2023 Volume 64 Issue 10 Pages 2530-2534
Pure copper, specifically oxygen-free copper (OFC), is widely used in superconductive and low-temperature refrigeration technologies owing to its superior electrical and thermal conductive properties at cryogenic temperatures. These properties, which can be expressed in terms of the residual resistivity ratio (RRR), are associated with the purity of copper or with its impurity concentration. High-purity copper with a low impurity concentration exhibits a high RRR. In addition to the impurity concentration, the existing form of impurities within the matrix affects the RRR, which is influenced by heat treatment. Specifically, for OFC, high-temperature heat treatment causes the impurities to dissolve into the matrix, resulting in a decrease in the RRR. Thus, to obtain a high RRR, the impurity concentration must be reduced, which often requires complex purification processes. In this study, we aimed to develop a pure copper material with a high RRR over a wide range of heat treatment temperatures using industrially feasible methods with OFC as the base material. For this, impurities with a negative effect on the RRR were investigated, and an additive element was selected to target the impurities and minimize the negative effect of these impurities without adversely affecting the RRR. We successfully developed a pure copper material using an extremely small amount of Ca as an additive element. This material not only exhibited an RRR comparable with that of high-purity copper but also maintained the high RRR value over a wide range of heat treatment temperatures, owing to the formation of CaS. The high RRR pure copper developed in this study is a promising material for use in components employed in superconductor and refrigeration applications such as in magnetic resonance imagining (MRI), in nuclear magnetic resonance (NMR), fusion reactors, maglevs, and particle accelerators.
This Paper was Originally Published in Japanese in Journal of Japan Institute of Copper 61 (2022) 329–333. Table 1, 2, Figs. 2, 4–6 and the caption of Table 1, Figs. 1–5 were slightly modified.
Fig. 4 RRR of Cu-17 atomic ppm Ca and OFC.
Pure copper, specifically oxygen-free copper (OFC), exhibits high electrical and thermal conductivity at cryogenic temperatures and is widely used in applications that are operational at such temperatures including superconductivity, refrigeration, and other technologies. Conductivity at cryogenic temperatures can be expressed in terms of the residual resistivity ratio (RRR). As indicated in eq. (1), the RRR equals to the ratio of the resistivity ρ293 K of a material at room temperature (293 K) to its resistivity ρ4 K at the liquid helium temperature (4 K).
| \begin{equation} \mathrm{RRR} = \rho_{\text{293$\,$K}}/\rho_{\text{4$\,$K}} \end{equation} | (1) |
The total resistivity of metals (ρtotal) can be expressed using Matthiessen’s rule, as shown in eq. (2), where ρ(L) denotes the resistivity based on phonon scattering (lattice vibration), ρ(D) denotes the resistivity based on lattice defects (dislocations and grain boundaries), and ρ(I) denotes the resistivity based on impurities.1,2)
| \begin{equation} \rho_{\textit{total}} = \rho(L)+\rho(D)+\rho(I) \end{equation} | (2) |
At room temperature, ρ(L), with a value of 1.676 × 10−8 Ωm,2) is the main resistivity component. At cryogenic temperatures, ρ(L) is almost negligible as lattice vibrations significantly decrease. The difference between room temperature resistivity and cryogenic temperature resistivity for both ρ(D) and ρ(I) is extremely small and nearly constant.2) Therefore, eq. (1) and (2) can be combined to obtain eq. (3). This equation indicates that the RRR is a function of ρ(D) and ρ(I).
| \begin{align} \mathrm{RRR} &= \rho_{\text{293$\,$K}}/\rho_{\text{4$\,$K}}\\ & = (\rho(L)+\rho(D)+\rho(I))/(\rho(D)+\rho(I)) \end{align} | (3) |
Previous studies have reported that the RRR is considerably affected by the concentration of impurities in OFC.3–5) The RRR is also affected by heat treatment processes as it affects the form of the impurities present in the material. In particular, when OFC undergoes high-temperature heat treatment, the RRR decreases significantly owing to the increase in concentration of dissolved impurities.3) Therefore, to obtain a high RRR, a special refining process is necessary to reduce the concentration of impurities in the material.6) A previous study has reported a method to improve the RRR via atmospheric heat treatment over a long period to internally diffuse oxygen and oxidize the impurities.7) However, this technique presents limitations in large-scale production.
Among the impurity elements, S demonstrates the largest effect on the resistivity of heat treated OFC.3,4) Therefore, in this study, we developed a design method to increase the RRR of pure copper by adding an extremely small amount of a selected element to OFC to precipitate the S in OFC as a compound. Consequently, we obtained pure copper that did not require a special refining process and exhibited a high RRR over a wide range of heat treatment temperatures.
Considering the standard Gibbs free energy of formation of sulfides8) and the increase in resistivity with dissolution,9–11) Ca, La, Mg, Sr, Ti, Y, and Zr were selected as additive elements to precipitate S from the matrix as a compound, as shown in Table 1.
Ingots were prepared in a vacuum melting furnace by adding the selected elements to commercially available OFC (S: 3 mass ppm) in the concentration range of up to 100 atomic ppm. An ingot without the added elements was also prepared as a control sample. The ingots were then heated under homogenization heat treatment, groove rolled, and cold drawn at an area reduction ratio of 99% or more to obtain 2 mm diameter wires, and this was followed by a final heat treatment at 373 K to 1073 K under hydrogen atmosphere for 3600 s to produce samples for evaluation. As illustrated in Fig. 1, the resistance of the sample was measured at room temperature (293 K) and liquid helium temperature (4 K) using the four-terminal measurement method. The RRR was then calculated using eq. (1).
Schematic diagram of electrical resistivity measurement method.
Figure 2 presents the change in the RRR with the addition of Ca, La, Mg, Sr, Ti, Y, and Zr after the final heat treatment at 773 K for 3600 s. The RRR of OFC was approximately 450, and a decrease in the RRR was observed for all samples with additive elements added at concentrations below 10 atomic ppm. This finding may be attributed to the low activity of the additive elements in the sample with Cu at a concentration of 10 atomic ppm, which prevents sufficient reaction with S. Moreover, the unreacted additive elements may dissolve into the matrix and cause increased resistivity. The RRR of the samples added with Ca, La, and Sr increased with additive contents ranging from 10 atomic ppm to 20 atomic ppm, and for the La and Sr added samples, the RRR decreased beyond the peak concentration, whereas Ca maintained a high RRR over a wide concentration range of 10 atomic ppm to 40 atomic ppm. By contrast, the RRR did not increase with the addition of Mg, Ti, Y, and Zr; instead, it decreased with increasing addition of these elements.
Effect of additive element on RRR for heat treatment at 773 K, 3600 s.
Figure 3 presents the results of scanning electron microscopy and energy-dispersive X-ray analyses of the Ca-added sample that exhibited a high RRR over a wide concentration range. Compounds containing Ca and S were confirmed at 10 atomic ppm of Ca addition. At Ca concentrations greater than 41 atomic ppm, compounds containing Ca were identified. Transmission electron microscopy and electron diffraction results indicated that these compounds were CaS and Cu5Ca.
Compounds of (a) Cu-10 atomic ppm Ca, and (b) Cu-41 atomic ppm Ca.
These results suggest that CaS formed at a concentration of 10 atomic ppm of Ca, and any addition of excessive Ca led to the formation of Cu5Ca due to its low solubility, thus RRR did not reduce over a wide concentration range of 10 atomic ppm to 40 atomic ppm.
3.2 Effect of Ca addition and heat treatment temperature on the RRRFigure 4 depicts the change in the RRR of the drawn and heat treated samples for OFC and Cu-17 atomic ppm Ca, which exhibited the highest RRR in Fig. 2. The final heat treatment temperatures of these materials were varied from 373 K to 1073 K.
RRR of Cu-17 atomic ppm Ca and OFC.
Similar RRR for the drawn Cu-17 atomic ppm Ca and OFC samples prior to the final heat treatment was observed, and the RRR was approximately 50. Heat treatment at 423 K increased the RRR of the Cu-17 atomic ppm Ca sample to approximately 400, which is more than eight times greater than that of the OFC sample. The RRR of both samples increased as the heat treatment temperature increased. For the OFC sample, the peak RRR was obtained at 773 K. Above this temperature, the RRR decreased significantly to approximately 200 at 973 K. In comparison, the Cu-17 atomic ppm Ca sample presented an RRR of approximately 700 at 773 K, and this value continued to increase as the heat treatment temperature increased, reaching approximately 900 at 1073 K, which is more than four times greater than that of the OFC sample.
Further, microstructural analysis of the samples demonstrated the presence of Cu2S in both the drawn and the heat treated OFC samples from 373 K to 773 K but not in the OFC samples heat treated above 773 K. In comparison, CaS was confirmed in the Cu-17 atomic ppm Ca sample heat treated up to 1073 K. This finding indicates that CaS is a more thermodynamically stable compound than Cu2S and can exist at high temperatures up to 1073 K.
To understand the mechanism underlying the high RRR of the Cu-17 atomic ppm Ca sample, a resistivity numerical prediction model was constructed based on eq. (2). The results of this model were then compared with the experimental data. Here, ρ(D) denotes the resistivity based on the dislocation density and grain boundaries, and ρ(I) is the resistivity based on the dissolution of S, Ca, and other impurities.
Table 2 summarizes the calculation method. The equation for the dislocation density in Table 2 is obtained by the combination of the equation that applies to the experimental data (table) of annealing temperature and dislocation density described in Ref. 12), and the relationship between the dislocation density and resistivity described in Ref. 13). The effect of grain boundaries at each heat treatment temperature was obtained from the resistivity based on grain boundaries,14) which was determined by measuring the grain size from the cross-sectional microstructure of each sample. Based on analytical values and thermodynamic calculations,8) the effects of S and Ca on ρ(I) were determined by obtaining the dissolution concentration at each investigated heat treatment temperature and multiplying it by the increase in resistivity per 1 atomic ppm.9,10) The effect of impurities was obtained from the analytical values for each element and increase in resistivity per 1 atomic ppm.9–11)
The resistivities of the Cu-17 atomic ppm Ca sample and the OFC sample at 4.2 K were calculated as above for heat treatment temperatures ranging from 473 K to 1073 K. Figure 5 presents a comparison of the calculated values with experimental values. The experimental ρ4K values were obtained by dividing ρ(L) by RRR, assuming ρ293K of all experimental samples is mainly ρ(L) and the effect of ρ(D) and ρ(I) is extremely small in comparison. This was conducted since the experiment measures the resistance of the samples and not the resistivity, and determining resistivity would require precise measurement of the surface area and length of the samples and this could result in errors due to the high sensitivity of the values.
Calculated resistivities of (a) Cu-17 atomic ppm Ca, and (b) OFC, with respect to heat treatment temperature.
The calculated values agreed well with the experimental values. The Cu-17 atomic ppm Ca sample and the OFC sample presented large differences in ρ(I) at heat treatment temperatures of 773 K and above owing to the dissolution of S. In other words, the RRR of the Cu-17 atomic ppm Ca sample did not decrease at temperatures of 773 K and above, most likely because CaS is a highly thermodynamically stable compound, and the added Ca and S do not dissolve into the matrix even at high heat treatment temperatures.
The decrease in ρ(D) with increasing heat treatment temperature may also be attributed to the decrease in dislocation density and grain boundary area owing to recovery, recrystallization, and grain growth. Figure 6 illustrates the change in hardness and grain size of the samples at each investigated heat treatment temperature.
(a) Vickers hardness and (b) grain size.
Figure 6 indicates a decrease in the hardness and coarsening of crystal grains in both the OFC and the Cu-17 atomic ppm Ca samples with an increase in the heat treatment temperature. As indicated in Fig. 5, ρ(D) decreases with increasing temperature. For the OFC sample, the increase in ρ(I) is relatively greater than the decrease in ρ(D) owing to the dissolution of S in the matrix. For the Cu-17 atomic ppm Ca sample, ρ(I) remains low and does not increase even at above 773 K, resulting in a decrease in the overall resistivity owing to a decrease in ρ(D). In other words, the increase in the RRR of the Cu-17 atomic ppm Ca sample above 773 K is caused by the decrease in ρ(D). The hardness value of both the Cu-17 atomic ppm Ca sample and the OFC drawn wire sample prior to heat treatment was 127 HV; notably, the hardness of the Cu-17 atomic ppm Ca sample was much lower than that of the OFC sample after heat treatment at 473 K. Comparing the coarsening of the crystal grains for the Cu-17 atomic ppm Ca sample and the OFC sample indicates that the Cu-17 atomic ppm Ca sample recovers and recrystallizes at lower heat treatment temperatures owing to the precipitation of S from the matrix with the addition of Ca.
This study focused on a design method to increase the RRR of pure copper with the addition of an extremely small amount of a selected element to OFC to precipitate the S in OFC as a compound. Pure copper exhibiting a high RRR was successfully obtained with the addition of an extremely small amount of Ca without any refining processes. The developed material exhibited an RRR greater than that of OFC over heat treatment temperatures ranging from 423 K to 1073 K. Specifically, at 423 K, the RRR of the developed material was approximately 400, which is more than eight times greater than that of OFC. At 773 K, the RRR was approximately 700, which is more than 1.5 times greater than that of OFC. At 1073 K, the RRR was approximately 900, which is more than four times greater than that of OFC.
The results of an RRR numerical prediction model indicated that Cu-17 atomic ppm Ca exhibited a high RRR over a wide temperature range as CaS is a highly stable compound, and S impurities do not undergo dissolution during high-temperature heat treatment. In the case of OFC, the RRR decreases owing to the dissolution of S impurities in the heat treatment temperature range of 773 K and higher.
