur was fixed mainly in the form of FeS. To regulate the bismuth extraction process, the effects of a range of parameters during the bismuthinite smelting were comprehensively investigated, including smelting temperature and time, dosage and composition of molten salts, dosage of coal powder. Confirmation experiments were also carried out and 79.63% of lead and 95.00% of silver were collected in the crude bismuth. This study is beneficial for the further optimisation of the bismuth glance smelting process and its potential application to other sulphide ore extractions.

---

1. Introduction

Bismuth is an important strategic metal because it is non-toxic and harmless in its performance and is widely used in low melting point alloy casting, the electronics industry, medicine, the chemical industry and other materials fields.^1–3 In 2014, bismuth consumption in the world was reported to be 15 000 t, which is approximately equal to the total yield. The major ores for production of bismuth are bismuthinite (Bi₂S₃) and bismite (Bi₂O₃).⁴ Other secondary sources, such as lead anode slime, bismuth slags, and copper converter dusts are also important bismuth-containing materials.^5–7

At present, the extraction processes for bismuth can be divided into two methods: pyrometallurgy and hydrometallurgy.⁸ The pyrometallurgical methods, including reduction smelting, bath smelting and mixed smelting, using coal powder as the reducing agent with scrap iron and soda addition, are operated at temperatures of 1200–1300 °C, which renders them high energy consumption (5.8 × 10¹⁰ J t⁻¹) processes.⁹ However, it is reported that both bismuth and its oxides are easy to volatilise and both exhibit significant vapour pressure above 930 °C.¹⁰ In addition, toxic arsenic is often associated with bismuth ores due to its similar properties and arsenic is also volatile at high temperatures.^11,12 Furthermore, sulfur contained in bismuth sulfide concentrate also escapes to the atmosphere in the form of low-concentration SO₂, resulting in air pollution.¹³ Smelting at high temperature will, therefore, result in environmental pollution, hazards and the loss of bismuth.

Hydrometallurgical processes can be used for both low-grade and complex bismuth ores. Normal industrial bismuth hydrometallurgical practices comprise two steps: leaching and preparation. The first, leaching process using FeCl₃, Cl₂ and HCl as oxidant and leaching agent,^14–16 then a variety of products can be produced form purification of solution by replacement, electrowinning and hydrolysis precipitation.^17,18 However, the serious shortcomings of the hydrometallurgical methods, such as tedious and complicated processes and low current efficiency, are fatal for industrial production.¹⁹

Alkaline smelting was first proposed by the former Soviet Union and was first used for lead extracting from its sulphide concentrate in molten NaOH;²⁰ this process featured a low smelting temperature and cleaning production. Subsequently, Margulis employed molten NaOH salt to recycle lead from lead metallic scrape.²¹ Molten salts have excellent thermal stability, low viscosity and solvent properties, which have led to their use in materials preparation,²² solar power plants²³ and other engineering fields.²⁴ Recently, Yang adapted an alkaline smelting process to smelt antimony and bismuth using the mixed NaOH–Na₂CO₃ molten salts.²⁵

Molten NaOH, however, has a strong activity at a high temperature, which makes it easy to react with sulphide, gangue and refractory bricks, resulting in difficulties in NaOH regeneration and brasque consumption.²⁶ To address this issue, we propose instead, NaCO₃–NaCl as the molten salt system to replace the NaOH system in the reaction. NaCO₃–NaCl exhibits alkalescence and has a low-temperature eutectic composition. In addition, ferric oxide is used as a desulphurisation agent, which provides the outstanding advantages of eliminating SO₂ discharge while utilising the large iron slag produced in the sulfuric acid industry. The present study investigated the effects of the main variables contributing to the yield of bismuth and considered the relevant reaction mechanism. Compared to the current process, the proposed approach is low-carbon and cleaning and can produce bismuth directly by one-step smelting.

2. Experimental

2.1. Materials

Bismuthinite concentrate was obtained from Hunan Shizhuyuan Nonferrous Metals Co. Ltd. (Hunan province, China) and the content of bismuth is 26.22%. Its chemical composition and XRD analysis results are given in Table 1 and Fig. 1, respectively. XRD study shows that the main composition occurs in the form of Bi₂S₃, MoS₂ and FeS₂. The coal powder used in this research has 84.14% fixed-carbon and the chemical composition is shown in Table 2. All other reagents used in the experiment, including Fe₂O₃, Na₂CO₃ and NaCl, were of analytical grade.

Table 1 Main chemical components of bismuthinite concentrate (wt%)

Bi	S	Fe	Ca	Pb	Si	Cu	Mo	Ag^a
a g t^−1.
26.22	20.15	14.94	5.53	2.75	5.54	3.6	1.75	300

---

Fig. 1 XRD pattern of bismuthinite concentrate.

Table 2 Chemical components of coal powder (wt%)

Materials	C	CaO	MgO	Al2O3	SiO2	TFe	S
Content	84.14	0.60	0.16	5.17	7.21	0.80	0.10

---

2.2. Process flow and procedure

In the smelting experiments, the bismuth concentration was maintained at 200 g and the amount of mixed salt and coal powder added were expressed as multiples and percentages of the ore content. The amount of Fe₂O₃ required to convert the sulfur released from bismuth concentrate to the form of FeS was calculated based on stoichiometry. After mixing to improve homogeneity, the mixture was charged in an alumina crucible. When the electric furnace (Yuandong Furnace Co. Ltd., Hunan province, China) reached the setting temperature, the crucible was located in the hot zone. After the predetermined reaction time, the crucible was quickly taken out and quenched to room temperature. Bismuth metal could be easily separated at the bottom of the quenched crucible and the content of Bi in the crude bismuth was determined by titrimetric analysis. The valuable metals content in the crude bismuth was also determined by inductively coupled plasma (ICP).

X-ray diffraction (XRD) studies were performed using a Rigaku D/max 2550VB + 18 kW powder diffractometer with the scanning rate of 8° per minute. The TG-DSG analysis was performed using a thermo-gravimetric analysis of Universal V4.0C TA instrument with SDT Q600 V8.0 Build 95 in an argon flow of 0.1 L min⁻¹ and at a heating rate of 10 °C per min and the dosage of mixture was 15 mg every time. The direct recovery rate of bismuth was calculated using the following equation:

where W₁, W₂ are the mass of bismuthinite concentrate and crude bismuth, and x₁ and x₂ are the content of Bi in the raw ore and crude bismuth, respectively.

3. Thermodynamic analysis and reaction process

3.1. Thermodynamic analysis

Direct smelting of bismuthinite concentrate was conducted with the addition of a reductant and a sulfur-fixing agent in NaCO₃–NaCl molten salt at a low temperature of 500–1000 °C. The principal products were crude bismuth and sulfur fixed in the form of FeS. The main chemical reactions can be described as follows:

Fe₂O₃ + C = 2FeO + CO(g) (2)

Bi₂S₃ + 3FeO + 3C = 2Bi + 3FeS + 3CO(g) (3)

Bi₂S₃ + 3FeO + 3CO(g) = 2Bi + 3FeS + 3CO₂(g) (4)

Bi₂S₃ + 3Na₂CO₃ + 6C = 2Bi + 3Na₂S + 9CO(g) (5)

Bi₂S₃ + 3Na₂CO₃ + 3CO(g) = 2Bi + 3Na₂S + 6CO₂(g) (6)

FeO + Na₂S + CO₂(g) = FeS + Na₂CO₃ (7)

Eqn (8) represents the Gibbs free energy function. By calculating ΔG^θ_T at different temperatures, ΔG^θ_T values for all the abovementioned reactions can be obtained. If ΔG^θ_T < 0, then the relevant reaction can be considered to have spontaneously occurred. Conversely, if ΔG^θ_T > 0, then the reaction cannot occur.

where ΔH^θ_T and ΔH^θ₂₉₈ are the standard enthalpy change at T (K) and 298 K, respectively; ΔS^θ_T and ΔS^θ₂₉₈ are the standard entropy change at T (K) and 298 K, respectively; ΔC_p is the heat capacity change and T is the temperature in Kelvin. All the data are from the thermodynamic handbook.²⁷ Fig. 2 gives the effect of temperature on the ΔG^θ_T of reactions (2)–(7).

Fig. 2 ΔG^θ–T graphs of reactions (2)–(7).

It can be observed from the Fig. 2 that it is feasible for reactions (3)–(5) and (8) to occur because the ΔG^θ_T of these equations are all negative when the smelting temperature is maintained between 500 and 1000 °C. Thus, Bi₂S₃ can easily react with FeO and carbon to produce metallic Bi. Furthermore, Na₂S can be regenerated into Na₂CO₃ spontaneously via reaction (7). However, the ΔG^θ_T values of reactions (5) and (6) are greater than zero when the reaction temperature is below 670 °C and 598 °C, respectively, and therefore, the two reactions are difficult to progress. Therefore, the overall reactions to extract Bi from Bi₂S₃ contained in bismuthinite can be denoted as reactions (3) and (4).

3.2. Reaction process of Bi₂S₃–Fe₂O₃–C system

The results of TG/DTA analysis of the Bi₂S₃–Fe₂O₃–C system at the mole ratio of 2 : 3 : 9 are shown in Fig. 3. According to the TG curves, the mass of the mixture slowly decreased ranging from 120 °C to 480.93 °C, which can be interpreted as the volatilization of Bi₂S₃.²⁸ Furthermore, the mass loss was rapid. The DTA curve appears to have an endothermic peak at 586.72 °C, which is caused by the reaction between C and Fe₂O₃.²⁹ The second endothermic peak at 617.72 °C is ascribed to the melting of Bi₂S₃.³⁰ A strong endothermic peak at 930–1000 °C is ascribed to the reaction between Bi₂S₃, Fe₂O₃ and C.

Fig. 3 Thermal behaviour of the Bi₂S₃–Fe₂O₃–C system under argon atmosphere.

Fig. 4 shows the XRD patterns of the products of the Bi₂S₃–Fe₂O₃–C system of mole ratio of 2 : 3 : 9 at different reaction times and temperatures. Fig. 4(a) shows that the peaks of Bi and Fe_1−xS emerged after 15 min at 850 °C, which indicates that the reaction proceeded quickly at this temperature. Fig. 4(b) shows that Fe₃O₄ appeared at 570 °C caused by the reaction between C and Fe₂O₃ and this temperature corresponds to the endothermic peak at 586.72 °C on the DTA curve. The peaks of Bi and Fe_1−xS emerged at 750 °C, but the reaction was inadequately completed at the temperature of 850 °C. However, when the temperature was further increased to 1000 °C, only Bi, Fe₃O₄ and Fe_1−xS can be observed from the XRD, which may indicate the reduction and sulfur-fixing reactions that have finished sufficiently.

Fig. 4 XRD patterns of the products in the Bi₂S₃–Fe₂O₃–C system at (a) different reaction times and (b) different temperatures.

4. Results and discussion

4.1. Effect of temperature

The effect of temperature on the direct recovery rate and grade of crude bismuth was investigated in the range from 750 °C to 950 °C, at the condition of W_NaCl = 30% W_salt, W_{molten salt}/W_bismuthinite = 3/1, 1.0 times stoichiometric Fe₂O₃, W_coal = 20% W_bismuthinite and smelting time 2 h. The relationship between the direct recovery rate and grade of crude Bi and the temperature are shown in Fig. 5.

Fig. 5 Effect of temperature on the direct recovery rate and grade of crude Bi (1.0 times stoichiometric Fe₂O₃, W_{molten salt}/W_bismuthinite = 3/1, W_coal = 20% W_bismuthinite, 2 h, W_NaCl = 30% W_salt).

As shown in Fig. 5, the direct recovery rate of bismuth increases quickly with the increasing of temperature from 750 °C to 850 °C, up to a maximum of 93.56%, which can be explained in two ways. First, the flow ability of the molten system increases as the temperature increases; consequently, the contact among reactants was sufficient and the smelting reactions could be carried out thoroughly. Moreover, a high temperature promotes the kinetics of the reaction. The direct recovery rate of bismuth keeps increasing until the profile becomes flat at 850 °C. However, the direct recovery rate of bismuth gradually decreases as the temperature increases. This response can be interpreted by more Fe₂O₃ being reduced to Fe with increasing temperature. In addition, high temperature accelerates the volatilisation of Bi and molten salt. Thus, the optimum temperature for smelting is 850 °C.

4.2. Effect of the amount of salt

Fig. 6 shows the effect of the amount of molten salt on bismuth recovery rate and its grade. The figure indicates that the direct recovery rate and grade of bismuth continuously increase as the W_bismuthinite/W_{molten salt} changes from 1/1.5 to 1/3 and the maximum values reach 94.28% and 92.47%, respectively, at 1/3; a small decrease was observed after that. This phenomenon can be interpreted by the fact that increasing the amount of salt correspondingly increases the flow ability and decreases the viscosity of the melt. As a result, the delamination and clarification of bismuth was beneficial, otherwise the bismuth would be lost due to physical inclusion. Excessive salt is, however, equivalent to diluting the concentration of the reactant, resulting in reduced productivity. Therefore, the suitable W_bismuthinite/W_{molten salt} ratio is 1/3.

Fig. 6 Effect of the amount of salt on the direct recovery rate and grade of bismuth (1.0 times stoichiometric Fe₂O₃, 850 °C, W_coal = 20% W_bismuthinite, 2 h, W_NaCl = 30% W_salt).

4.3. Effect of the amount of coal powder

Fig. 7 shows the effect of coal powder dosage on the bismuth extraction and its grade. As can be observed from the figure, the direct recovery rate of Bi increases from 85.9% to 95.18% when the amount of coal powder increases from 5% to 10%. Above 15%, both recovery rate of Bi and grade of crude Bi remain nearly constant. A strong reducing atmosphere caused by the coal powder dosage increasing was conducive to moving the reaction towards to the right to enable the smelting process to be completed. In addition, other impurity metals will be reduced and result in the bismuth grade decreasing and the carbon consumption will be increased with the excess coal. Therefore, the appropriate dosage of coal powder is found to be 10% of the bismuthinite mass.

Fig. 7 Effect of the amount of coal powder on the direct recovery rate and grade of bismuth (1.0 times stoichiometric Fe₂O₃, 850 °C, W_{molten salt}/W_bismuthinite = 3/1, 2 h, W_NaCl = 30% W_salt).

4.4. Effect of smelting time

Fig. 8 shows the effect of smelting time on bismuthinite smelting result. The direct recovery rate of bismuth reaches the greatest value of 95.68% within 1.5 h. Beyond 1.5 h, prolonging the smelting time results in no significant increase in recovery rate of Bi because the reactions were completed quickly through kinetics at the high temperature. However, the grade of crude bismuth slowly decreases from 96.58% to 92.13% when the time increases from 0.5 h to 1.5 h. It can be suggested that increasing time is beneficial for other metals, including Pb and Ag, to be collected in the crude bismuth. However, prolonging the time results in volatilization of Bi and higher energy consumption. Therefore, the suitable smelting time appears to be 1.5 h.

Fig. 8 Effect of smelting time on the direct recovery rate and grade of bismuth (1.0 times stoichiometric Fe₂O₃, 850 °C, W_{molten salt}/W_bismuthinite = 3/1, W_coal = 10% W_bismuthinite, W_NaCl = 30% W_salt).

4.5. Effect of the composition of the salt

The effect of the composition of salt on the bismuth extraction and its grade is presented in Fig. 9. The direct recovery rate of Bi increases from 91.08% to 97.35% as the W_NaCl/W_salt ratio increases from 10% to 20%. This can be interpreted as follows: the viscosity of the molten salt decreases with increasing NaCl contribution because NaCl has a lower viscosity than Na₂CO₃. Furthermore, the melting point of molten salt can be decreased and the flow ability of the melt can be improved as NaCl increases in a certain range; as a result, the delamination and clarification of Bi was beneficial. However, the direct recovery rate of metallic bismuth decreased to 94.15% at 50% NaCl. It can be explained that decreasing of Na₂CO₃ proportion weakens the reaction between Na₂CO₃ and Bi₂S₃. The grade of crude bismuth reserved is nearly constant though the amount of NaCl was increasing. Therefore, the optimum composition of salt appears to be W_NaCl/W_salt of 20%.

Fig. 9 Effect of composition of salt on the direct recovery rate and grade of bismuth (1.0 times stoichiometric Fe₂O₃, 850 °C, W_{molten salt}/W_bismuthinite = 3/1, W_coal = 10% W_bismuthinite, 1.5 h).

4.6. Confirmation experiment

Based on the single factor experiments, the optimum conditions for the low-temperature smelting process to extract bismuth from bismuthinite in Na₂CO₃–NaCl molten salt are presented in Table 3. These conditions were used in confirmation experiments for three times to smelt bismuthinite using 200 g of ore each time. The average direct recovery rate and grade of crude Bi can be calculated from the results listed in Table 4 as 97.35% and 91.34%, respectively, while only 0.59% of the bismuth was contained in the final slag.

Table 3 The optimal parameters for the molten smelting process

Condition	T (°C)	Wmolten/Wbismuthinite	Wcoal/Wbismuthinite	t (h)	WNaCl/Wsalt
Parameters	850	3/1	10%	1.5	20%

---

Table 4 Confirmation experiment results (%)

No.	Bi grade	Bi in slag	Bi recovery rate
1	91.81	0.50	97.71
2	90.74	0.81	96.22
3	91.46	0.47	98.13
Average	91.34	0.59	97.35

---

An image and the XRD results of the smelting products are shown in Fig. 10. It can be observed from the figure that the cooled melt was separated to two clear layers, which from the XRD analysis, were salt slag and crude bismuth. However, it can be observed from Fig. 10(B) that the composition of salt slag located in different layers is different, for example, the Na₂CO₃ content in the top slag is higher than in the bottom slag. Sulfur was detected mainly in the form of FeMo₄S₆ and CuFe₂S₃; the lack of apparent FeS peak is mainly because of its low relative content in the salt slag.

Fig. 10 (A) Image and (B) XRD patterns of the products.

The XRD pattern of the water leaching residue of salt slag is shown in Fig. 11. It can be deduced from the figure that the salt and soluble metal ions are dissolved in the leaching solution, which can be comprehensively recovered in a further process. Insoluble components in salt slag are collected in the residue. However, FeS can be detected because the relative content is increased in the leaching residue, but a part of the FeS was oxidized to Fe₂(SO₃)₄ in the water leaching process.

Fig. 11 XRD pattern of water leaching residue.

The chemical composition of the crude Bi is presented in Table 5. As shown in the table, the grade of Bi reached is 91.81% and the major impurity is Pb. By calculation, 79.63% of Pb and 95% of Ag are gathered in the crude bismuth, which can be recycled in further refining processes.

Table 5 Chemical composition of crude bismuth (wt%)

Bi	Fe	S	Pb	Mo	Ag^a	Cu
a g t^−1.
91.81	0.003	0.002	7.84	0.001	285	0.14

---

Comparison of some key indices between the low-temperature molten salt smelting and the current pyrometallurgical smelting process is made in Table 6. In the molten salt smelting process, low-concentrate SO₂ pollution is minimized because sulfur is fixed in the form of FeS. On the other hand, reducing the smelting temperature achieves two main advantages. First, energy consumption and the requirement for refractory materials are reduced. Second, the volatilization of hazardous compounds, such as arsenic and lead, is less at low temperature. Finally, new processes can directly extract bismuth from bismuthinite by a one-step smelting with a high recovery rate.

Table 6 Comparison between (a) the process of molten salt smelting at low temperature and (b) the current smelting process

Process	Bi grade (%)	Bi recovery rate (%)	Bi in slag (%)	Smelting temperature (°C)	The forms of sulfur
(a)	>91	>97	0.5	<1000	FeS
(b)	<90	<90	>1.0	>1200	Low-concentrate SO2

---

5. Conclusions

Molten salt smelting of bismuthinite concentrate to extract Bi at low temperature (700–1000 °C) was developed. Research into the reaction behaviours of the Bi₂S₃–Fe₂O₃–C system indicated that the main products are Bi and Fe_1−xS, and reaction can complete sufficiently at 850 °C after 30 min. The smelting parameters, such as reaction temperature and time, the dosage and composition of the molten salt and the dosage of coal powder, had a significant effect on the recovery of bismuth and the grade of crude bismuth. The optimum conditions were established by single experiments and were as follows: temperature of 850 °C, time of 1.5 h, W_salt/W_ore = 3/1, W_NaCl = 20% W_salt, and W_carbon = 10% W_ore. Under the optimum conditions, the direct recovery rate of Bi and grade of crude Bi were 97.71% and 91.81%, respectively. High comprehensive recovery of Pb and Ag were also demonstrated in the experiments. The crucial advantages of this new process are the high recovery rate of Bi, the decrease of the smelting temperature and reduced SO₂ emission, making this new molten salt smelting a cleaner and low-carbon process for bismuth extraction.

Acknowledgements

This study was financially supported by the National Natural Science Foundation of China (No. 51104128, No. 51234009) and we gratefully acknowledge many helpful comments and suggestions from anonymous reviewers.

Notes and references

1. H. Wu and J. Li, Hunan Nonferrous Met., 2006, 22(01), 25–27.
2. L. Cesarino, C. Gouve-Caridade and R. Pauliukaite, et al., Electroanalysis, 2010, 22(13), 1437–1445.
3. F. K. Ojebuoboh, JOM: Journal of the Minerals, metals, and Materials Society, 1992, 44(4), 46–49.
4. L. G. Wang, Electrochem. Ind., 1986, 14–15.
5. B. L. Liu, H. J. Wei and B. W. Wang, Multipurp. Util. Miner. Resour., 2013, 02, 48–50.
6. B. Zhang, Q. Li and W. Q. Shen, et al., Rare Met., 2012, 31(1), 102–106.
7. T. K. Ha, B. H. Kwon and K. S. Park, et al., Sep. Purif. Technol., 2015, 142, 116–122.
8. F. Habashi and S. Mridha, Encyclopedia of Materials: Science and Technology, 2nd edn, 2001, pp. 332–336.
9. R. X. Liu and H. Wu, Trans. Nonferrous Met. Soc. China, 1996, 6(1), 44–46.
10. C. M. B. Hincapié, M. J. P. Cárdenas and J. E. A. Orjuela, et al., Dyna, 2012, 79(176), 139–148.
11. Y. Chen, T. Liao and G. B. Li, et al., Miner. Eng., 2012, 39, 23–38.
12. W. Levason, G. M. Reid and J. A. M. J. Meyer, Comprehensive Coordination Chemistry II, Pergamon, Oxford, 2003, pp. 465–544.
13. S. Q. Li, J. He and M. T. Tang, Nonferrous Metals, 2003, 03, 39–40.
14. D. Kim and S. Wang, Can. Metall. Q., 2008, 47(3), 317–326.
15. L. E. Derlyukova, M. V. Vinokurrova and T. A. Anufrieva, Russ. J. Inorg. Chem., 2008, 53(9), 1371–1375.
16. G. Q. Zheng and M. T. Tang, Trans. Nonferrous Met. Soc. China, 2002, 5, 987–991.
17. C. Y. Wang and D. F. Qiu, Nonferrous Metals, 2001, 53(4), 15–18.
18. Y. P. Li, C. Y. Liu and S. H. Wu, et al., Hydrometallurgy, 2007, 26(3), 118–122.
19. C. Wang, D. Qiu and P. Jiang, Nonferrous Metals, 2001, 53(4), 15–18.
20. M. P. Smimov, Tsvetn. Metall., 1990, 5, 34–36.
21. E. V. Margulis, Erzmetall, 2000, 53, 85–89.
22. G. Chen, D. J. Fray and T. W. Farthing, Nature, 2000, 407, 361–364.
23. D. Zhou, C. Zhao and Y. Tian, Appl. Energy, 2012, 92, 593–605.
24. Y. Long, J. Li, D. H. Timmer, R. E. Jones and M. A. Gonzalez, J. Cleaner Prod., 2013, 68, 243–251.
25. J. G. Yang, C. B. Tang, Y. M. Chen and M. T. Tang, Metall. Mater. Trans. B, 2011, 42, 30–36.
26. X. Y. Guo, J. X. Liu and Q. H. Tian, et al., Nonferrous Met. Sci. Eng., 2013, 8–13.
27. I. Brain, O. Knacke and O. Kubaschewski, Thermochemical properties of inorganic substances, Springer-Verlag, New York. 1972.
28. V. Piacenie, V. D. Gianfreda and G. Bardi, J. Chem. Thermodyn., 1983, 15(7), 7–14.
29. H. G. Li, Principles of Metallurgy, Science Press, Changsha, Hunan, China, 2005, pp. 192–198.
30. J. Krüger, P. Winkler, M. Lück and H. U. Wolf, Ullmann's Encyclopedia of Industrial Chemistry, 2003.

---