Unusual Wetting of Liquid Metals on Iron Substrate with Oxidized Surface in Reduced Atmosphere

Unusual Wetting of Liquid Metals on Iron Substrate with Oxidized Surface

in Reduced Atmosphere

Nobuyuki Takahira

, Toshihiro Tanaka

, Shigeta Hara

and Joonho Lee

1

Division of Materials and Manufacturing Science, Graduate School of Engineering, Osaka University, Suita 565-0871, Japan 2_{Department of Mechanical Engineering, Fukui University of Technology, Fukui 910-8505, Japan}

3_{Division of Materials Science and Engineering, Korea University, Seoul, Korea}

The authors found that a liquid Cu droplet wetted and spread very widely on a solid substrate of Fe in a reduced atmosphere after the surface oxidation of the substrate. The mechanism of the unusual wetting behavior was investigated by using surface-oxidized Fe substrate with liquid Cu, Ag, Sn, and In. It was found that under a reduced atmosphere condition, fine pores were formed at the surface of the substrate which had been oxidized, and that the pores were connected to each other continuously over the whole surface. The liquid metals penetrate into these pores by capillary force to cause the unusual wetting behavior.

(Received July 25, 2005; Accepted October 13, 2005; Published December 15, 2005)

Keywords: wettability, high temperature oxidation, porous structure, capillary phenomenon, copper, silver, tin, indium

1. Introduction

The recycle processes of various materials have been in-vestigated to realize a recyclable society.1,2) _{In particular,}

many procedures to avoid Cu mixing with Fe have been tried in an effort to recycle Fe and Cu mixed scrap. After these metals are mixed, it is very difficult to remove Cu from Fe through the usual refining processes.3) Even though Cu in steels can be diluted with raw iron, a little amount of Cu in the steels causes surface hot shortness.4)This is a phenomen-on caused by Cu in the hot rolling process of steels cphenomen-ontaining Cu. That is to say, Fe is preferentially oxidized at the steel surface and liquid Cu droplets appear at the steel/scale inter-face. The liquid Cu does not have good wettability to the scale layer, and it infiltrates into grain boundaries during the hot rolling process to cause cracks and strength degrada-tion of the steels. In this phenomenon, it is important to un-derstand wettability between liquid Cu and scales, or the oxi-dized surface of solid iron. Some experimental results on wettability have been reported by Brown et al.,5) _{Hara et}

al.,6)_{and Mukaiet al.,}7)_{and there have been many studies}8–10)

about the mechanisms of surface hot shortness. It is generally known that liquid metals are not wetted with solid oxide and that there is good wettability between liquid metals and solid metals.11)However, the authors have found, in experiments on the wettability of liquid Cu, a peculiar phenomenon in which liquid Cu unusually wetted on solid Fe substrate with the oxidized surface. In this paper, we describe the unusual wetting behavior of liquid metals on surface-oxidized Fe sub-strate and the mechanism of this unusual wetting behavior.

2. Unusual Wetting Phenomenon

2.1 Liquid Cu wetting on solid Fe substrate with

oxidized surface

The authors found a wetting phenomenon in which liquid Cu, in reduction atmosphere, unusually wetted and spread

onto solid Fe substrate, of which the surface had been oxidized. This phenomenon was concerned with iron oxide formed on the surface of the iron substrate and therefore different from typical wetting to the clean iron surface, which is described later in detail. The experimental procedure was as follows: Iron rods of 99.5% purity were used. Each rod was 5 mm in diameter and 50 mm in length. After ultrasonic cleaning of these rods in acetone and ethanol, some samples were oxidized at 1073 K for 1 h in air. Two kinds of solid Fe samples, a pure metallic Fe rod and a surface-oxidized Fe rod, were contacted with a liquid Cu droplet settled on a graphite plate at 1423 K in Ar–10%H2, as shown in Fig. 1. The rate of heating and cooling was 400 K per hour and the sample was kept at 1423 K for 20 min.

Iron rod without

oxidized surface

liq. Cu

Iron rod with

oxidized surface

Fig. 1 Experiment for wettability of liquid Cu on solid iron rods in Ar– 10%H2atmosphere.

*1_{This Paper was Originally Published in Japanese in J. Jpn. Inst. Met.69}

(2005) 465–471.

*2_{Graduate Student, Osaka University}

Figure 2 shows the appearances of the samples after the experiment. The result of using Fe without surface oxidation is shown in Fig. 2(a). Liquid Cu wetted just at the tip of the rod where liquid Cu contacted directly. Figure 2(b), on the other hand, shows that liquid Cu, which contacted at the

lower end of the Fe rod with an oxidized surface, climbed up to the top of the rod. The results of this experiment suggest that this unusual wetting phenomenon is concerned with the oxide layer formed on the surface of the Fe sample.

To confirm this unusual wetting behavior of liquid Cu observed on a solid iron rod, as shown in Fig. 2, further experiments were carried out by using a screw made of Fe with a hole as well as a cup-shaped iron rod (purity 99.5%,

10mm30mm) with a hole (5mm20mm). The experimental procedure is illustrated in Fig. 3. These two samples were oxidized at 1073 K for 1 h in air. After forming the oxide layer on the surface, each cup-shaped sample with Cu shots in the hole was heated up to 1423 K in reduction atmosphere. The rate of heating and cooling was 400 K per hour, and the sample was kept at 1423 K for 20 min.

Figure 4 shows the samples after the above experiments were completed. As shown in Figs. 4(a) and (b), liquid Cu in the hole of the sample climbed up along the inside wall and spread over to the lower edge of the screw. Similar behavior was also seen for the cup-shaped iron rod, as shown in Fig. 4(c).

2.2 Wettability of various liquid metals for solid iron substrate in reduction atmosphere

As described in the preceding section, it was found that

(a)

5mm

(b)

wetting

wetting

Fig. 2 Wettability of liq. Cu on iron substrate. (a) Iron substrate without oxidized surface (b) Iron substrate with oxidized surface.

Iron screw

Making a hole and

oxidizing in the air

Iron rod

Cu particles put

into the hole

Heating up

in Ar-10%H

Cup-shaped

iron

Fig. 3 Experiment for wettability of liquid Cu on cup-shaped solid iron.

5mm

(a)

(b)

(c)

liquid Cu wetted unusually on the iron sample with an oxidized surface. Subsequently, whether the unusual wetting of other metals also occurs was investigated. Cu, Ag, Sn, and In, all of which have relatively small solubility of Fe, were used as liquid metals. The effect of temperature on the unusual wetting was also investigated by using Sn and In, which have low melting points. The cup-shaped iron rod (purity 99.5%, 10mm30mm) with a hole (5mm

20mm) was oxidized at 1073 K for 1 h in air. After putting small metal particles in the hole, the sample was heated up to the following temperature, which was determined consider-ing the meltconsider-ing points of these metals.

Cu: 1423 K Ag: 1273 K

Sn and In: 1173, 973, and 773 K

When each surface of the solid iron rod wetted by liquid metals was observed by EDX, we found that all of the four metals (Cu, Ag, Sn, and In) spread even to the outside surface of the rods by the unusual wetting. Thus the unusual wetting behavior of each liquid metal was confirmed at all of the experimental temperatures stated above, although some different wetting behaviors were observed for these metals due to differences in the solubility of Fe in each liquid metal, as described later.

3. Mechanism of Unusual Wetting Behavior

3.1 Surface of reduced iron substrate after oxidization

In the experiments stated in sections 2.1 and 2.2, the

sample of the iron substrate was oxidized before the wetting experiments in the reduction atmosphere. Therefore it is meaningful to observe the surface structure of the iron substrate after the reduction of the oxide layer on the substrate. An iron substrate (99.5%Fe) with a size of 20

301mm was oxidized at 1073 K for 1 h in air and was reduced at 1373 K for 1.5 h in Ar–10%H2. The rate of heating and cooling was 400 K per hour. Figure 5 illustrates the above process. The surface and the cross-section of the substrate were observed by using SEM.

Figure 6 shows SEM images of the surface and the cross-section of the reduced iron substrate. After the oxidization and the reduction, it was observed that a lot of pores were formed on the surface, as shown in Fig. 6(a). The upper side of the photograph corresponds to the surface of the substrate. Figure 6 demonstrates that a porous layer of approximately 80mmthickness was formed with many pores on the surface of the substrate in the above redox condition. It is supposed that the pores were made by reducing the oxide and removing the oxygen from the layer. Some pores are found at the outermost surface of the substrate as micro-size holes, as shown in Fig. 6(a).

3.2 Porous structure on the surface layer of reduced iron

The formation of a porous layer with many pores was found after the reduction of the iron oxide layer, as described in section 3.1. In order to observe the 3-dimensional porous structure, we tried to penetrate liquid Ag into the porous structure and then remove the iron substrate by an acid

oxidation

Iron substrate

Surface-oxidized

iron

reduction

Reduced iron

Observation of the surface and

cross-section by using SEM

Fig. 5 Sample preparation for the observation of the surface and the cross-section.

10

µ

m

(a)

(b)

aqueous solution. The microstructure penetrated by Ag corresponds to the porous layer originally formed on the iron substrate. The reduced iron substrate was prepared in the procedure stated in section 3.1, and the reduced substrate was contacted with liquid Ag at 1273 K. The iron sample with Ag penetrated in the surface layer was soaked in hydrochloric acid, and a thin film of Ag was obtained after the dissolution of Fe.

Figure 7 shows the Ag film obtained from the above procedure. When we observed the microstructure of the film, it was found that, as shown in Fig. 8, liquid Ag penetrates into almost all of the porous structure of the iron substrate. In addition, these pores in the porous layer of the reduced iron substrate do not exist individually, but are connected to each other continuously. The average diameter of the pores is 2– 3mm, as can be seen in Fig. 8(b).

3.3 Unusual wetting of various liquid metals on reduced iron substrate

In order to investigate the wetting behavior of various liquid metals, wetting experiments were carried out using the

5mm

Fig. 7 Ag film obtained from the penetration of liq. Ag in the porous layer on Fe substrate.

Metal plate

(Cu, Ag, Sn, In)

Reduced iron

substrate

Heating up

by using SEM

30mm

1mm

20mm

Fig. 9 Sample preparation in the wetting experiments.

80

µ

m

(a)

(b)

5

µ

m

following four metals at each experimental temperature: Cu at 1423 K, Ag at 1273 K, Sn at 1273 K and 873 K, and In at 1073 K. The reduced iron substrates were prepared in the way described in section 3.1. The substrate was vertically placed in a furnace, with each metal plate attached on the bottom, as shown in Fig. 9, and heated up to the above temperature in Ar–10%H2. The rate of heating and cooling was 400 K per hour, and the sample was kept at the above temperature for 20 min. The surface and the cross-section of the five samples fabricated with the four metals were observed by using SEM, and the concentration distribution of each element was analyzed with EDX.

3.3.1 Factors of the unusual wetting of liquid Cu

The cross-section of the reduced iron substrate that liquid Cu wetted was observed by using SEM, and the concentration of Cu was analyzed by EDX, as shown in Fig. 10. Liquid metal penetrated into the substrate from left to right. In order to investigate the forefront of the penetration of liquid Cu into the porous layer, a little amount of Cu was used to prepare a substrate in which the liquid Cu had penetrated into half of the porous layer. Figure 10 shows the forefront of this penetration of Cu into the layer on the iron substrate. In the left side of the photograph, Cu was observed at all pores of the porous layer. Pores are seen in the upper part of Fig. 10(a) but are not present in the bottom part. Figure 10(b) shows that liquid Cu penetrated into the pores of the porous layer on the surface of the reduced iron substrate. We could therefore conclude that the liquid metal penetrated into these pores formed in the porous layer by the capillary force to cause the unusual wetting phenomenon.

3.3.2 Effect of the solubility of Fe in liquid metals on their wettability with iron substrate

Figure 11 shows the experimental results for the wetting of liquid Ag, Sn, and In. Each metal existing in the porous layer is indicated as a white area in Fig. 11. The wetting

behaviors of the liquid metals penetrating into the porous layer of the substrate have almost the same characteristics of Cu in Fig. 10.

Figure 12 indicates the outermost surfaces of the iron substrates, in which the above four metals (Cu: 1423 K, Ag: 1273 K, Sn: 1273 K, In: 1073 K) wetted unusually. The way that each liquid metal wetted to the outermost surfaces illustrates its own distinctive manner of wetting. Although liquid Cu and Sn widely cover the surfaces of the iron substrates, liquid Ag and In only appear in some limited areas. We guess that these differences depend on the solubility of Fe in each liquid metal. Under the present experimental conditions, Fe is solved by about 4.7% into Cu and 8.0% into Sn, but little into Ag and 0.4% into In.12)As shown in Fig. 12, liquid metals that have relatively high solubility of Fe were found to be viable at almost the entire surface area of iron substrates, but metals with low solubility of Fe were observed only at pores on the surface or just around them. Thus, the extent of the wetted surface area of liquid metals seems to have a rough correlation with the solubilities of those liquid metals in solid Fe although liquid metals should have small solubilities in solid Fe. Then, the solubilities of liquid metals in solid Fe, in other words, attractive interactions of liquid metals with solid Fe, is considered to be one of driving forces to control the spreading of liquid metals on the substrate by the unusual wetting behavior.

Figure 13 shows the EDX analysis of Sn at the outermost surface on the substrate obtained from the wetting experi-ment at 873 K. The white area shows Sn distribution in Fig. 13. Sn has approximately 0.5% solubility of Fe at

(a)

(b)

20

µ

m

Fig. 10 Cross-section of the reduced iron substrate wetted by liq. Cu: (a) SEM image, (b) EDX image.

(a)

(b)

(c)

20µm

873 K, and this solubility is almost the same value as that of In at 1073 K, as shown in Fig. 12(d). Therefore, though it had been predicted that liquid Sn would hardly spread in the present experimental conditions, it in fact wet widely, in contrast to the case of In. In light of this result, it is supposed that since Sn is the only metal among the above four metals which forms an intermetallic compound with Fe,12)then the wetting proceeds to the extent that the metallic compound is formed.

4. Conclusions

In the present study, unusual wetting behavior was found for liquid Cu on a reduced iron substrate with a pre-oxidized surface. This wetting investigation leads to the following conclusions.

(1) The unusual wetting of liquid metals to the iron substrate is caused by capillary penetration into the porous structure formed on the surface of the iron sample that is reduced after oxidized in air.

(2) Except for Cu, unusual wetting was observed by Ag, Sn, and In, all of which have Fe solubility of only a few percent.

(3) The wetting behavior at the outermost surface depends on the solubility of Fe in each metal or the formation of an intermetallic compound.

Acknowledgments

The authors would like to thank Mr. S. Wakita, Mr. S. Nakatsuka, and Ms. T. Kitamura for their kind experimental assistance.

REFERENCES

1) K. Harada: Bull. Iron Steel Inst. Jpn.7(2002) 243–251. 2) S. Hayashi: Bull. Iron Steel Inst. Jpn.8(2003) 879–882. 3) H. Katayama and Y. Mizukami: Mater. Jpn.35(1996) 1283–1289. 4) K. Shibata, K. Kunishige and M. Hatano: Bull. Iron Steel Inst. Jpn.7

10

µ

m

(a) Cu

(b) Ag

(c) Sn

(d) In

Fig. 12 EDX images of Cu, Ag, Sn, and In on the surfaces of the reduced iron substrates wetted by: (a) Cu at 1423 K, (b) Ag at 1273 K, (c) Sn at 1273 K, and (d) In at 1073 K.

10

µ

m

Sn

(2002) 252–256.

5) R. B. Brown and F. E. Block: U.S. Bureau of Mines, Report of Invest.

7218(1968) 1–15.

6) S. Hara:Report of ‘‘Fundamental Research on Tramp Elements in Iron Base Scraps’’ed. by Division of High Temperature Processes, (Iron and steel Institute of Japan 1996) pp. 105–108.

7) K. Mukai, S. Yamakawa, T. Matsubara and T. Mori: CAMP-ISIJ7

(1994) 5.

8) N. Imai, N. Komatsubara and K. Kunishige: ISIJ Int.37(1997) 217–

223.

9) M. Hatano, K. Kunishige and Y. Komizo: Tetsu-to-Hagane´88(2002) 142–147.

10) K. Shibata, S. J. Seo, K. Asakura and Y. Akiyama: Trans. MRSJapan24

(1999) 333–336.

11) T. Choh and T. Oki: J. Jpn. Inst. Met.28(1989) 285–293.