I. INTRODUCTION
Owing to the excellent combination of high electrical conductivity and strength, copper (Cu)–beryllium (Be) alloys are used for numerous applications such as lead frames, electrical connectors, elastic components, and precise instruments (Wilkes, Reference Wilkes1968). To avoid environmental pollution because of the toxic character of Be, advanced metallic materials are required to be developed to replace Cu–Be alloys. Therefore, worldwide research has been aimed at developing a substitute for the noxious and costly Cu–Be alloys. Age-hardenable copper–titanium (Cu–Ti) alloys containing approximately 1–6 at.% Ti are attractive as a candidate since their mechanical strength is comparable with Cu–Be alloys with higher wear resistance and superior stress relaxation behavior (Datta and Soffa, Reference Datta and Soffa1976; Nagarjuna et al., Reference Nagarjuna, Srinivas, Balasubramanian and Sarma1999; Soffa and Laughlin, Reference Soffa and Laughlin2004). Unfortunately, their electrical conductivity is rather low (Nagarjuna et al., Reference Nagarjuna, Balasubramanian and Sarma1997; Suzuki et al., Reference Suzuki, Hirabayashi, Shibata, Mimura, Isshiki and Waseda2003).
To obtain a good balance between conductivity and strength, much work has been done (Nagarjuna et al., Reference Nagarjuna, Srinivas, Balasubramanian and Sarma1999; Kamegawa et al., Reference Kamegawa, Iwaki and Okada2010, Reference Kamegawa, Kuriiwa and Okada2013; Semboshi, Reference Semboshi2007; Semboshi et al., Reference Semboshi, Al-Kassab, Gemma and Kirchheim2009, Reference Semboshi, Nishida, Numakura, Al-Kassab and Kirchheim2011a, Reference Semboshi, Orimo, Suda, Gao and Sugawara2011b). Semboshi et al. (Reference Semboshi, Nishida, Numakura, Al-Kassab and Kirchheim2011a, Reference Semboshi, Orimo, Suda, Gao and Sugawara2011b) made a systematic investigation of the Cu–Ti alloys and found that the electrical conductivity and the strength of Cu–(1–6) at.% Ti alloy increased greatly when aging in a hydrogen (H2) atmosphere. Moreover, they demonstrated that good balance between conductivity and strength can be obtained by aging at low temperature and high H2 pressure and by prior deformation of cold rolling (Semboshi et al., Reference Semboshi, Nishida, Numakura, Al-Kassab and Kirchheim2011a, Reference Semboshi, Orimo, Suda, Gao and Sugawara2011b). Recently, high-strength and high-conductivity Cu–4.2 mol% Ti alloy wire was fabricated by Semboshi and Takasugi (Reference Semboshi and Takasugi2013).
Mechanical properties of Cu–Ti alloys are limited to intermediate temperatures up to 723 K. Unfortunately, these properties drop down at higher temperatures because of the coarsening of the β-phase precipitates with the nominal composition of Cu4Ti (Kato et al., Reference Kato, Mori and Schwartz1980; Ardell, Reference Ardell1985). As an important kind of alloying additive for metallic materials, rare earth (RE) can significantly improve the properties of alloys by affecting the microstructure and refining grain. To discover further application characteristics and regularities concerning phase formation in the Cu–Ti–RE ternary system, it is necessary to investigate phase relationships in this system. Up to now, reports on the Cu–Ti–RE phase diagrams are limited, except the Cu–Ti–Y (Hu et al., Reference Hu, Zhan, She, Zhang and Peng2009) and Cu–Ti–Er ternary systems (Zhan et al., Reference Zhan, Peng and She2012). The work presented in this paper is aiming to determine the Cu–dysprosium (Dy)–Ti phase equilibria at 973 K.
II. BINARY SYSTEMS
A. Cu–Ti binary system
For the Cu–Ti binary system, the existence of six intermediate phases, i.e., CuTi2, CuTi, Cu4Ti3, Cu3Ti2, Cu4Ti, and Cu2Ti is accepted without question. But the existence of the phase CuTi3 and the stability range of the phase Cu2Ti is controversial. Eremenko et al. (Reference Eremenko, Buyanov and Prima1966) reported that the temperature for the invariant reaction Cu2Ti ↔ Cu4Ti + Cu3Ti2 was 1123 K; however, this temperature suggested by Murray (Reference Murray1983) was about 1143 K. Canale and Servant (Reference Canale and Servant2002) suggested that CuTi3 was a stable phase on the basis of differential thermal analysis results. Kumar et al. (Reference Kumar, Ansara, Wollants and Delaey1996) also reported the same results when thermodynamically evaluated the Cu–Ti system. After that, Karlsson (Reference Karlsson1951) reported the possible existence of CuTi3. Recently, Zhan et al. (Reference Zhan, Peng and She2012) confirmed the CuTi3 in the microstructure of the alloy 25Cu75Ti. They heated the 25Cu75Ti alloys to 1023 K and then kept warm for 90 h. The temperature of the eutectoid transformation, namely, βTi↔αTi + CuTi3, was determined to be 1078 K in their work. They suggested that the formation of the CuTi3 phase was controlled by the annealing temperature and annealing time. At the same time, it should be a metastable phase in light of the works carried out by the other groups (Massalski et al., Reference Massalski, Murray, Bennett and Baker1986; Xu et al., Reference Xu, Du, Huang and Liu2005; Liu et al., Reference Liu, Wang, Zhang, Chen, Zheng and Jin2006a; Wang et al., Reference Wang, Liu, Zhang, Zheng and Jin2006)
B. Cu–Dy binary system
For the Cu–Dy binary system, much work has been done so far (Baenziger and Moriarty, Reference Baenziger and Moriarty1961; Storm and Benson, Reference Storm and Benson1963; Copeland and Kato, Reference Copeland and Kato1964; Buschow et al., Reference Buschow, Van Der Goot and Birkhan1969; Buschow and Van Der Goot, Reference Buschow and Van Der Goot1971; Zheng and Xu, Reference Zheng and Xu1982; Franceschi, Reference Franceschi1982; Subramanian and Laughlin, Reference Subramanian and Laughlin1988; Zhang et al., Reference Zhang, Huang, Qi, Jia, Liu and Jin2009). The first systematical study of the Cu–Dy system was performed experimentally by Zheng and Xu (Reference Zheng and Xu1982). At the same time, the Cu–Dy system was also investigated experimentally by Franceschi (Reference Franceschi1982). Zheng and Xu (Reference Zheng and Xu1982) reported that the maximum solid solubility of Cu in Dy was 12.5 at.% Cu, while Franceschi (Reference Franceschi1982) reported that there was no appreciable solid solutions of Cu in Dy. Subramanian and Laughlin (Reference Subramanian and Laughlin1988) assessed the Cu–Dy system and most of their work was accepted except the stable temperature range of Cu7Dy phase and the solid solubility of Cu in Dy. Zhang et al. (Reference Zhang, Huang, Qi, Jia, Liu and Jin2009) thought that the result of Zheng and Xu (Reference Zheng and Xu1982) was better than that of Franceschi (Reference Franceschi1982) because the properties of Gd and Dy were quite similar and the solid solubility of Cu in Gd was 15 at.% (Carnasciali et al. Reference Carnasciali, Cirafici and Franceschi1983a). To obtain a convincible result, Zhang et al. (Reference Zhang, Huang, Qi, Jia, Liu and Jin2009) reassessed the Cu–Dy system and found that the maximum terminal solubility of Cu in Dy was 12 at.% Cu.
CuDy, Cu2Dy, and Cu5Dy have been confirmed so far. About the Cu5Dy, Buschow et al. (Reference Buschow, Van Der Goot and Birkhan1969) found Cu5Dy_H and Cu5Dy_L phases. For the Cu7Dy, much investigation confirmed that it was a high-temperature phase, but the range of the temperature was ambiguous. Buschow and Goot (Reference Buschow and Van Der Goot1971) found that the Cu7Dy was not stable below 973 K, while Zheng and Xu (Reference Zheng and Xu1982) found that the temperature range of existence of Cu7Dy was from 1121 to 1163 K. However, later in Franceschi’ s work (Reference Franceschi1982), this compound was reported to be stable between 1048 and 1133 K. So to obtain a more reliable result, Zhang et al. (Reference Zhang, Huang, Qi, Jia, Liu and Jin2009) reassessed the Cu–Dy system and they found that the temperature range of existence of Cu7Dy was from 1054 to 1174 K.
To resolve the differences, the solid solubility of Cu in Dy and the range of the temperature of Cu7Dy are restudied in this work, which will be discussed in the phase analysis part.
C. Dy–Ti binary system
The previous work has indicated that no binary compound exists in the Dy–Ti system (Zhuang et al., Reference Zhuang, Huang and Li1996; Bulanova et al., Reference Bulanova, Podrezov, Fartushnaya, Meleshevich and Samelyuk2004; Liu et al., Reference Liu, Ding, Qian, Zhuang and Huang2006b; Yan et al., Reference Yan, Liu, Huang, Long, Zhuang, Li and Baenziger2009). Linus Pauling file entry prototype of the binary Cu–Ti and Cu–Dy phases are summarized in Table I. Additionally, the lattice parameters calculated in this work are also showed in Table I.
Table I. Binary crystal structures in the Cu–Dy–Ti system quenched from 973 K.
III. EXPERIMENTAL DETAILS
The alloy samples were produced by arc melting on a water-cooled Cu cast with a non-consumable tungsten electrode under pure argon atmosphere. Titanium was used as O2 getter during the melting process. The sample was prepared with a total weight of 2 g and the purity of Ti, Cu, and Dy were all 99.9 wt.%. The arc-cast button was turned around after each melting and remelted at least three times for better homogeneity. In total, 75 samples were prepared and the weight losses were almost all <1% after melting. The melted alloy buttons were sealed in small glass tubes for homogenization heat treatment. The homogenization temperatures of the samples were determined according to the phase diagrams of the Cu–Ti, Dy–Ti, and Cu–Dy systems and the results of differential scanning calorimetry (DSC) of some typical ternary alloys. The alloys were annealed at 973 K for more than 1 month. Finally, the samples were quenched in ice-water.
Room temperature X-ray powder diffraction (XRD) and scanning electron microscopy (SEM) with energy dispersive analysis (EDX) were used in the present investigation. Samples for the XRD analysis were initially placed in a metal mortar and the pestle was used to break apart the sample and then repeated grinding, until it becomes powder (the finer, the better). Usually, the mean particle size should be 30–50 μm. The samples were measured with the help of Rigaku D/Max 2500 V diffractometer with CuKα radiation and graphite monochromator operated at 40 kV and 200 mA. The scan ranges of the samples were from 20° to 60° (2θ) with a scanning speed of 10° min−1. The Materials Data Inc. software Jade 5.0, PCW (powder Cell Windows) software and PDF-2 2010 Inorganic (Dr. Soorya Kabekkodu, Reference Kabekkodu2010) were used for phase identification. For the SEM/EDX analyses, the alloys were prepared by following the standard metallographic procedures: hot mounting in resin, grinding in the sequence of no. 400–3000 SiC sand paper, and polishing with colloidal silica suspension (OP-S). DSC experiments were performed by DSC. The DSC curve (heat flow vs. temperature) of alloy Cu7Dy was recorded at a rate of 10 K min−1 under a flow of pure argon on the NETZSCH STA 449 C instrument with the crucible-type DSC/TG. Pan A12O3 and TAB separator are used, and the temperature was calibrated by pure aluminum. By all these means, the phase relations of the Cu–Dy–Ti ternary system were determined.
IV. RESULT AND DISCUSSION
A. Phase analysis
In the Cu–Ti system, the existence of six compounds, i.e., CuTi3, CuTi2, CuTi, Cu4Ti3, Cu3Ti2, and Cu4Ti has been confirmed at 973 K. Canale and Servant (Reference Canale and Servant2002) suggested that the phase CuTi3 was a stable phase on the basis of their experimental results, though it was generally regarded as a metastable phase. Fortunately, in our work, CuTi3 was observed from the XRD patterns of the equilibrated alloys with composition equal to or near to Cu:Ti = 1: 3. Furthermore, the XRD pattern of #18 sample (44 at.% Cu, 18 at.% Dy, and 38 at.% Ti) clearly indicates the existence of CuTi3 (25-1144), CuTi2 (72-0441), and Cu2Dy (39-1344), as shown in Figure 1, which is in accordance with the previous reports from Zhan et al. (Reference Zhan, Peng and She2012) and Eremenko et al. (Reference Eremenko, Buyanov and Prima1966). The XRD pattern of #19 sample (57 at.% Cu, 20 at.% Dy, and 23 at.% Ti) shows the existence of CuTi (07-0114), CuTi2 (72-0441), and Cu2Dy (39-1344), as shown in Figure 2.
Figure 1. (Color online) The XRD pattern of #18 sample (44 at.% Cu, 18 at.% Dy, and 38 at.% Ti) indicating the existence of CuTi3, CuTi2, and Cu2Dy.
Figure 2. (Color online) The XRD pattern of #19 sample (57 at.% Cu, 20 at.% Dy, and 23 at.% Ti) indicating the existence of CuTi, CuTi2, and Cu2Dy.
In the Cu–Dy binary system, CuDy, Cu2Dy, and Cu5Dy have been confirmed at 973 K, which can be seen from the XRD pattern or the SEM micrograph in Figures 1 and 2 mentioned above and in the following Figures 5, 6 and 8.
Figure 3 shows the XRD pattern of the equilibrated alloy of the Cu1Dy99. It is obviously seen that CuDy can be detected, which reveals that Cu does not show remarkable solubility in Dy, which is in agreement with the result from Franceschi (Reference Franceschi1982).
Figure 3. (Color online) The XRD pattern of the equilibrated alloy of the Cu1Dy99.
The DSC curve (heat flow vs. temperature) of alloy Cu7Dy is shown in Figure 4. The change in the slope of the heat flow temperature curve can be observed. The temperature of the extrapolated onset is used as the phase transformation temperature in this work, so T 1, T 2, and T 3 represent three phase transformation temperatures, respectively. An obvious endothermic peak related to the eutectic transformation L ⇔ Cu + Cu7Dy is observed at T 2 = 1135 K. The temperature T 1 is related to the phase transition temperature of the eutectoid reaction, namely, Cu7Dy ⇔ Cu + Cu5Dy_L at 1112 K. The temperature T 3 is related to the peritectic transition action L + Cu5Dy_L ⇔ Cu7Dy. The temperature range of the existence of Cu7Dy is from 1112 to 1183 K, which agrees well with the work of Zheng and Xu (Reference Zheng and Xu1982). Table II shows the reactions in this work and the results from previous references (Zheng et al., Reference Zheng and Xu1982; Franceschi, Reference Franceschi1982; Subramanian et al., Reference Subramanian and Laughlin1988).
Figure 4. (Color online) DSC heating curve of the alloy Cu7Dy.
Table II. Invariant reactions of the Cu7Dy in the Cu–Dy system.
Figure 5 shows the XRD pattern of #1 sample (25 at.% Cu, 50 at.% Dy, and 25 at.% Ti), illustrating the existence of Ti (44-1294), Dy (89-2926), and CuDy (65-4143). The SEM photograph in Figure 6 also clearly indicates the existence of the above phases (identified by EDX). In the Dy–Ti system, it is confirmed that no binary compound exists in this work, which is in good agreement with the results from (Zhuang et al., Reference Zhuang, Huang and Li1996; Bulanova et al., Reference Bulanova, Podrezov, Fartushnaya, Meleshevich and Samelyuk2004; Liu et al., Reference Liu, Ding, Qian, Zhuang and Huang2006b; Yan et al., Reference Yan, Liu, Huang, Long, Zhuang, Li and Baenziger2009).
Figure 5. (Color online) The XRD pattern of #1 sample (25 at.% Cu, 50 at.% Dy, and 25 at.% Ti) containing Ti, Dy, and CuDy.
Figure 6. SEM micrograph of the equilibrated alloy 25 at.% Cu, 50 at.% Dy, 25 at.% Ti containing Ti, Dy, and CuDy. (The white areas relate to the Dysprosium oxide.)
In the previous work, no ternary compound has been reported in the ternary Cu–Dy–Ti system. This result is also confirmed in the present work (at 973 K).
B. Isothermal section
The isothermal section of the Cu–Dy–Ti ternary system at 973 K has been determined on the basis of XRD and SEM, as shown in Figure 7. This isothermal section consists of ten ternary phase regions, 21 binary phase regions, and 12 single-phase regions. Two ternary phase regions, i.e., CuTi2 + CuTi3 + Cu2Dy and CuTi3 + Ti + Cu2Dy, are supposed based on the stability of the CuTi3 phase. The XRD pattern of #6 sample (40 at.% Cu, 26 at.% Dy, and 34 at.% Ti) illustrates the existence of Ti (44-1294), CuDy (65-4143), and Cu2Dy (39-1344), as shown in Figure 8.
Figure 7. (Color online) The phase diagram (with the present experimental data) of the Cu–Dy–Ti ternary system at 973 K.
Figure 8. (Color online) The XRD pattern of #6 sample (40 at.% Cu, 26 at.% Dy, and 34 at.% Ti) illustrates the existence of Ti, CuDy, and Cu2Dy.
The constitutions of the ternary phase regions and compositions of the typical alloys are listed in Table III. The XRD results confirm that nine binary compounds, namely Cu4Ti, Cu3Ti2, Cu4Ti3, CuTi, CuTi2, CuTi3, CuDy, Cu2Dy, and Cu5Dy exist in this system at 973 K.
Table III. Details of the phase regions and typical samples in the Cu–Dy–Ti system at 973 K.
The solid solubility ranges in the isothermal sections are determined using the phase-disappearing method and comparing the XRD patterns of samples near the compositions of the binary phases. Variation of the lattice parameters was determined to obtain the solid solubility. None of the other phase found in this system has a remarkable homogeneity range at 973 K.
V. CONCLUSIONS
In this work, the nine binary compounds, i.e., Cu4Ti, Cu3Ti2, Cu4Ti3, CuTi, CuTi2, CuTi3, CuDy, Cu2Dy, and Cu5Dy have been confirmed. No binary compound is found in the Dy–Ti binary system at 973 K. The isothermal section of the Cu–Dy–Ti ternary system at 973 K consists of 12 single-phase regions, 21 binary phase regions, and ten ternary phase regions. The temperature range of Cu7Dy is from 1112 to 1183 K. The solid solubility of Cu in Dy is undetectable. None of the other phase in this system shows a remarkable solid solution at 973 K. No ternary compound is found in the present work.
ACKNOWLEDGMENTS
This research work is supported by the National Natural Science Foundation of China (Grant numbers 51161002 and 51361002), the Program for New Century Excellent Talents in the University of China (No. NCET-12-0650), the Guangxi Science and Technology Development Project (No. 11107003-1, 12118001-2B), and the Science and Technology Development Project of Guangxi Department of Education (No. 2013ZL010).
