Ultra-Purification of Electrolytic Iron by Cold-Crucible Induction Melting

Materials Transactions, JIM, Vol. 41, No. 1 (2000) pp. 2 to 6 Special Issue on Ultra-High Purity Metals c 2000 The Japan Institute of Metals

Ultra-Puriﬁcation of Electrolytic Iron by Cold-Crucible Induction Melting and Induction-Heating Floating-Zone Melting in Ultra-High Vacuum

Seiichi Takaki and Kenji Abiko

Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan

The effect of melting under ultra-high vacuum of 107 Pa on the purification of the high-purity electrolytic iron developed was investigated using a new cold-crucible induction melting furnace which was able to melt high-purity iron of 10 kg and a new induction-heating floating-zone melting furnace, which were designed and constructed using ultra-high vacuum technology. Ultra-high vacuum melting was quite useful even for the ultra-purification of iron 10 kg in weight. The high-purity electrolytic iron of 7.5 kg with residual resistivity ratio of 2000 to 2700 was melted using the cold-crucible induction melting furnace, and the pressure during melting after melt-down was kept from 1 106 to 8 107 Pa for 9 min. The 7.5 kg ultra-pure iron ingot of more than 99.9988 mass% purity after the chemical analysis of 33 elements, most of which were lower than each detection limit, was successfully obtained, even if this ingot contains their all elements of the detection limit value. Also the ultra-pure iron 35 g in weight of more than 99.9988 mass% purity was obtained from electrolytic iron of 99.996 mass% purity by the combination of cold-crucible induction melting in ultra-high vacuum and induction-heating floating-zone melting in UHV of 4 107 Pa after the analysis of 35 elements. The concentration of C + N + O + S decreases from 14.4 mass ppm before zone-leveling to 2.4 mass ppm after four zone-leveling passes in ultra-high vacuum. For further ultra-purification of iron it is absolutely necessary to develop and establish the trace analysis techniques of measuring accurately the concentration of non-metallic impurity elements at levels below 0.1 mass ppm and metallic impurity elements at levels below 0.01 or 0.001 mass ppm.

(Received January 12, 2000) Keywords: high-purity electrolytic iron, ultra-high purity iron, ultra-purification, ultra-high vacuum, cold-crucible induction-melting, induction-heating floating-zone melting, electron-beam floating-zone melting, trace analysis

Ref. 6). 1. Introduction The purpose of the present work is to investigate the effect of melting under ultra-high vacuum of 107 Pa on the purifi- For the fundamental research on the intrinsic properties of cation of the high-purity electrolytic iron as follows and to es- iron and also to determine the inherent effects of each impu- tablish a new purification procedure to obtain ultra-pure iron: rity element on the properties ultra-high purity iron is essen- (1) two kinds of electrolytic iron about 10 kg in weight are tial. There have been many studies of the purification of iron melted under UHV of 107 Pa in CCIM furnace and (2) the in the literature,1) but few which describe the effect of ultra- iron bars less than 50 g in weight prepared from iron ingot ob- high vacuum (UHV) on it, because it is not easy to attain to tained by CCIM in UHV of 106 Pa are further zone-melted ultra-high vacuum better than 1 106 Pa during melting of in UHV of 107 Pa using the IHFZM furnace. The UHV at- iron in a large melting furnace. tainable in both our CCIM and IHFZM furnaces is better than We previously reported2) that ultra-high purity iron 15 g 1 107 Pa. in weight with the residual resistivity ratio (RRRH=64 kA/m) of more than 5000 was prepared by electron-beam floating- 2. Experimental Procedure zone melting (EBFZM) under an ultra-high vacuum of the order of 107 Pa using Johnson-Matthey pure iron as start- 2.1 Starting materials ing iron bars. The production of high-purity electrolytic iron Two kinds of electrolytic iron (EFe) developed by us,3) has been developed by us3) and the purity attainable is in- high-purity iron and higher-purity iron were used as the creasing year by year. On the other hand, we have made starting materials. The analytical results of two batches of surface analysis experiments on Fe–P alloys under an ultra- high-purity electrolytic iron, EFe-1 and EFe-2, are shown in high vacuum of 4 109 Pa.4) We also designed and made Table 1. The purity was more than 99.996 mass% after the a new high-temperature optical microscope using UHV tech- analysis of 35 elements. For the higher-purity electrolytic nology.5) Then, on the basis of such our experience and the iron, EFe-3, not chemical analysis but residual resistivity ratio background mentioned above, we started to design and build (RRRH ) measurement was performed and the RRRH value new purification apparatus to prepare ultra-high purity base was from 2000 to 2700 for 16 batchs of higher-purity elec- metal specimens, and first in UHPM-94 we reported on the trolytic iron. construction of a new arc melting furnace by using UHV technology,6) second in UHPM-95 on the construction of a new 2.2 Cold-crucible induction melting induction-heating floating-zone melting (IHFZM) furnace7) The new induction melting furnace with a cold-copper cru- and then on the construction of a new cold-crucible induction- cible is described in detail in Ref. 8). This furnace can melting (CCIM) furnace8) in UHPM-97 and a part of the re- make an iron ingot up to 10 kg. The power supply provides sults of the purification effect by CCIM.9,10) The importance max. 240 kW and 10 kHz. The main chamber which has two of UHV in a purification apparatus is described in detail in flanges of 1.4 m in diameter and a volume of about 2.3 m3 can Ultra-Purification of Electrolytic Iron by Cold-Crucible Induction Melting and Induction-Heating Floating-Zone Melting in Ultra-High Vacuum 3

Table 1 Results of chemical analysis of iron samples (mass ppm).

Starting iron Iron ingot IFe-1 IFe-2 IFe-3 EFe-1 EFe-2 EFe-1+EFe-2 EFe-1+EFe-2 EFe-3 CCIM, >4 106 Pa CCIM, >2 106 Pa CCIM, >8 107 Pa C 3.2 3.5 1.0 0.7 0.6 N <0.1 0.1 <0.1 0.8 <0.1 O 24 12 7.2 6.0 1.9 S 1.3 1.1 0.5 0.7 0.8 H 1.3 1.0 0.7 ------Al 3.1 <0.2 0.2 <0.4 <0.4 As 1.2 0.8 1.1 1.6 1.4 B 0.6 0.3 0.03 0.38 0.12 Ba <0.1 <0.1 <0.1 <0.1 <0.1 Bi <0.1 <0.1 <0.1 <0.1 <0.1 Ca <0.2 <0.2 <0.2 Cd <0.001 <0.001 <0.001 <0.001 <0.001 Co 0.3 <0.1 0.7 <0.1 0.6 Cr <0.3 <0.3 0.2 0.5 <0.5 Cu 0.4 0.2 0.5 0.2 0.2 Ga <0.2 0.3 0.2 <0.3 <0.3 Hf <0.06 <0.06 <0.06 <0.07 <0.07 Mg <0.01 <0.01 <0.01 <0.02 <0.02 Mn <0.01 <0.01 <0.01 <0.01 <0.01 Mo 0.1 <0.1 0.1 <0.3 <0.3 Nb <0.05 <0.05 <0.05 <0.03 <0.04 Ni <0.1 0.1 0.1 <0.1 0.2 P 0.1 0.7 0.1 0.3 0.3 Pb <0.01 0.12 0.04 <0.03 <0.03 Sb <0.3 <0.3 <0.3 <0.3 <0.3 Se <0.04 <0.04 <0.04 <0.04 <0.04 Si11111 Sn <0.4 <0.4 <0.4 <0.5 <0.5 Ta <0.6 <0.6 <0.6 <0.9 <0.9 Te <0.03 <0.03 <0.03 <0.02 <0.02 Ti <0.2 <0.2 <0.1 <0.2 <0.2 V 0.03 0.03 0.05 <0.03 <0.03 W <1 <1 <111 Zn 1.6 14.2 0.6 <0.1 <0.1 Zr <0.2 <0.2 0.2 <0.07 <0.07 Purity >99.996% >99.996% >99.9982% >99.9983% >99.9988% be evacuated to a base pressure of 6.7 108 Pa by a system ﬂow of high-purity hydrogen or in UHV. The UHV attainable of an oil diffusion pump and a cold trap. The total pressure in this IHFZM furnace is better than 1 107 Pa. An exam- and the partial pressure of mass number between 1 and 50 ple of the vacuum during each zone-leveling pass in UHV is in the main chamber were measured before, during and after shown in Table 3. The zone travel rate was 2.5 mm/min and melting as seen below. the distance was about 90 mm for FZM only in UHV, about The ingot obtained was heated in high-purity Ar atmo- 120 mm for FZM only in hydrogen and about 55 mm for FZM sphere at 1150 K and then forged, rolled and machined to rods in hydrogen and then in UHV. This new IHFZM apparatus is 8 mm in diameter for FZM. described in detail in Ref. 7).

2.3 Induction-heating floating-zone melting 2.4 Analysis Iron bars 8 mm in diameter for FZ refining were prepared The samples for analysis were cut from the starting ma- from the ingot IFe-1 obtained by cold-crucible induction terials and the ingots and bars melted in the present exper- melting of EFe-1 and EFe-2 at the pressure of 106 Pa. The iment, and polished mechanically and then electrolytically. further purification by floating-zone melting was performed The gaseous impurities in these samples were analyzed by under the following conditions. Four zone-leveling passes, combustion infrared absorption method for carbon and sulfur which means two up-and-down passes, were made in the slow on the LECO CS444LS and by fusion thermal conductivity 4 S. Takaki and K. Abiko

Fig. 1 Changes in the total pressure and the partial pressure of gases of mass numbers 2, 18, 28, 32 and 44 during induction heating of higher-purity electrolytic iron, EFe-3. method for oxygen and nitrogen on the LECO TC43611) in the accuracy of 0.1 mass ppm in Japan Analyst Corporation. The metallic impurities in these samples were analyzed by several methods in our institute.12,13)

3. Results and Discussion

3.1 Purification by cold-crucible induction melting A high-purity iron ingot of 8 kg, IFe-1, was made from high-purity electrolytic iron 7 kg EFe-1 and 1 kg EFe-2,9,10) 10 kg IFe-2 from 8 kg EFe-1 and 2 kg EFe-2, and 7.5 kg IFe-3 from 7.5 kg EFe-3. Figure 1 shows the changes in the total pressure and the partial pressure of gases of mass numbers 2, 18, 28, 32 and 44, which seem to be H2,H2O, CO, O2 and CO2, during induction heating and melting of higher- purity electrolytic iron, EFe-3. Roughly speaking, the total pressure and the partial pressure are two orders of magnitude lower than those in the case of IFe-1 melting (see Fig. 2 of Ref. 9)) during heating before melt-down and one order of magnitude lower during melting after melt-down. It is the Fig. 2 The high-purity iron ingot, IFe-3. point worthy of mention that the pressure during melting after melt-down for IFe-3 of weight 7.5 kg was kept from 1 106 to 8 107 Pa for 9 min, because the lowest pressure during if these ingots contain all impurities analyzed of each detec- melting was 2 106 Pa for IFe-2 and 4 106 Pa for IFe-1. tion limit value. The picture of the high-purity iron ingot, IFe-3, is shown in Fig. 2. 3.2 Purification by floating-zone melting The analytical results of these high-purity iron ingots made Table 2 shows the changes in concentration of gaseous im- by cold-crucible induction melting in UHV are shown in purities in high-purity iron before and after floating-zone lev- Table 1. Almost every impurity but oxygen is less than eling in hydrogen and/or in UHV. An iron bar was melted 1 mass ppm and most of impurities analyzed are lower than only in UHV by FZL. The changes in pressure during the FZL each detection limit of analysis. The total amount of im- passes are shown in Table 3, comparing with the results of purities was reduced from about 40 mass ppm in electrolytic EBFZM in UHV,2) by which we successfully prepared ultra- iron (99.996 mass% purity) to less than 18 mass ppm in ingot high purity iron with RRR of more than 50002) and oxygen IFe-1 (more than 99.9982 mass% purity) after the analysis of content of less than 3 mass ppm,14) which was the detection 35 elements, to less than 17 mass ppm in IFe-2 (more than limit in those days. At the first pass the vacuum in the present 99.9983% purity) and to less than 12 mass ppm in IFe-3 (more FZL is already two orders of magnitude better than that in than 99.9988% purity) after the analysis of 33 elements, even EBFZM and after four passes the pressure in the present FZL Ultra-Purification of Electrolytic Iron by Cold-Crucible Induction Melting and Induction-Heating Floating-Zone Melting in Ultra-High Vacuum 5

Table 2 Changes in concentration of gaseous impurities in high-purity iron before and after ﬂoating-zone leveling in hydrogen and/or in UHV (mass ppm).

After FZL Before FZL 4 passes in H2 4 passes in UHV 4 passes in H 2 +4 passes in UHV C 6.1 0.2 0.7 0.5 N 0.7 0.0 0.6 0.3 O 7.0 2.0 1.0 0.8 S 0.6 0.2 0.6 0.6 C + N + O + S 14.4 2.4 2.9 2.2

Table 3 Changes in vacuum with zone-melting passes (Pa). which is almost on the level of the detection limit. Pass number The present IHFZL EBFZM [2] Oxygen content decreased from 7.0 to 2.0 mass ppm after four passes in UHV. To decrease oxygen further, an- 11 106 1 104 other iron bar was melted by FZL in hydrogen and the 26 107 1 105 half length of the bar cut was further melted by FZL in 35 10 7 7 10 6 44 107 3 106 UHV; as a result, oxygen decreases in concentration from 81 106 7.0 mass ppm to 1.0 mass ppm after four passes in hydrogen 10 4 107 and to 0.8 mass ppm after further four passes in UHV. The 11 4 107 melting in hydrogen is known to be more effective in reduc- 12 4 107 ing oxygen in high-purity iron. In the present investigation ultra-high purity iron with the concentration of C + N + O + S of 2.4, 2.9 and 2.2 mass ppm were obtained after four zone-leveling passes in UHV, in hydrogen and further four zone-leveling passes in UHV, respectively. Thus it follows that ultra-high purity iron less than 50 g in weight of more than 99.9988 mass% purity were obtained from electrolytic iron of 99.996 mass% purity by the present puriﬁcation process, even if the total amount of metallic impurities of 9.1 mass ppm was not reduced during zone-melting process.

3.3 Effects of ultra-high vacuum on purification Ultra-high vacuum attained in the chamber of a melting furnace is thought to have three effects on the purification of iron; first, a direct purification effect by evaporation of impurities of which vapor pressure is high at the melting point, for example, in CCIM and EBFZM in UHV, and second, an indirect effect by less contamination of gas atmosphere by outgassing from the inner wall and parts of the chamber, for example, on arc melting, CCIM and IHFZM in gas atmosphere such as Ar, He or H2. In addition to these effects, third, the ultra-high vacuum acts as a reducing atmosphere, because the residual gas in ultra-high vacuum is not water vapor but mostly hydrogen, if the chamber is clean and well outgassed by baking; thus, the oxygen concentration is reduced by the ultrahigh vacuum melting. These agree with the results ob- Fig. 3 Carbon and nitrogen plot of ultra-pure iron after four zone-leveling tained in the present experiments as seen in Tables 1 and 2. passes in UHV from LECO CS444LS and TC436, respectively. We conclude that ultra-high vacuum is very important and effective for the ultra-purification of iron, in particular for the reduction of gaseous impurities. became 4107 Pa, which corresponds to that after 10 passes in EBFZM. This means that the purity of the present starting 4. Conclusion iron bar is better than that of JM-iron used in the EBFZM experiment. After four zone-leveling passes in UHV, carbon The effect of melting under ultra-high vacuum of 107 Pa and nitrogen decreased to 0.2 and 0.0 mass ppm, respectively. on the purification of the high-purity electrolytic iron was in- Figure 3 shows the signal in the carbon and nitrogen analysis, vestigated using a new cold-crucible induction melting fur- 6 S. Takaki and K. Abiko nace and a new induction-heating floating-zone melting fur- to thank Toho Zinc Co. Ltd. for supplying high-purity elec- nace. It is concluded that ultra-high vacuum melting is very trolytic iron. We are grateful to Mr. Y. Morimoto, Japan An- important and effective even for the ultra-purification of iron alyst Co., Dr. K. Takada and the members of the analytical 10 kg in weight. group in our institute for analysis of trace amounts of ele- (1) The high-purity electrolytic iron of 7.5 kg with the ments in high-purity iron and also to Mr. N. Harima and Mr. residual resistivity ratio of 2000 to 2700 was melted using T. Nakajima for help in this experiments. the cold-crucible induction melting furnace and the pressure during melting after melt-down was kept from 1 106 to REFERENCES 8107 Pa for 9 min. The 7.5 kg ultra-pure iron ingot of more than 99.9988 mass% purity after the chemical analysis of 33 1) Ultra High Purity Base Metals, Proc. 1st. Int. Conf. on Ultra High Pu- elements, most of which were lower than detection limit, was rity Base Metals (UHPM-94), ed. by K. Abiko, K. Hirokawa and S. Takaki, The Japan Institute of Metals, Sendai, (1995). successfully obtained, even if this ingot contains their all ele- 2) S. Takaki and H. Kimura: Scripta Metall., 10 (1976), 1095–1100. ments of detection limit value. 3) T. Daicho, N. Arisawa and K. Abiko: Ultra High Purity Base Metals, (2) Ultra-pure iron 35 g in weight of more than eds. K. Abiko, K. Hirokawa and S. Takaki, The Japan Institute of Met- 99.9988 mass% purity were obtained from electrolytic iron of als, (1995), pp. 397–402. 4) K. Abiko: Ultra High Purity Base Metals, eds. K. Abiko, K. Hirokawa 99.996 mass% purity by the combination of cold-crucible in- and S. Takaki, The Japan Institute of Metals, (1995), pp. 1–26. duction melting in UHV and induction-heating floating-zone 5) K. Abiko: Proc. JIMIS-4 Conf. on Grain Boundary Structure and Re- melting in UHV of 4 107 Pa after the analysis of 35 el- lated Phenomena, Supplement to Trans. JIM, Vol. 27 (1986), pp. 633– ements. The concentration of C + N + O + S decreases 640. 6) R. Oiwa, Y. Ashino, S. Takaki and K. Abiko: Ultra High Purity Base from 14.4 mass ppm before zone-leveling to 2.4 mass ppm af- Metals, eds. K. Abiko, K. Hirokawa and S. Takaki, The Japan Institute ter four zone-leveling passes in ultra-high vacuum. of Metals, (1995), pp. 516–519. (3) For further ultra-purification of iron it is absolutely 7) S. Takaki, Y. Ashino, Y. Morimoto, M. Tanino and K. Abiko: J. necessary to develop and establish the analysis techniques of Physique (IV) C7, 5 (1995), 159–164. 8) K. Abiko, T. Nakajima, N. Harima and S. Takaki: Phys. Stat. Sol. (a), measuring accurately the concentration of non-metallic impu- 167 (1998), 347–355. rity element at levels below 0.1 mass ppm and metallic impu- 9) K. Abiko and S. Takaki: Vacuum, 53 (1999), 93–96. rity element at levels below 0.01 or 0.001 mass ppm. 10) S. Takaki and K. Abiko: Vacuum, 53 (1999), 97–100. 11) D. Lawrenz: Ultra High Purity Base Metals, eds. K. Abiko, K. Hirokawa and S. Takaki, The Japan Institute of Metals, (1995), pp. 183– Acknowledgements 190. 12) K. Hirokawa and K. Takada: Ultra High Purity Base Metals, eds. We would like to acknowledge the support from the Min- K. Abiko, K. Hirokawa and S. Takaki, The Japan Institute of Metals, (1995), pp. 161–170. istry of Education, Culture and Science, and from the Pro- 13) K. Takada: Phys. Stat. Sol. (a), 160 (1997), 561–565. gram of Core Research for Evolutional Science and Technol- 14) R. Onodera, S. Takaki, K. Abiko and H. Kimura: Mater. Sci. Eng., A141 ogy (CREST), Japan Science and Technology Co. We wish (1991), 261–267.