Introduction

Copper-iron alloys (CFAs) are known for their optimum combination of electrical conductivity and mechanical strength [1]. This alloy can be used in magnetic recording, optical devices, electrical contact materials and sensors [1]. The advantage of CFA over alternative alloy compositions like Cu-Nb and Cu-Ag lies in the cheap cost of Fe as compared to both Nb and Ag. CFAs have been produced by various methods, by casting [2], mechanical alloying [6], or rapid solidification processes such as melt spinning [3] and gas atomization [4]. However, this alloy has limited industrial applications due to segregation that takes place during solidification. [5]. Previous works hint that rapid solidification has great potential in the manufacturing of this alloy with expected microstructure [6].

Atomization is a process with the highest potential for large-scale exploitation of the microstructural benefits achieved by rapid solidification [7]. CFA is a typical example of a peritectic system [8]. It has a nearly flat liquidus line and a retrograde solidus line; therefore it has the thermodynamic tendency to form an immiscibility gap in the liquid state [9], as shown in Fig. 1. This phase separation was first observed in 1958 [10]. Researchers have investigated rapid solidification as a fabrication method, and they have also studied the thermodynamics and modeling of the microstructural evolution of CFAs [11-13]. For gas-atomized CFAs, however, there has been no information on the microstructural evolution with composition and powder size. The authors consider it important to characterize CFA with composition and powder size because grain boundaries present scattering centers for conduction electrons and there may exists a direct relationship between composition, microstructure, grain size and material properties of this alloy. In this study, the authors characterized the microstructural and structural properties of CFA and investigated the relationship between composition, powder size and microstructure.

Fig. 1

Equilibrium phase diagram of copper-iron alloys (CFAs) drawn using FactSage (Thermfact/CRCT, Canada; GTTTechnologies, Germany). The vertical arrows (Alloy 1 = CFA50, Alloy 2 = CFA30 and Alloy 3 = CFA10) show the compositions of CFAs used in the experiments. L is the liquid phase, whereas γ-Fe and α-Fe are the phases of Fe. The dashed curve represents the bimodal line of the miscibility gap.

Experimental

The composition and powder size of CFA are summarized in Table 1. Alloy powders of desired composition were manufactured in gas atomization system. High purity (99.99%) Cu and Fe pellets were melted inside alumina (Al₂O₃) crucible placed in induction melting system ((Model: DTIH-0050MF, Dongyang induction melting furnace, Korea) with boron nitride carbon (BNC) orifice and stopper. Before melting, vacuum environment was ensured in the melting and atomization chamber and argon gas was sprayed. 10 min preheating was performed to completely remove moisture from crucible, stopper and pellets. Regular intervals of induction powder i.e. 2KW per 2 min were given to the crucible. Additional power was given for superheating the melt above 100°C to ensure uniform melting. Liquid melt was atomized with argon gas at 2MPa (20 bar) pressure. After atomization, powder was collected and meshed into power size ranges as specified in Table 1. Powder samples were analyzed using X-ray diffraction (XRD; D8-Advance, Bruker Miller, USA) and scanning electron microscope (FEG-SEM; JEOL, JSM-7100F, Japan) to reveal microstructure, grain size and structural properties. Elemental concentrations were determined and mapping was done with energy dispersive spectroscopic (EDS) detector installed inside the SEM. Volume fractions were measured using image analysis software ImageJ (ImageJ; Java, NIH, USA).

Table 1

Composition and powder size of CFA used in the experiment.

Alloy	Composition (wt.%)	Category by size	Particle size
CFA50	Cu50Fe50	p1	~24 µm
CFA30	Cu70Fe30	p2	25-49 µm
CFA10	Cu90Fe10	p3	50-99 µ
		p4	100-200 µm

Results and Discussion

Shown in Fig. 2, 3 and 4 are the microstructures of CFA depending upon composition and powder size. The bright areas in the SEM images correspond to Cu phase and dark areas belong to Fe phase. In all of the three compositions, Cu phase acts as a matrix in which Fe which is a secondary phase is distributed homogeneously. Some metastable microstructures can also be seen in these SEM images. Gas atomization provides high cooling rates which are usually in excess of 10⁶ K/ sec [6]. Such high cooling rates can hint for the supplication of adequate energy for the formation of metastable phases during the powder synthesis. From the phase diagram between Cu and Fe (Fig. 1), it can be seen that there exists no intermetallic compound between Cu and Fe over the entire range of composition.

Fig. 2

Microstructures of CFA10 (a) p1, (b) p2, (c) p3 and (d) p4 samples respectively. Darker areas correspond to Fe phase and brighter areas correspond to Cu phase.

Fig. 3

Microstructures of CFA30. (a) p1, (b) p2, (c) p3 and (d) p4 samples respectively. Darker areas correspond to Fe phase and brighter areas correspond to Cu phase.

Fig. 4

Microstructures of CFA50 (a) p1, (b) p2, (c) p3 and (d) p4 samples respectively. Darker areas correspond to Fe phase and brighter areas correspond to Cu phase.

Cu and Fe are completely miscible in all compositions in liquid state. A horizontal dotted curve in the phase diagram (Fig. 1) denotes the boundary of the miscibility gap; below which Cu and Fe phase separate. Upon decreasing temperature during cooling, the alloys melt fall into the metastable miscibility gap and separates into Fe-rich and Cu-rich liquids. The melting point of Fe is 1538°C while that of Cu is 1083°C. During cooling, Fe crystallizes first followed by the crystallization of Cu. Fe rich phase precipitates first due to high crystallization temperature than Cu. Fe-rich phase is first solidified as small spheres, remaining liquid becomes Cu rich and Fe deficient so crystallization and coalescence of Fe phase takes place at elevated rate. Because of the rapid processing of alloy melt into powder, Fe-rich droplets are supersaturated with Cu. Some of the Fe exists as a solid solution in Cu and this is referred as the Cu rich phase while Cu exists as a solid solution in Fe, referred as the Fe-rich phase.

Same sized powders of different compositions yield different microstructures (Fig. 2-4). From the phase diagram (Fig. 1) it can be seen that miscibility gap has different value for different compositions i.e. the vertical interval of the miscibility gap of the alloy is larger when the alloy composition is equiatomic. Therefore Fe-rich phase has a relatively longer time to grow by diffusion and coarsen through collisions and coagulations between the Fe-rich spheres, owing to the spheres’ motion. The miscibility gap is ~20 K for the equiatomic composition, i.e. CFA50, and it increases as the copper content increases [11]. For CFA10 the miscibility gap is ~90 K [11]. It is relatively easy for the CFA50 alloy to phase separate compared to other two compositions. Therefore, the Fe phase in CFA50 grows by collision and coagulation with other Fe particles produced. In CFA10, the miscibility gap is relatively higher and this allows more time for CFA10 to phase separate and consequently the solidification proceeds much faster than phase separation. The shape of the Fe-rich phase for CFA10 is almost spherical and it changes to an irregular shape for CFA50. The mea- sured average size of Fe rich phase with composition and powder size is illustrated in Fig. 5. Therefore, the size of Fe-rich phase follows this sequence with composition:

Fig. 5

Average Fe-rich phase size in all the compositions of CFA.

CFA10 < CFA30 < CFA50.

Within the same composition, the size of Fe-rich phase increases as the powder size increases (Fig. 2-4). Mass effect plays a vital role in this phenomenon i.e. a smaller droplet cools faster and hence achieves a deeper undercooling [12]. Due to this faster cooling rate, it allows less time for the individual Cu and Fe phases to separate. Increase in the size of the Fe phase appears to be due to high solidification time available for large sized powders and the Fe phase grows by the collision, coagulation, and galloping of the smaller Fe phase particles.

Fig. 6, 7 and 8 show the grain size of CFA with composition and powder size. After comparison, it can be seen that with the increase in Fe content in the alloy melt, grain size of Cu matrix decreases. Fe phase has a grain refining effect on Cu phase. Among three compositions, CFA10 has relatively largest grain size as compared to other two compositions. The decrease in grain size with increase in Fe content also follow the same approach of phase separation between Cu and fe upon cooling. For CFA10, larger miscibility gap provides less time for phase separation and as a result the grain size is large. Whereas in other two compositions, due to relatively lower miscibility gap it becomes easy to phase separate. Fe phase size becomes large and it exerts the pushing force on the matrix phase upon cooling. Furthermore the pinning effect increases due to which grain size becomes small.

Fig. 6

Grain sizes distribution of Cu matrix phase in (a) p1, (b) p2, (c) p3 and (d) p4 samples of CFA10 respectively. Brighter areas correspond to Cu-rich phase and dark areas correspond to Fe rich phase.

Fig. 7

Grain sizes distribution of Cu matrix phase in (a) p1, (b) p2, (c) p3 and (d) p4 samples of CFA30 respectively. Brighter areas correspond to Cu-rich phase and dark areas correspond to Fe rich phase.

Fig. 8

Grain sizes distribution of Cu matrix phase in (a) p1, (b) p2, (c) p3 and (d) p4 samples of CFA50 respectively. Brighter areas correspond to Cu-rich phase and dark areas correspond to Fe rich phase.

Within the same composition, grain size increases with the increase in powder size (Fig. 6-8). In CFA10, p1 sample has a grain size of 5.45 μm whereas p2, p3 and p4 samples have grain sizes of 5.56 μm, 5.61 μm and 5.74 μm respectively. This is also probably due to the mass effect during solidification of the alloy. Smaller atomized droplets have relatively large undercooling; therefore, relatively lower time is available to phase separate. Fe size is relatively smallest so it can be said that grain refining effect is higher in smaller powder particles. Similarly p1, p2 p3 and p4 samples of CFA30 and CFA50 have grain sizes of 3.41 μm, 3.49 μm, 3.53 μm, 3.59 μm, 1.57 μm, 1.63 μm, 1.69 μm and 1.86 μm respectively. Fig. 9 illustrates the measured grain sizes of CFA with composition and powder size.

Fig. 9

Grain sizes distribution of Cu-rich matrix phase in all three compositions of CFA.

The XRD patterns of CFA10, CFA30 and CFA50 are shown in Fig. 10, 11 and 12 respectively. In these diffractograms of all compositions, fcc-Cu and bcc-Fe peaks can be seen. Cu and Fe in their elemental stages have fcc and bcc structures at room temperatures. So, it can be said that unlike mechanical alloying [6] gas atomization of CFA with atomizing gas pressure of ~ 20 bar does not produce atomic level alloying between the two elements. This result is also in agreement with the phase diagram. The only change that happens with composition is in intensity of bcc-Fe peaks that increase with Fe contents in the alloy. Peak positions of both Cu and Fe remain same in all the compositions regardless of composition and powder size. Lattice parameter was also computed for each composition with powder size and it was also found that lattice parameter remain essentially same.

Fig. 10

XRD pattern of CFA10. p1, p2, p3 and p4 are the powder sizes of the alloy used.

Fig. 11

XRD pattern of CFA30. p1, p2, p3 and p4 are the powder sizes of the alloy used.

Fig. 12

XRD pattern of CFA50. p1, p2, p3 and p4 are the powder sizes of the alloy used.

The only change that can be theoretically possible with gas atomization is the amount of solid solution. Cu and Fe have atomic radii of 1.28 Å and 1.26 Å (1 Å = 1 × 10^-10 m) respectively, the ≤15% difference in atomic radii satisfy the condition for formation of solid solution. Additionally the electro-negativities of both Cu and Fe are also comparable which are 1.90 and 1.83 respectively. Since the system was rapidly solidified from liquid state, where complete miscibility existed between Cu and Fe therefore it can be said that CFA system can contain random substitutional solid solution. From the phase diagram and XRD, it can be said that CFA10 will have relatively highest amount of solid solution whereas CFA50 will have relatively lowest. In between them is the solid solution for CFA30.

Conclusion

Copper-iron alloy (CFA) powder produced with gas atomization has composition and size dependent microstructure. Copper phase exist as matrix in all the three composition whereas Fe is distributed homogeneously in it. Size of Fe-rich phase increases with increase in Fe contents. Within the same composition, Fe-rich phase size increases due to mass effect. Grain size of matrix phase decreases with increase in Fe contents whereas it increases with increase in powder size for a given composition. XRD analysis has proved that there exists no atomic level alloying between Cu and Fe across the composition in CFA produced at 20 bar gas atomization pressure.

Further in-depth studies are required to completely assess the kinetics and thermodynamics of microstructural evolution with composition and powder size of CFAs produced at 20 bar atomization pressure.

• 1. C. Biselli and D.G. Morris: Acta Mater.., (1996) 44 493.Article
• 2. J.C. Crivello, T. Nobuki and T. Kuji: Mater. Trans.., (2008) 49 527.Article
• 3. G.A. Jerman, I.E. Anderson and J.D. Verhoven: Metall. Mater. Trans., A Phys. Metall. Mater. Sci.., (1993) 24 35.ArticlePDF
• 4. A. Munitz: Metall. Mater. Trans., B, Process Metall. Mater. Proc. Sci.., (1987) 18 565.ArticlePDF
• 5. J. Eckert, J.C. Holzer, C.E. Krill and W.L. Johnson: J. Appl. Phys.., (1993) 73 2794.ArticlePDF
• 6. C.G. Levi and R. Mehrabian: Metall. Mater. Trans., A Phys. Metall. Mater. Sci.., (1982) 13 221.ArticlePDF
• 7. P.C. Angelo and R. Subramanian, Powder Metallurgy: Science, Technology and Applications., New Delhi PHI Learning Private Limited, (2008) 217.
• 8. L.J. Swartzendruber, Binary Alloy Phase Diagrams., In : T.B. Massalski, editor. New York ASME, (1990) 1408.
• 9. V.A. Beloshenko, V.Yu. Dmitrenko and V.V. Chishko: Phys. Met. Metallogr.., (2015) 116 456.
• 10. J. He and J.Z. Zhao: Mater. Sci. Eng. A., (2005) 404 85.
• 11. J. He and J.Z. Zhao and et al: Gao: Metall. Mater. Trans. A., (2005) 36 2449.
• 12. J. He and J.Z. Zhao: J. Mater. Sci. Technol.., (2005) 21 759.
• 13. J. He, J.Z. Zhao and L. Ratke: Acta Mater.., (2006) 54 1749.

Figure & Data

References

Citations

Citations to this article as recorded by