Trans. Nonferrous Met. Soc. China 23(2013) 1501-1505

Mechanism of debismuthizing with calcium and magnesium

Dian-kun LU, Zhe-nan JIN, Yong-feng CHANG, Shu-chen SUN

School of Materials and Metallurgy, Northeastern University, Shenyang 110819, China

Received 9 April 2012; accepted 2 July 2012

Abstract: In order to further understand the debismuthizing mechanism with calcium and magnesium, the influence of adding amount of calcium on the bismuth removal from lead was examined. A part of the debismuthizing dross samples were studied by electron probe microanalysis (EPMA). The results show that the calculated bismuthide governing debismuthizing process is Ca₃Mg₇Bi₈, which may be a mixture of Ca₃Bi₂, CaBi, CaBi₃ and Mg₃Bi₂. And the bismuthide formed during the debismuthizing process exists in two states: one is free bismuthide in the matrix of dross, and the other is symbiotic with Pb-Ca phase.

Key words: debismuthizing mechanism; Kroll-Betterton process; Ca₃Mg₇Bi₈; Pb-Ca phase; Bi-Ca-Mg phase

1 Introduction

Most of the alkali and alkaline earth metals can form intermetallic compound with bismuth, and this sort of reactions were often used for the research work on the debismuthizing of lead bullion [1-3]. The Kroll- Betterton (K-B) process with calcium and magnesium reagents is the only practical process in operation for pyrometallurgical bismuth removal [4,5]. There still have been arguments on the mechanism of debismuthizing, and the arguments mostly focused on the compounds which act on bismuth removal.

On the mechanism of bismuth removal from lead by the K-B process, BETTERTON and LEBEDEFF [6] suggested that it resulted from the formation of both Ca₃Bi₂ and Mg₃Bi₂ compounds according to the microscopic study of the debismuthizing dross. While EVERS [7] believed that it resulted from the formation of CaMg₂Bi₂. The latter opinion has been accepted later by a majority of researchers [8-11]. According to this view, DAVEY [12] summarized the thermodynamic relationships that govern the removal of bismuth from lead by the K-B process. The removal of bismuth from lead is governed by the reaction equations (1) and (2):

2Mg+Ca+2Bi=Mg₂CaBi₂                       (1)

K_sp=[Mg]²[Ca][Bi]²                          (2)

where K_sp is the solubility product of the magnesium, calcium and bismuth in lead bullion, and it should be a constant at the same temperature. But the values of this solubility product obtained by many researchers changed in a very wide range [9-12]. The difference among solubility product values of CaMg₂Bi₂ obtained by different works indicates that there may be some unknown reasons affecting the debismuthizing results. Apart from these problems, the insufficiency of direct examination on the dross causes the debismuthizing mechanism lack of direct evidence to support.

This study focuses on ascertaining the mechanism governing debismuthizing with calcium and magnesium, especially on ascertaining the bismuthide and its occurence in debismuthizing dross. In addition to revealing debismuthizing mechanism, this research can also play a significant role in understanding the influence of bismuth on the performance of Pb-Ca alloy, which is a very important material in the storage battery industry.

2 Experimental

By addition of bismuth (99.99% purity) into a steel pot containing about 15 kg molten lead (99.99% purity) at 450 °C, the lead-bismuth alloy (0.023% Bi) was obtained. Calcium was added as lead-calcium alloy (1.36% Ca) and magnesium (99.99% purity) as magnesium piece.

Five experiments were conducted in a sealed vertical tubular crucible electric furnace under the protection of a nitrogen atmosphere. For each test, 500 g of prepared Pb-Bi alloy was used, and the initial Mg concentration was held at 0.155%, the initial Ca concentration varied between 0.040% and 0.135%. The temperature for bismuth removal was set at 330 °C, and the experimental procedure was the same as that in our previous work [13]. After the separation of lead bullion and dross, the alloys were weighed and sampled for ICP-AES analysis, and four of the dross samples were sent to electron microscopic and electron probe microanalysis. The results are shown in Table 1.

3 Results and discussion

The dross samples were taken from tests No. 2-5. The results show that all the samples are composed of matrix, Pb-Ca phase and Bi-Ca-Mg phase. The matrix and Pb-Ca phase account for about 60% and 30% of the dross, respectively. Bi-Ca-Mg phase is very little in the dross.

3.1 Analysis of matrix of dross

The composition of the matrix obtained by electron probe microanalysis is shown in Table 2. The backscattered electron image of the matrix and Pb-Ca phase is shown in Fig. 1, and the X-ray surface distribution images of Fe, Ca and Pb can be seen in Figs. 2(a-c), respectively.

Table 2 shows that the matrix is basically constituted by lead and iron. A large amount of iron in the matrix may come from the corrosion of stainless net by liquid lead alloy [14] during the separation of dross samples. Bi is enriched in matrix compared with the final lead (see Table 1). Through the average relative mass of Bi, Ca and Mg in matrix, the mass ratios of Ca and Mg to Bi were calculated as 0.243 and 0.234, respectively. They are all higher than 0.096 and 0.116, which are the mass ratios of Ca and Mg to Bi in CaMg₂Bi₂ compound, respectively. The surplus Ca and Mg may exist as Ca₃Mg₄Pb_x [12].

In Fig. 1 it can be found that the distribution of Pb-Ca phase in the dross is uneven. This uneven distribution is strict in accordance with the surface distribution image of Ca (see Fig. 2(a)), which forms Pb-Ca phase. Pb-Ca phase has a relatively larger grain size, generally larger than 20 μm, and takes the shape of round, strips or branch-like in the matrix. The boundary between the Pb-Ca phase and the matrix is smooth. Fe in the dross is also unevenly distributed (see Fig. 2(b), and is complementary with the distribution areas of Ca in the X-ray surface distribution images, which means that the Fe contaminated in the dross is corroded by lead bullion matrix, but not by Pb-Ca phase.

3.2 Analysis of Pb-Ca phase in dross

In contrast to the composition of matrix (see Table 2), the content of Ca in Pb-Ca phase (see Table 3) is well above that in the matrix. The mass ratio of Ca to Pb in Pb-Ca phase is about 5.71%, which is a little lower than that of compound CaPb₃ (6.45%). So, there must be a little lead left in Pb-Ca phase under the present dross separation conditions, and Pb-Ca phase can be taken as a solid solution of CaPb₃ and final lead alloy. Bi content in Pb-Ca phase is also higher than that in the matrix. But there is still some Bi in the matrix. This means that a part of bismuthide floats out of the lead before the formation of Pb-Ca phase during the falling of temperature. So, the compounds governing debismuthizing process consist of both free bismuthide and symbiotic bismuthide with Pb-Ca phase. Fe content in the Pb-Ca phase is well below that in the matrix. This also indicates that the stainless steel net is corroded by the final lead alloy, but not by Pb-Ca phase.

Fig. 1 Backscattered electron image of matrix and Pb-Ca phase (bright part)

Table 1 Initial conditions and analysis of final lead alloy in calcium dependent tests

Table 2 Composition of matrix in debismuthizing dross samples

Fig. 2 X-ray surface distribution images of Ca (a), Fe (b) and Pb (c) in dross (bright part)

Table 3 Composition of Pb-Ca phase in dross samples

3.3 Analysis of Bi-Ca-Mg phase in dross

The Bi-Ca-Mg phase generally has a little grain size, the majority of which is about 10 μm, and only a little reaches to 20 μm (see Fig. 3). The X-ray surface distribution images of Ca, Mg and Bi are very similar (see Figs. 4(a-c), which means that elements Bi, Ca and Mg are really enriched in the dross to form Bi-Ca-Mg compounds.

Fig. 3 Backscattered electron image of matrix and Bi-Ca-Mg phase

Apart from these mentioned above, it can be seen that this Bi-Ca-Mg phase is circulated by Pb-Ca phase (see Fig. 3). So, the Bi-Ca-Mg phase should be formed together with Pb-Ca phase. Then the reactivity of the Bi-Ca-Mg phase may decrease and the Bi content in final lead alloy also decreases.

We know that there is a peritectic point in Pb-Ca phase diagram. When the peritectic reaction occurs, a part of Pb-Ca phases (see Fig. 5) will be consumed and some Bi-Ca-Mg phases will be exposed to liquid lead. This will result in a bad debismuthizing effect due to the redissolution of some Bi-Ca-Mg phases. And it leads to the disturbance of Bi content in final lead according to the difference of Pb-Ca phase consumption (see Table 1). This disturbance will cause the change of the solubility product of the Bi-Ca-Mg compound.

Fig. 4 X-ray surface distribution images of Bi (a), Mg (b) and Ca (c) (bright part)

The composition of Bi-Ca-Mg phase is determined as 63.312% Bi, 1.900% Pb, 6.580% Mg, 4.580% Ca, 0.128% Fe, 23.280% O. A large amount of oxygen in this phase shows that the dross is significantly oxidized during the dross separation. If the contents of O, Pb and Fe are ignored, the compositions of the simplified Bi-Ca-Mg phase and CaMg₂Bi₂ compound can be compared (see Table 4).

Fig. 5 Pb-Ca phase diagram and Pb-rich region [13]

Table 4 Compositions comparison between of Bi-Ca-Mg phase and CaMg₂Bi₂ compound

Table 4 shows that the composition of Bi-Ca-Mg phase deviates from the stoichiometric coefficient of CaMg₂Bi₂ compound. Molecular formula of Ca_0.76Mg_1.79Bi₂ can be derived from the composition of simplified Bi-Ca-Mg phase. When the mole number in Ca_0.76Mg_1.79Bi₂ is changed into round figure, Ca₃Mg₇Bi₈ can be obtained. The total amount of Ca and Mg in Ca₃Mg₇Bi₈ is not enough to maintain the ratio of Bi to Ca and Mg for both Ca₃Bi₂ and Mg₃Bi₂. This demonstrates that the compound governing debismuthizing process is not produced by the union of Ca₃Bi₂ and Mg₃Bi₂. So, we cannot conclude that CaMg₂Bi₂ forms during removing bismuth from lead with calcium and magnesium. If we intend to take Ca₃Mg₇Bi₈ taken apart into three CaMg₂Bi₂, one MgBi₂ will be left aside. But there is no MgBi₂ compound in Bi-Mg system [15], so the Bi-Ca-Mg phase contains excessive Bi but lacks of Ca. Ca can form a series of compounds with Bi, such as Ca₃Bi₂, CaBi and CaBi₃ [16]. This means that when the Ca content is not enough to form Ca₃Bi₂, the other two compounds, CaBi and CaBi₃ may yield. So, we may view the calculated compound Ca₃Mg₇Bi₈ as the combination of Ca₃Bi₂, CaBi and CaBi₃ with Mg₃Bi₂. On the other hand, although there is only one compound in the Bi-Mg system, the phase diagram of Bi-Mg system shows that the composition of the compound has more Mg than Mg₃Bi₂ (see Fig. 6 [15]). It appears that the compound presents more like a solid solution of Mg₃Bi₂ and Mg. So, the calculated compound Ca₃Mg₇Bi₈ should not be taken as a compound with fixed compositions, and can be taken as a solid solution of Ca₃Bi₂, CaBi, CaBi₃, Mg₃Bi₂ and Mg, similar to Ca₃Mg₄Pb_x in Pb-Ca-Mg system [12].

Fig. 6 Optimized phase diagram of Mg-Bi system [16]

4 Conclusions

1) The calculated bismuthide governing debismuthizing process is Ca₃Mg₇Bi₈, which can be taken as a solid solution of Ca₃Bi₂, CaBi , CaBi₃, Mg₃Bi₂ and Mg, similar to Ca₃Mg₄Pb_x in Pb-Ca-Mg system.

2) The bismuthide formed during debismuthizing process exists in two states: one is free compound in the matrix of the dross, and the other is symbiotic bismuthide with the Pb-Ca phase.

3) When the peritectic reaction occurs, a part of the Pb-Ca phase will be consumed and a part of Bi-Ca-Mg phase dissolves into lead. This results in a bad debismuthizing effect and the disturbance of Bi content in final lead, according to the difference of Pb-Ca phase consumption.

References

[1] JOLLIVET L. Equilibria in systems rich in lead and containing alkali and alkaline earth metals [J]. Comptes Rendus, 1949, 228: 129.

[2] CHEN Guo-fa. Lead metallurgy [M]. Beijing: Metallurgy Industry Press, 2000: 126-127. (in Chinese)

[3] XIN H X, QIN X Y, ZHU X G, ZHANG J, KONG M G. Fabrication of nanocrystalline Mg₃X₂ (X=Bi, Sb) with supersaturated solid solubility by mechanical alloying [J]. Materials Science and Engineering B, 2006, 128: 192-200.

[4] CASTLE J F, RICHARDS J H. Lead refining: Current technology and a new continuous process [C]// DENHOLM W T. Advance in Extractive Metallurgy. London: The Institution of Mining and Metallurgy, 1977: 217-234.

[5] HIBBINS S G, CLOSSET B, BRAY M. Advances in the refining and alloying of low-bismuth lead [J]. Journal of Power Sources, 1995, 53: 75-83.

[6] BETTERTON J O, LEBEDEFF Y. Debismuthing lead with alkaline earth metals [J]. Trans AIME, 1936, 121: 205-225.

[7] EVERS D. Debismuthing by the Kroll-Betterton process [J]. Metallhuttenw, 1949, 2: 129-133.

[8] DAVEY T R A. Debismuthing of lead [J]. Journal of Metals, 1956(3): 341-350.

[9] ILEY J D, WARD D H. Development of a continuous process for the fine debismuthizing of lead [C]//DENHOLM W. Advance in Extractive Metallurgy. London: The Institution of Mining and Metallurgy, 1977: 133-139.

[10] HANCOCK P, HARRIS R. Solubility of calcium-magnesium- bismuth intermetallic in molten lead [J]. Candian Metallurgy Quarterly, 1991, 30: 275-291.

[11] LU Dian-kun, LIU Xue-shan, YE Guo-rui, HE Jia-qi. Thermodynamical analysis of debismuthizing mechanism with calcium and magnesium [J]. Journal of Shenyang Institute of Gold Technology, 1997, 16(2): 110-115. (in Chinese)

[12] DAVEY T R A. The physical chemistry of lead refining [C]// OBERBECKMANN F K, PONEN M. Lead-Zinc-Tin'80. TMS-AIME: Metallurgical Society of AIME, 1980: 477-506.

[13] LU Dian-kun, JIN Ze-nan, JIANG Kai-xi. Fine debismuthizing with calcium, magnesium and antimony [J]. Transactions of Nonferrous Metals Society of China, 2011, 21(10): 2311-2316.

[14] ZHANG J S. A review of steel corrosion by liquid lead and lead–bismuth [J]. Corrosion Science, 2009, 51: 1207-1227.

[15] MANAS P, JUNG I H. Thermodynamic modeling of the Mg-Bi and Mg-Sb binary systems and short-range-ordering behavior of the liquid solutions [J]. Computer Coupling of Phase Diagrams and Thermochemistry, 2009, 33: 744-754.

[16] NOTIN M, MEJBAR J, BOUHAJIB A, CHARLES J, HERTZ J. The thermodynamic properties of calcium intermetallic compounds [J]. Journal of Alloys and Compounds, 1995, 220: 62-75.

钙镁除铋的机理

路殿坤，金哲男，畅永锋，孙树臣

东北大学 材料与冶金学院，沈阳 110819

摘  要：为了深入了解钙镁除铋的机理，研究了钙添加量对铅中铋脱除深度的影响。将除铋渣样进行电子探针分析(EPMA)。结果表明，控制除铋过程的化合物是Ca₃Mg₇Bi₈。与Pb-Ca-Mg体系中存在的Ca₃Mg₄Pb_x化合物类似， Ca₃Mg₇Bi₈化合物可以看成是由Ca₃Bi₂、CaBi、CaBi₃和Mg₃Bi₂形成的固溶体。同时，在除铋过程中所形成的Ca₃Mg₇Bi₈化合物有两种赋存状态，一种以自由铋化物状态存在于渣相的基体中，另一种则与Pb-Ca相共生。

关键词：除铋机理；克罗尔-别特尔顿法；Ca₃Mg₇Bi₈；Pb-Ca相；Bi-Ca-Mg相

(Edited by Xiang-qun LI)

Corresponding author: Dian-kun LU; Tel: +86-24-83681319; E-mail: ludk@smm.neu.edu.cn

DOI: 10.1016/S1003-6326(13)62622-9