Mechanical Dissolution of Cu5Zr Phase and Formation of Supersaturated Solid-Solution Nanocrystalline Structure by High-Pressure Torsion in a Hypoeutectic Cu–2.7 at%Zr Alloy

Kenta Miyamoto; Takahiro Kunimine; Chihiro Watanabe; Ryoichi Monzen; Reza Gholizadeh; Nobuhiro Tsuji

doi:10.2320/matertrans.MT-D2023008

Abstract

Microstructural evolution and changes in hardness and electrical conductivities of a cast hypoeutectic Cu–2.7 at%Zr alloy processed by high-pressure torsion (HPT) were investigated. The cast alloy had a net-like microstructure composed of a primary Cu phase and a eutectic consisting of layered Cu and Cu₅Zr phases. The Cu and Cu₅Zr phases in the eutectic had a cube-on-cube orientation relationship. The cast alloy with the hardness of 137 HV exhibited a value of electrical conductivity of 32%IACS. With increasing the number of HPT-revolutions, the eutectic was severely sheared and elongated along the rotational direction. In addition, mechanical dissolution of the Cu₅Zr phase into the Cu phase by the HPT was confirmed after 5 HPT-revolutions through XRD measurements and TEM observations. After 20 HPT-revolutions, the Cu phase was significantly refined and formed the lamellar structure having an average grain size of 15 nm. The electrical conductivity decreased and saturated at a value of 8%IACS after 50 HPT-revolutions. The significant decrease in the electrical conductivity was primarily attributable to the mechanical dissolution of the Cu₅Zr phase into the Cu phase by the HPT, followed by the formation of a nanocrystalline Cu–Zr supersaturated solid-solution alloy with the hardness of 430 HV.

This Paper was Originally Published in Japanese in J. Japan Inst. Copper 60 (2021) 98–103.

1. Introduction

As electronic devices have become smaller and more advanced in recent years, miniaturization of electronic components such as connectors has played an important role. Thus, Cu alloys as electronic materials are required to possess high strength and high electrical conductivity to realize miniaturization. The severe plastic deformation (SPD) is an efficient way to increase strength of metallic materials by downsizing grain size to submicron- or nano-scales and applying high dislocation density. Among the SPD processes, the high-pressure torsion (HPT) processing¹^,²⁾ can easily apply very large strain into materials without a change in sample dimensions compared to other SPD processes. In many single-phase Cu solid-solution alloys, as the stacking fault energy (SFE) decreases with increasing the concentrations of solid-solution atoms,³⁾ hardnesses and strengths of their alloys are significantly improved by the SPD processing. However, when these single-phase Cu solid-solution alloys are subjected to the SPD processing until the hardnesses and strengths are saturated, the attainable upper limit of hardness is about 300 HV⁴^,⁵⁾ at most; the attainable upper limit of tensile strength at room temperature (RT) is about 1000 MPa.⁵^,⁶⁾ On the other hand, some researchers have reported on achieving mechanical properties exceeding the upper limit values mentioned above by forming nanostructured microcomposites from two-phase Cu alloys through the SPD processing. Tian et al. reported that nanostructured microcomposites which had an elongated microstructure of the eutectic consisting of Cu and Ag phases oriented in the rotational direction of the HPT processing, with a tensile strength of 1420 MPa, was obtained by applying the HPT processing to an as-cast Cu–28 mass%Ag alloy.⁷⁾ Kormout et al. performed mechanical alloying on Cu–6, 37, and 84 at%Ag powder mixtures composed of Cu and Ag powders by the HPT processing at various temperatures until hardnesses of the mechanical alloyed samples saturated.⁸⁾ They have reported that single-phase supersaturated solid-solution alloys were obtained by the HPT processing performed at RT for Cu–6 and 84 at%Ag alloys, and at a temperature under liquid nitrogen for a Cu–37 at%Ag alloy, respectively.⁸⁾ In their study, the highest hardness of 386 HV was obtained in the Cu–6 at%Ag alloy.⁸⁾

In the present study, a cast hypoeutectic Cu–2.7 at%Zr alloy, composed of two phases: Cn and Cu₅Zr, was subjected to the HPT processing. Microstructural evolution of the hypoeutectic structure by the HPT processing and associated changes in hardness and electrical conductivity were investigated. Important experimental results reporting characteristics of microstructural evolution of Cu–Zr alloys caused by the SPD processing are as follows. Dobatkin et al. have reported that the electrical resistivity of a Cu–0.13 at%Zr alloy increased with increasing the number of HPT-revolutions.⁹⁾ They have explained that this increase in the electrical resistivity was caused by mechanical dissolution of Cu₅Zr particles as the second phase into the matrix in addition to the grain refinement.⁹⁾ Besides, Sun et al. prepared Cu–29, 38, and 47 at%Zr alloys by the accumulative roll-bonding (ARB) with stacked Cu and Zr sheets as an initial material, and performed the HPT processing on the ARB processed alloys.¹⁰⁾ They have reported that all of the alloys eventually became amorphous structure after the HPT processing.¹⁰⁾ In this study, based on these results, the HPT processing was applied to the hypoeutectic Cu–2.7 at%Zr alloy until the hardness of the alloy was saturated to reveal relationships among the final microstructure, hardness, and electrical conductivity of this two-phase alloy.

2. Experimental Procedures

A Cu–2.7 at%Zr alloy round-bar with a diameter of 14 mm produced by the vertical upward continuous casting (VUCC) was used for experiments. The round bar was cut to a one with a diameter of 10 mm using a wire electrical discharge machine (EDM). Then, disk-shaped samples with a diameter of 10 mm and a thickness of 0.8 mm were obtained by cutting the round bar using the EDM. These disk-shaped samples were subjected to the HPT processing with up to 100 HPT-revolutions under a condition of 6 GPa, 0.2 rpm, and RT. Hereafter, for example, a sample subjected to the HPT processing with 5 HPT-revolutions is referred to as a 5R sample.

For microstructural observations, a disk-shaped sample was cut at a distance from the center, r, of 4 mm as shown in Fig. 1(a). The center of the cross section of the cut sample was observed using a field-emission scanning electron microscope (FE-SEM: JSM-7100F, JEOL, Tokyo, Japan) to obtain backscattered electron (BSE) images. In the same way, a disk-shaped sample was cut into a strip at r = 4 mm, followed by polishing. Then, a thin film specimen was obtained from the polished strip using an ion slicer (EM-09100IS, JEOL, Tokyo, Japan), and observed using a transmission electron microscope (TEM: Tecnai G² 30, FEI, Hillsboro, OR, USA).

Fig. 1

Drawings showing the specimens and their positions in the HPT processed discs for (a) SEM and TEM observations and (b) electrical resistivity measurements.

Lattice constants of the various samples were also determined by the X-ray diffraction (XRD) using Cu-Kα₁ radiation (the wavelength λ = 0.15406 nm) in an X-ray diffractometer (RINT-2500, Rigaku Corporation, Tokyo, Japan). First, each lattice constant was obtained from the diffraction angles, 2θ, of the XRD peaks from the {111}, {200}, {220}, and {311} reflecting planes in a sample. Then, the lattice constant of the sample was determined by plotting the values of the lattice constants obtained from the {111}, {200}, {220}, and {311} reflecting planes with the Nelson-Riley function,¹¹⁾ and extrapolating to 2θ = π.

The Vickers microhardness was measured at RT using a hardness testing machine (HM-103, Akashi Corporation, Zama, Japan). Hardness tests were conducted at 375 µm intervals in the range of r = 0 to 4.5 mm in eight directions of the samples; average values of the Vickers microhardness at a distance from the center, r, were calculated by excluding the maximal and minimal values.

The electrical conductivity was determined by the following procedures. The electrical resistivity, ρ, of the each specimen was determined by the four-terminal method at RT using a resistance meter (RM3545, Hioki E.E. Corporation, Ueda, Japan). Using the electrical resistivity of pure Cu (ρ₀ = 17.2 nΩm), values of the electrical conductivity of each of the specimens were calculated by the relation: E = (ρ₀/ρ) × 100. Two types of specimens were prepared for the electrical resistivity measurements: a diametral specimen and an outer peripheral specimen as shown in Fig. 1(b).

3. Results

3.1 Microstructure of the as-cast alloy

Figure 2 shows an SEM-BSE image of the vertical cross-section of the as-cast Cu–2.7 at%Zr alloy. The as-cast alloy possessed a net-like eutectic structure; the dark area was composed of the primary Cu phase, and the light area was composed of the eutectic consisting of layered Cu and Cu₅Zr phases. The area ratio of the Cu phase to the Cu₅Zr phase in the as-cast alloy was 81:19 (the area ratio of the primary Cu phase to the Cu phase in the eutectic was 70:11). Figures 3(a) and (b) show a bright-field TEM image taken inside the eutectic in the as-cast alloy and the corresponding [112]_Cu selected-area diffraction pattern (SADP), respectively. The Cu₅Zr intermetallic compound has a face-centered cubic (fcc) structure,¹²⁾ and the diffraction spots from the Cu₅Zr phase were observed at inner side of all the diffraction spots from Cu phase in the SADP, indicating a cube-on-cube orientation relationship between the Cu phase and the Cu₅Zr phase in the eutectic. Besides, small spherical precipitates with an average size of about 7 nm were observed in the primary Cu phase and the Cu phase in the eutectic of the as-cast alloy. These spherical precipitates were composed of an ordered fcc structure with a lattice constant of 0.42 nm. It has been reported that these spherical precipitates possessed a cube-on-cube orientation relationship with the Cu matrix.¹³⁾ Hereafter, these spherical precipitates are referred to as fcc precipitates.

Fig. 2

SEM-BSE image of the vertical cross-section of the as-cast Cu–2.7 at%Zr alloy.

Fig. 3

(a) Bright-field TEM image of the as-cast Cu–2.7 at%Zr alloy; zone axis is parallel to [112]_Cu, (b) [112]_Cu selected-area diffraction pattern corresponding to (a).

3.2 Microstructural evolution of the hypoeutectic structure caused by the HPT processing

Figures 4(a) to (f) depict SEM-BSE images of the Cu–2.7 at%Zr alloy taken at a distance of 4 mm from the center of the HPT processed disk-shaped samples (r = 4 mm) with 0.5, 1, 5, 10, 20, and 50 HPT-revolutions, respectively. In the early stage of the HPT processing (the 0.5R and 1R samples), the microstructure of the net-like eutectic structure observed in the as-cast alloy was severely shear-deformed and fragmented; the fragmented eutectic structures began to be oriented in the rotational direction (the directions of the double-sided arrows in Fig. 4). The layered structure of Cu and Cu₅Zr phases observed in the eutectic of the as-cast alloy was still maintained in many fragmented eutectic structures. In the 5R sample, the eutectic structures were elongated in the rotational direction and became a short-fiber-like structure. The observed eutectic structures were slightly obscured because the Zr atoms in the Cu₅Zr phase began to mechanically dissolve into the Cu phase by the HPT processing, as described later. In the 10R and 20R samples subjected to further shear deformation by the HPT processing, as the mechanical dissolution of the Zr atoms in the Cu₅Zr phase into the Cu phase progressed, the Cu phase and eutectic structures were further elongated, resulting in the formation of a nanoscale lamellar grain structure. Figures 5(a) and (b) depict a bright-field TEM image observed at r = 4 mm in the 20R sample and the corresponding SADP, respectively. In the SADP of Fig. 5(b), the amorphous structure was not confirmed, and diffraction spots of the Cu₅Zr intermetallic compound were not observed as well. By defining the grain size as the grain boundary spacing parallel to the elongation direction, it was found that the average grain size of the Cu phase observed at r = 4 mm in the 20R sample was about 15 nm; the grain size was refined to the nanoscale by the HPT processing. In the 50R sample, the SEM-BSE image showed almost no contrast, indicating that a compositionally uniform microstructure was formed.

Fig. 4

SEM-BSE images of the Cu–2.7 at%Zr alloy taken at a distance of 4 mm from the center of the HPT processed disk-shaped samples with (a) 0.5, (b) 1, (c) 5, (d) 10, (e) 20, (f) 50 HPT-revolutions. The double-sided arrows indicate the rotational direction.

Fig. 5

(a) Bright-field TEM image of the HPT processed disk-shaped sample with the 20 HPT-revolutions, (b) selected-area diffraction pattern corresponding to (a).

3.3 Variation of the Vickers microhardness caused by the HPT processing

Figure 6 shows relationships between the average Vickers microhardness, H, and the distance from the center, r, in the HPT-processed disk-shaped samples. The H of the as-cast alloy was 137 HV. With increasing the number of the HPT revolutions, the H at each position in the samples were increased. In the 0.25R to 1R samples, the differences between the values of H around the center and the outer periphery of the samples were not so large. However, the H around the outer periphery of the 5R sample increased significantly; the difference between the values of H around the center and the outer periphery of the 5R sample became more significant. The H of the 20R sample saturated to about 430 HV in the range of r ≥ 2.25 mm. In the 50R sample, the H around the center almost reached the saturation value as well; the H became almost uniform in the entire area of the 50R sample. In the 100R sample subjected to further shear deformation by the HPT processing, the values of H were same with the ones measured in the 50R sample.

Fig. 6

Average Vickers microhardness vs. the distance from the center in the HPT-processed disk-shaped samples of the Cu–2.7 at%Zr alloy with various number of HPT-revolutions.

3.4 Variation of the electrical conductivity caused by the HPT processing

Figure 7 presents the electrical conductivity, E, of the diametral and outer peripheral specimens as a function of the number of HPT-revolutions. In the 0.25R to 1R samples, both the diametral and outer peripheral specimens possessed higher values of E than those in the as-cast alloy; the outer peripheral specimen of the 1R sample exhibited the electrical conductivity, E, of 42%IACS. After that, the E of the outer peripheral specimen rapidly decreased in the 5R sample, and became constant at about 8%IACS in the 10R to 50R samples. The E of the diametral specimens in the 5R to 20R samples gradually decreased and approached to the minimal value of E measured in the outer peripheral specimen. The E of the diametral specimen in the 50R sample finally became almost equal to the one measured in the outer peripheral specimen of the 50R sample.

Fig. 7

Electrical conductivity of the diametral and outer peripheral specimens vs. the number of HPT-revolutions.

4. Discussion

In this section, the reasons of the changes in the electrical conductivity and microhardness are discussed in terms of the microstructural evolution caused by the HPT processing. The electrical conductivity of alloys is affected by the dislocation density, the grain boundary density, the concentrations of solute atoms, and the presence of a second phase, etc.¹⁴⁾ Linde has reported that the amount of increase in the resistivity of Cu by adding Zr solute atoms was 110 nΩm/at%.¹⁵⁾ The concentrations of solute atoms significantly affect the electrical conductivity compared to other factors. First, the improvement in the electrical conductivity of the 0.25R to 1R samples, compared to the as-cast alloy, is considered. The microstructure of the as-cast alloy was the net-like eutectic structure. On the other hand, in the microstructure of the 1R sample shown in Fig. 4(b), the severely shear-deformed and fragmented eutectic structure began to be oriented in the rotational direction. It can be understood that the fragmented eutectic structures oriented in the rotational direction led to the improvement in the electrical conductivity due to the partially removed barrier, composed of the Cu₅Zr phase which possessed a low electrical conductivity, for the flow of electrons. The reason why the increase in electrical conductivity of the diametral specimen was smaller than that of the outer peripheral specimen can be considered as follows. The amount of strain introduced around the center of the HPT disc-shaped sample was smaller than that introduced around the outer periphery. Therefore, the microstructural evolution around the center of the sample was delayed, resulting in retardation of fragmenting the net-like eutectic. These results led to the hindering of the flow of electrons by the barrier composed of the Cu₅Zr phase.

Next, let us consider why the electrical conductivity decreased after applying the HPT processing with over 5 HPT-revolutions. The fragmented eutectic structures were elongated and developed to a lamellar structure consisting of the primary Cu phase and eutectic in the 5R, 10R, and 20R samples. Besides, the observed SEM-BSE images became more and more obscure as the number of HPT-revolutions increased. Figure 8 shows the XRD line profiles of the as-cast alloy, 5R and 50R samples. In the 5R sample, the Cu₅Zr peak observed in the as-cast alloy disappeared; the peaks of the Cu phase shifted to the lower diffraction angles. In the 50R sample, the peaks of the Cu phase shifted further to the lower diffraction angles than that in the 5R sample. The diffraction spots of the Cu₅Zr phase were not observed in the SADP shown in Fig. 5(b), indicating that most of the Cu₅Zr phase disappeared in the 20R sample. Therefore, the changes in the lattice constant, a, of the Cu phase caused by the HPT processing was determined from the XRD line profile analyses. Then, the concentration of Zr solutes, C_Zr, in the Cu phase was calculated using the Vegard’s law¹⁶⁾ expressed as the following eq. (1).

\begin{equation} a = (1 - C_{\text{Zr}})a_{\text{Cu}} + C_{\text{Zr}}a_{\text{Zr}} \end{equation}

(1)

Here, a_Cu is the lattice constant of Cu (a_Cu = 0.3615 nm), and a_Zr is the lattice constant of Zr with fcc structure (a_Zr = 0.4520 nm).¹⁷⁾ Figure 9 shows the lattice constant, a, of the Cu phase and the concentration of Zr solutes, C_Zr, in the Cu phase, as a function of the number of HPT-revolutions. According to the Cu–Zr binary phase diagram at equilibrium state, the maximal solid-solubility limit of Zr solutes in Cu is about 0.12 at% at the eutectic temperature.¹³⁾ In the as-cast alloy, a = 0.3617 nm, and C_Zr = 0.2 at% were obtained. This C_Zr value is approximately in agreement with the maximal solid-solubility limit shown in the literature. The values of a obtained in the 0.25R, 0.5R, and 1R samples were almost unchanged compared to the as-cast alloy. However, the value of a increased significantly to 0.3625 nm in the 5R sample, and the value of C_Zr was estimated to be 1.1 at%. This value of C_Zr is much higher than the maximal solid-solubility limit of Zr solutes in Cu of 0.12 at%. Furthermore, the values of a and C_Zr increased with increasing the number of HPT-revolutions after applying the HPT processing with over 5 HPT-revolutions: a = 0.3637 nm and C_Zr = 2.4 at% in the 50R sample, and a = 0.3638 nm and C_Zr = 2.6 at% in the 100R sample. These results indicate that most of the Zr atoms in the alloy were mechanically dissolved into the Cu phase by the HPT processing. As described in the section 3.1, the Cu phase and Cu₅Zr phase inside the eutectic, and the Cu phase and fcc precipitates possessed a cube-on-cube orientation relationship, respectively. Therefore, it can be understood that dislocations were repeatedly passed through the Cu₅Zr phase and fcc precipitates due to the shear deformation during the HPT processing, resulting in the mechanical dissolution of the Zr atoms in the Cu₅Zr phase and fcc precipitates into the Cu phase. By considering these results, it can be understood that the reason for the decrease in the electrical conductivity of the specimens with more than 5 HPT-revolutions is due to the mechanical dissolution of the amount of Zr atoms, which was significantly exceeded the solid-solubility limit, into the Cu phase by the HPT processing. As shown in Fig. 9, the mechanical dissolution of Zr atoms into the Cu phase continued up to 100 HPT-revolutions. However, the electrical conductivity at the outer periphery of the HPT processed sample was already saturated after 10 HPT-revolutions. Although the electrical conductivity of the Cu phase decreases due to the solid-solution of Zr atoms in the Cu phase, the Cu₅Zr phase with low electrical conductivity disappears as the mechanical dissolution of Zr atoms progresses by the HPT processing. Therefore, in the case of applying the HPT processing with more than 10 HPT-revolutions, the decrease in electrical conductivity of the Cu phase associated with the progress of HPT processing was counterbalanced by the increase in electrical conductivity due to the disappearance of the Cu₅Zr phase; the electrical conductivities of the specimens as a whole became constant.

Fig. 8

X-ray diffraction profiles of the Cu–2.7 at%Zr alloy of the as-cast, and HPT processed samples with 5 and 50 revolutions.

Fig. 9

Lattice constant and concentration of Zr solutes in the Cu phase vs. the number of HPT-revolutions.

It has also been reported that the minimal grain size of the metals and alloys obtained by the SPD processing became smaller as the concentration of solid-solution atoms in the Cu alloy increased.⁴^,⁶⁾ In particular, since the solid-solution of Zr atoms in Cu markedly increased the recovery and recrystallization temperature,¹⁸⁾ it is assumed that Cu–Zr alloys can be introduced higher dislocation density than other Cu alloys by the SPD processing; it can be judged that the average grain sizes of the Cu–2.7 at%Zr alloy were markedly refined after the SPD processing. Furthermore, it can be understand that dissolving Zr atoms mechanically by the HPT processing to supersaturated solid-solution, which significantly exceeded the solid-solubility limit, led to the grain refinement of the Cu phase to the nanocrystalline scale of about 15 nm, resulting in extremely high hardness as a Cu alloy. In spite of the continued mechanical dissolution of Zr atoms into the Cu phase, the hardnesses at the outer periphery of the samples were already saturated after 20 HPT-revolutions. These results can be judged to be caused by a tradeoff relationship between the solid-solution strengthening associated with the mechanical dissolution of Zr atoms and the disappearance of the composite strengthening or dispersion strengthening by the Cu₅Zr phase.

5. Conclusions

Microstructural evolution and changes in hardness, and electrical conductivity of the as-cast Cu–2.7 at%Zr alloy subjected to the HPT processing were investigated. The obtained results can be summarized as follows.

(1) The net-like eutectic structure was fragmented and began to be oriented in the rotational direction by 1 HPT-revolution. In the 5R sample, the eutectic structure was fragmented and became the short-fiber-like structure. Besides, the mechanical dissolution of Zr atoms, which were located inside the Cu₅Zr phase and fcc precipitates, into the Cu phase began to be confirmed. Furthermore, in the 10R and 20R samples subjected to further shear deformation by the HPT processing, the Cu phase and fragmented eutectic structures were elongated in the rotational direction, and formed a nanoscale lamellar grain structure. The average grain size of the Cu phase in the 20R sample was about 15 nm. The grain size was reduced to the nanocrystalline scale by the HPT processing. In the 100R sample, it can be judged that most of the Zr atoms were mechanically dissolved into the Cu phase, and the Cu₅Zr phase was disappeared.
(2) By the HPT processing, the microhardness was significantly improved compared with the one of the as-cast alloy, reaching a saturated microhardness of about H = 430 HV in almost the entire area of the 50R sample. On the other hand, the electrical conductivity significantly decreased, and reached the minimal value of E = 8%IACS in the 50R sample. This decrease in electrical conductivity was attributed to the mechanical dissolution of almost the Zr atoms in the alloy to the Cu phase caused by the repeated dislocation glides through the Cu₅Zr phase in the eutectic and fine fcc precipitates in the Cu phase by shear deformation during the HPT processing.

Acknowledgments

This research was supported by a Grant-in-Aid for the 2020 academic year from Japan Institute of Copper. This work was also supported in part by a Grant-in-Aid for the 2020 academic year (No. 0321028-A) from Iketani Science and Technology Foundation. The authors are grateful to Dr. Naokuni Muramatsu, NGK INSULATORS, LTD., for providing the as-cast Cu–Zr alloys.

REFERENCES

Corresponding author

Register with J-STAGE for free!