Fundamental study on utilization of tin and zinc-bearing iron concentrate by selective chlorination

JIANG Tao(½ª ÌÎ), ZHANG Yuan-bo(ÕÅÔª²¨), HUANG Zhu-cheng(»ÆÖù³É),

LI Guang-hui(Àî¹â»Ô), GUO Yu-feng(¹ùÓî·å),YANG Yong-bin(ÑîÓÀ±ó), JIN Yong-shi(½ðÓÂÊ¿)

(School of Resources Processing and Bioengineering, Central South University,Changsha 410083, China)

Abstract: The feasibility and technologies of comprehensive recovery of tin, zinc, arsenic and iron from the complex iron ores by selective chlorination roasting were studied by thermodynamic analysis and roasting experiments. Investigation shows that the product pellets with the compression strength of 2625N/P, the tumble index of 97.26%, the abrasion index of 1.35%, tin, arsenic and zinc residue of 0.043%, 0.046% and 0.058% respectively can be achieved if balling a concentrate containing 0.39% tin, 0.40% arsenic and 0.28% with addition of 8% coke breeze and 0.5% CaCl₂ and roasting the pellets at 1060-1080¡æ for 40min. The volatilization of tin, arsenic and zinc is 91.75%, 93.42% and 81.12% respectively. The performances of the product pellets are able to meet the requirements of blast furnace ironmaking.

Key words: selective chlorination; comprehensive recovery; tin; zinc; complex iron ore CLC number: TF046

Document code: A

1 INTRODUCTION

China is rich in iron ore resources, but most of these ores are low-grade, polymetallic refractory ones^{[1, 2]}. With the rapid development of iron and steel industry, the comprehensive utilization of these ores is becoming more and more important. The complex iron ores containing tin and zinc are typical intractable ones, great reserves of which are found in Inner Mongolia, Guangdong, Guangxi and Hunan of China. Huanggang Iron Mine of Inner Mongolia is a large-size complex iron ore deposite associated with nonferrous elements of tin, arsenic, zinc and so on^{[3, 4]}, among which the reserve of iron is above 1.1¡Á10⁸t, and the reserves of tin, arsenic and zinc are respectively about 4¡Á10⁵, 5.6¡Á10⁵ and 2.5¡Á10⁵t. It is very difficult to efficiently recover and utilize iron, tin, arsenic and zinc from these ores because of the complex association of these elements. The literatures indicate that a lot of research has been done since 1970, but the methods are mainly limited to reduction roasting to produce metallic pellets^[5]. Furthermore, these methods with high production costs and low efficiency lack systematical research and have not been put into commercial production. Therefore, these ores have not been efficiently utilized hitherto.

Because of the low melting point and high volatility of most metal chlorides, chlorination volatilization has been adopted in extractive metallurgy. The metal elements accompanied are easy to be transferred into their chlorides, selectively volatilized into gaseous phase and recovered in the dust collecting system^[6-8]. In this investigation, a new process of tin and zinc recovery and preparation of pellets for blast furnace from tin and zinc containing iron concentrate by chlorination roasting is developed.

2 EXPERIMENTAL

2.1 Characteristic of material

The iron concentrate containing tin and zinc used in this investigation is taken from Inner Mongolia Autonomous Region of China. The mass percent of particles below 0.075mm of the concentrate is found to be 61.16%, and the chemical composition is listed in Table 1.

It can be seen from Table 1 that the iron con- centrate belongs to magnetite type^[9]. The contents of tin and zinc are 0.39% and 0.28% respectively, each of which exceeds the limitation of ironmaking^[10]. And the contents of SiO₂ and Al₂O₃ are 3.35% and 0.70% respectively. Moreover, there is 0.40% of arsenic in the sample, which needs to be removed or recovered.

Table 1  Chemical composition of iron concentrate (mass fraction, %)

Optical microscope (OM) and scanning electronic microscope (SEM) observations(Fig.1) show that tin in the concentrate mostly occurs as SnO₂ in fine intergrowth or inclusion and a little in monomer, and partial SnO₂ is adherent to magnetite or encapsulated by silicate and fluorite. And arsenic in the concentrate almost occurs as FeAsS₂, and closely adheres to magnetite. The primary existence minerals of zinc are sphalerite and calamine(including a little of zinc oxide), part of which are embedded in the gangue alone and metasomaticly intergrow with magnetite and gangue. The other metallic minerals hardly contain zinc.

Fig.1  Existence forms of tin, arsenic and zinc in concentrates

   In this investigation coke breeze is used as internal reductants. The mass percent of coke particles below 0.15mm and 0.075mm is 100% and about 80% respectively, and the industrial analyses of coke breeze are listed in Table 2.

Table 2  Industrial analysis of coke breeze (mass fraction, %)

   Fuel ratio (the ratio of fixed carbon to percent volatile) is an important criterion of coal combustibility, which is closely correlated to the coal reactivity^[11]. The greater the fuel ratio is, the worse the combustibility and reactivity are, and the more obvious the lingering combustion is. From Table 2, it can be calculated that the fuel ratio of coke breeze is 57.52, which indicates that the combustibility and reactivity of coke breeze are inferior, so [BJ(,,,][BJ)] Vol.15 ¡í.4       Utilization of tin and zinc-bearing iron concentrate it can offer durative combustion and feeble reduction atmosphere.

2.2 Research methods

The experimental process includes mixing, balling, drying, preheating, roasting and so on. The green ball is prepared in the disc pelletizer of 1m in diameter. The bench-scale experiments of preheating and roasting are carried out in a vertical electrically heated tube furnace with an internal and external diameter of 80mm and 100mm respectively. The mixture of N₂ and CO₂, with a ratio of 2¡Ã1, and the flow-rate of 10L/min and 5L/min respectively, is led into from the bottom of the furnace. The experimental flowsheet is shown in Fig.2.

3 CHLORINATION THERMODYNAMICS OF IRON, TIN, ARSENIC AND ZINC

Thermodynamic analysis indicates that the standard formation free enthalpy of FeO and SnO₂ is approachable, and the reduction balance curves of SnO₂(s) and SnO(g) are also proximate^[12]. Therefore, it is difficult to realize the effective volatilization of tin in the form of SnO(g) by selective reduction under the condition of the only existence of carbon. But effective separation of Sn from Fe becomes easy in the coexistence of carbon and chlorides. On one hand, carbon is easily compatible with O₂, so the partial pressure produced by chlorination reactions is decreased; and on the other hand, SnO₂ is easily inverted into SnO(g) and Sn(l) in the presence of carbon^[8]. Therefore, there will be the following chlorination reactions:

SnO₂(s)+2Cl₂+2C=SnCl₄(g)+2CO(1)

¦¤G^¦È/J=-160900-232.06T

SnO(g)+Cl₂=SnCl₂(g)+1/2O₂(2)

¦¤G^¦È/J=38284-75.88T

Sn(l)+Cl₂=SnCl₂(g)(3)

Fig.2  Experimental flowsheet

  ¦¤G^«ß/J=-251458+30.21T

¦¤G^«ß for the above three reactions is below zero when the temperature is between 800¡æ and 1200¡æ. It is obvious that SnO₂(s), SnO(g) and Sn(l) will easily form SnCl₄(g) or SnCl₂(g) and can be volatilized from the ores.

For arsenic, the following reaction will take place at first during chlorination roasting:

4FeAsS₂+14O₂=2Fe₂O₃+As₄O₆+8SO₂(4)

As₄O₆ exists steadily between 25¡æ and 900¡æ, and will be decomposed if the temperature is higher^[8]:

As₄O₆=2As₂O₃(5)

As₂O₃ can be chlorinated into AsCl₃, but As₂O₃ has a higher steam pressure because of a low boiling point (about 460¡æ) at the chlorination roasting temperature. Therefore, once As₂O₃ is formed, it will be volatilized into gaseous phase in which a little of it may be chlorinated into AsCl₃. So, arsenic is mostly volatilized in the form of As₂O₃.

As far as ZnO and FeO are concerned, there will be the following reactions in the presence of chlorides:

ZnO(l)+Cl₂=ZnCl₂(g)+1/2O₂(6)

¦¤G^¦È/J=-800-72.8T

FeO(l)+Cl₂=FeCl₂(g)+1/2 O₂(7)

¦¤G^¦È/J=-41800+20.0T

But seen from the ¦¤G^«ß¡ªT chloride curves of MeO-Cl₂^[12-14], in the three curves of FeO-Cl₂, ZnO-Cl₂ and SnO-Cl₂, the curve of FeO-Cl₂ lies in the upper side, ZnO-Cl₂ in the middle, and the curve of SnO-Cl₂ in the lower side. It shows FeO(l) is most difficult to be chlorinated among the three oxides, and SnO(g) is easy to be transferred into SnCl₂. Therefore, if only the proper chloride content and given volatile conditions are controlled, tin, arsenic and zinc can be chlorinated and volatilized step by step, and iron will not be chlorinated and remain in the roasted pellets.

Practical chlorination agent is calcium chloride, and through the activation of gangues, such as SiO₂ and Al₂O₃, CaCl₂ reacts with O₂ and H₂O to form Cl₂ or HCl in the following chlorination reactions:

CaCl₂+SiO₂+1/2O₂=CaO¡¤SiO₂+Cl₂(8)

¦¤G^¦È/J = 65960-34.89T

CaCl₂+Al₂O₃+1/2O₂=CaO¡¤Al₂O₃+Cl₂(9)

¦¤G^¦È/J = 140400-56.23T

CaCl₂+SiO₂+H₂O=CaO¡¤SiO₂+2HCl(10)

¦¤G^¦È/J =175200-103.55T

CaCl₂+Al₂O₃+H₂O=CaO¡¤Al₂O₃+2HCl(11)

¦¤G^¦È/J =199700-124.89T

Compared with no existence of SiO₂ and Al₂O₃ gangues, the trend of oxidation and hydrolyzation of CaCl₂ is apparently increased and the reaction equilibrium constant multiples 10⁴ in the existence of the two gangues^{[15, 16]}. Furthermore, the melting point of CaCl₂ is only 772¡æ, and CaO generated in the oxidation and hydrolyzation reactions of CaCl₂ is easy to dissolve in the melting CaCl₂, so the activity of CaO is reduced. Thus, CaCl₂ actually acts an accelerator of the formation of CaO¡¤SiO₂ or CaO¡¤Al₂O₃ and the chloride volatile reactions can proceed.

4 EXPERIMENTAL RESULTS AND DISCUSSION

In batch-scale experiments, the mixture of N₂ and CO₂ is led into the roasting furnace during whole roasting in order that the internal reductants in the pellets react with CO₂ at high temperature and are gasified into CO to maintain feeble reduction atmosphere for the roasting procedure. In this test, coke breeze is used as an internal reduction agent.

4.1 Effect of reductant dosage

The test conditions are fixed as follows: the roasting temperature 1050¡æ, time 60min, and CaCl₂ dosage 4%. The effect of internal coke breeze dosage on the roasting results is studied and illustrated in Fig.3.

Fig.3  Effect of coke dosage on roasting results

It can be seen from Fig.3 that the compression strength of pellets presents a downward trend, but all of them are above 2100N/P with the increase of coke breeze dosage from 4% to 8%. When the dosage is increased to 10%, the compression decreases obviously. The possible reason is that there will be more pores formed inside the pellets after roasting with increasing dosage of internal coke breeze, which results in a looser structure.

Fig.3 also shows that tin volatilization, which remains above 90%, is nearly unchanged and the remains of tin are below 0.06% when the dosage of coke breeze varies from 4% to 10%. It can be concluded that compounds of tin can be easily chlorinated and volatilized effectively in the coexistence of coke breeze and CaCl₂. Arsenic volatilization is increased continuously with the increase of coke dosage. The volatilization of arsenic is 90.03% and the remaining arsenic is below 0.08% when the coke dosage is 8%. Furthermore, zinc volatilization is also enhanced with the increase of coke breeze dosage. Volatilization of zinc is also increased with the increase of coke breeze dosage. When the coke breeze dosage is in excess of 8%, zinc volatilization is above 82%, and the remains of zinc is below 0.06%.

4.2 Effect of CaCl₂ dosage

The test conditions are as follows: roasting at 1050¡æ for 60min, and 8% internal coke breeze dosage. Effects of the dosage of CaCl₂ are shown in Fig.4.

Fig.4  Effect of dosage of CaCl₂ on roasting results

It is illustrated in Fig.4 that the compression strength of roasted pellets is reduced gradually with the increase of CaCl₂. The possible reason is that some low-melting-point compounds(such as CaO¡¤SiO₂) are formed. It also can be seen that the volatilizations of tin and zinc are enhanced with the increase of CaCl₂ dosage. When the dosage of CaCl₂ is below 2%, the volatilizations of tin and zinc increase markedly with the dosage. And when the dosage exceeds 2%, the volatilization of tin keeps almost unchanged and remains above 90%. As far as arsenic is concerned, the change of CaCl₂ dosage hardly has effect on arsenic volatilization, which is between 89.4% and 90.4% when the dosage of CaCl₂ is from 1% to 4%. However, the volatilization of zinc still keeps increasing.

4.3 Effect of roasting temperature

The test conditions are: the roasting time of 40min, coke breeze dosage of 8%, and CaCl₂ dosage of 2%. Effects of roasting temperature on the roasting results are shown in Fig.5.

Fig.5  Effect of roasting temperature on roasting results

From Fig.5, it can be concluded that roasting temperature has great effect on the results. With the rising of roasting temperature, the compressive strength keeps increasing. But the compression strength exceeds 2000N/P only when the temperature gets to 1100¡æ. When the temperature reaches 750¡æ, tin has been volatilized notably. When the temperature exceeds 850¡æ, over 80% tin has been volatilized, and the remains of tin in the pellets are below 0.08%. When the temperature gets to 950¡æ, the volatilization increases slowly. Volatilization of arsenic is kept above 85.83% and changes a little with the temperature when it surpasses 850¡æ. The volatilization of zinc is continuously increased with the rising of temperature(750-1100¡æ). When the pellets are roasted at 1100¡æ, zinc volatilization is 82.55% with the remains of 0.058% zinc in the finished pellets. Therefore, the suitable roasting temperature should be between 1050¡æ and 1100¡æ.

4.4 Effect of roasting time

Effect of roasting time on the roasted results is shown in Fig.6. Experimental conditions are kept at 8% coke breeze, 2% CaCl₂ and 1100¡æ roasting temperature.

It can be seen from Fig.6, the compression of the roasted pellets and the volatilization of tin and zinc are continuously increased with the increase of roasting time. When time is over 30min, the compression strength is greater than 2100N/P. When roasting time increases from 10min to 20min, the volatilizations of tin, arsenic and zinc increase quickly and about 90% of tin, 85% of arsenic and 80% of zinc are volatilized into gaseous phase in 20min. With the further increase of roasting time, the volatilization is increased slowly. When the time is above 40min, remaining contents of tin, arsenic and zinc are all under 0.08%.

Fig.6  Effect of roasting time on roasting results

   Seen from the results and discussion, the suitable roasting time is between 30 and 40min.

4.5 Dynamic roasting experiment

Previous tests are conducted in a static furnace. The dynamic experiment is performed in an intermittent rotary kiln with a diameter of 1m. The experimental conditions are that the pellets with coke breeze (8%) and CaCl₂ (0.5% and 2%) are roasted for 40min and the temperature is controlled between 1060¡æ and 1080¡æ. The performances of the product pellets are presented in Table 3.

The results in Table 3 indicate that the compression strength of the product pellets is greater than 2500N/P, the tumble index is higher than 97%, and the contents of tin, arsenic and zinc in the pellets are all less than 0.06%. All the indexes of the pellets meet the requirement of ironmaking in blast furnaces. It is found that the dosage of CaCl₂ can be greatly reduced to 0.5% from 2%, which may be due to the improved dynamic conditions for tin, arsenic and zinc volatilization in the rotary kiln.

Table 3  Performances of product pellets by roasting in rotary kiln

   Since tin, arsenic and zinc are easily volatilized into gaseous phase, they can be selectively recovered in the dust collecting system. As well known, the traditional chlorination processes often use over 4% chlorinating agent and anticorrosive process has been successfully applied in the whole production. This process, only using 0.5% CaCl₂, possesses a promising perspective.

5 CONCLUSIONS

1) SEM analysis shows that the symbiosis of valuable metals in tin, zinc-bearing iron concentrates are so complex that it is difficult to effectively recover them by beneficiation and traditional sintering and pelletizing process.

2) The thermodynamic analyses show that tin, arsenic, zinc and iron can be separated from one another by selective chlorination. Tin, arsenic and zinc can be chlorinated and volatilized selectively and iron will remain in the roasted product by properly controlling chloride dosage and roasting conditions.

3) The static bench-scale experiments show that appropriate conditions of comprehensive recovery of tin, arsenic and zinc containing iron concentrates are balling the concentrate with 8% coke breeze and 2% CaCl₂, and roasting at 1050-1100¡æ for 40min.

4) The dynamic rotary kiln experiments show that the dosage of CaCl₂ can be reduced to 0.5%, and the product pellets with the compression strength of 2625N/P, the tumble index of 97.57%, the abrasion index of 1.24%, tin, arsenic and zinc residues of 0.045%, 0.056% and 0.060% are achieved. The performances of the pellets meet the requirements of ironmaking.

REFERENCES

[1]ZOU Jian. Status of China¬ðs iron ore resource and measures to ensure continuous supply of home produced iron ore[J]. Mining Engineering, 2003, 2(1): 1-4. (in Chinese)

[2]LI Pei-liang, XIE Ying-liang. How to effectively utilize foreign iron ore resources[J]. Conservation and Utilization of Mineral Resources, 2003, 4(2): 13-16. (in Chinese)

[3]XIE Chang-jiang. Study on the separation-smelting process of skarn Fe-Sn ores[J]. Hunan Nonferrous Metals, 1996, 11(6): 13-17. (in Chinese)

[4]WANG Li-juan, WANG Jing-bin, WANG Yu-wang, et al. REE geochemistry of the Huanggangliang skarn Fe-Sn deposit of Inner Mongolia[J]. Acta Petrologica Sinica, 2002, 18(4): 575-584. (in Chinese)

[5]CHEN Yao-ming, XU Jing-cang, JIANG Tao, et al. Study on pelletizing direct reduction process of high Sn, Zn, As content iron ore concentrate[J]. Sintering and Pelletizing, 1997, 5(3): 17-20. (in Chinese)

[6]CHEN Xin-min. Physical Chemistry of Pyrometallurgical Process[M]. Beijing: Metallurgical Industry Press, 1984. 471-476. (in Chinese)

[7]Fletcher A W, Jackson D V, Valentine AG. Chloride route for the recovery of tin metal from low-grade concentrates[J]. Institute of Mining & Metallurgy Tarnsactions. Section C: Mineral Processing & Extractive Metallurgy, 1967, 76(730): 145-153.

[8]HUANG Wei-sen. Tin[M]. Beijing: Metallurgical Industry Press, 2001. 346-348. (in Chinese)

[9]FU Ju-ying, JIANG Tao, ZHU De-qing. Sintering and Pelletizing [M]. Changsha: Central South University of Technology Press, 1995. 15. (in Chinese)

[10]Beneficiation Design Manual Editorial Committee. Beneficiation Design Manual [M]. Beijing: Metallurgical Industry Press, 1988. 26-27. (in Chinese)

[11]XIE Ke-chang. Coal Structure and Its Reactivity [M]. Beijing: Science Press, 2002. 50.

[12]CHEN Guo-fa. Heavy Metal Metallurgy[M]. Beijing: Metallurgical Industry Press, 1992. 224-226. (in Chinese)

[13]Kubaschewski O, Alcock C B. Metallurgical thermochemistry(5th Rev. ed) [M]. Oxford: Pergamon Press, 1979. 449-450.

[14]Chris C, Kirk C Y, Donald W. Behavior of metals under the conditions of roasting MSW incinerator fly ash with chlorinating agents [J]. Journal of Hazardous Materials, 1999, 1(1): 75-89.

[15]Lythe R G, Prosser A P. Catalytic effects in the chlorination of cassiterite [J]. Institution of Mining & Metallurgy Transactions. Section C: Mineral Processing & Extractive Metallurgy, 1969, 78(757): 206-213.

[16]Tskhai G F, Lebedev B N, Buyanov V I, et al. Decomposition of calcium chloride with the separation of chlorine during heating [J]. Chemical Abstracts, 1998, 70(22): 98872d.

(Edited by YANG Bing)

Received date: 2004-12-10; Accepted date:2005-03-08

Correspondence: JIANG Tao, PhD; Tel: +86-731-8830542; E-mail: jiangtao@mail.csu.edu.cn.