A Comparative Differential Scanning Calorimetry Study of Precipitation Hardenable Copper-Based Alloys with Optimized Strength and High Conductivity

Abstract

Copper alloys with chromium, hafnium, and scandium combining enhanced strength as well as high electrical and thermal conductivity are analyzed in depth. The aim is to compare the precipitation process during temperature exposure to meet increasing material requirements. This research focuses on alloying elements having a limited, maximum 1 wt.%, and with temperature decreasing solubility in copper. For the simultaneous enhancement of mechanical strength and conductivity, precipitation hardening is the utilized mechanism during the processing of as-casted annealed and quenched specimens and in combination with optional cold-rolling prior to the aging process. Extensive DSC measurements, accompanied by metallographic investigations, and the analysis of hardness and electrical conductivity, lead to a versatile description and comparison of different alloying systems. CuCr0.7 starts to precipitate early and is mainly influenced by the temperature of aging. Provoking the solid solution with cold deformation has a less significant influence on the following precipitation. CuSc0.3 and CuHf0.7 precipitate at higher temperatures and are highly influenced by cold deformation prior to aging. Furthermore, CuHf0.7 and CuSc0.3 show advantages regarding the recrystallization behavior, making them especially applicable for higher operating temperatures. Future research will assess ternary alloy combinations to further scoop the potential.

1. Introduction

For copper alloys in high-performance applications, sufficiently high electrical, respectively, thermal conductivity, and a high strength are essential. Precipitation hardening enables a favorable combination of good conductivity, hardness and strength, ductility, and less softening at higher temperatures. This can be used in various current and future applications of copper in the automotive and electronic industries, including electromobility components, welding electrodes, switches, casting molds, or die-casting plungers.

Potentially applicable alloying elements for precipitation hardening show limited solubility in copper, that decreases further at lower temperatures [1]. Regarding conductivity, the solute amount of alloying element within the copper matrix is the most crucial factor due to the disadvantageous scattering of conduction electrons [2,3,4]. Therefore, low-alloyed precipitation hardening, binary, copper alloys containing less than 1 wt.% alloying elements are the focus. Alternative alloying elements, including scandium and hafnium, are compared with the industrially commonly used chromium to further exploit the precipitation potential of low-alloyed copper alloys for high-performance applications.

Both hafnium and scandium appeared to provide beneficial mechanical properties during aging treatments, paired with benefits regarding recrystallization behavior and grain refinement, but are rarely analyzed in literature so far, as described in the following.

1.1. Copper Alloys Containing Chromium

Chromium has a limited solubility of 0.65 wt.% at 1075 °C in copper, which decreases at lower temperatures [5,6]. The corresponding binary phase diagram visualizes these basic requirements for precipitation in Appendix A Figure A1.

With a similar atomic radius of chromium and copper (r_Cu = 0.128 nm and r_Cr = 0.127 nm), the rapidly and continuously growing precipitates, which do not have preferred growth sites [2,7], assume a spherical shape for the reason of minimized surface energy [8]. The necessary nucleation barrier for a growth-capable fcc nucleus, which orients to the crystal structure of the copper matrix, is significantly lower than that of a corresponding bcc nucleus [9].

Nevertheless, the chromium-rich phases later strive for their stable bcc equilibrium structure, which is at the expense of the formed fcc nuclei [7,8,9,10,11]. According to Bodyakova et al., chromium supersaturation of the copper matrix has a significant effect on lowering the activation energy of segregation, increasing the nucleation rate, and improving stability against coarsening [12]. In the case of fast-growing partially coherent, as well as incoherent precipitates (8–10 nm), the strengthening Orowan-mechanism is active [8,10,13,14,15]. The cutting mechanism dominates at huge amounts of smaller coherent precipitates (3–4.5 nm) [10,12,16].

In summary, the precipitation mechanism of chromium in copper has been the subject of scientific investigation for decades due to its industrial importance and continues to be debated by recent findings. This research uses CuCr as the benchmark for directly comparing the two other alternative alloying systems.

1.2. Copper Alloys Containing Scandium

Scandium has a maximum solubility of 0.35 wt.% at 865 °C [17,18,19] in copper, the lowest solubility of the compared systems in this study. The corresponding binary phase diagram is added in Appendix A Figure A2. A quenched solution can be used in aging temperature treatments to optimize mechanical properties and electrical conductivity simultaneously [20,21,22,23].

During the aging treatment, Hao et al. describe the precipitation process with homogeneous nucleation following solid solution and scandium-enriched clusters [24]. During the growth, more atom layers attach, and the distortion of the surrounding matrix increases due to scandium’s larger atom size and the resulting mismatch [24]. Lamellar precipitates follow semi-coherent and incoherent stages to the tetragonal orientated lamellar Cu₄Sc-precipitates [24]. Therefore the strengthening uses the dislocation interaction with precipitates in the shearing and Orowan by-passing mechanism [24], which is analyzed by Hao et al., especially for strongly deformed and cryorolled specimens [24], and by Franczak et al. without any deformation prior to the aging [21].

Combined with grain refining [25,26] and potentially recrystallization-inhibiting effects, scandium’s thermal stabilizing influence is interesting for applications at higher operating temperatures [27].

Overall, current research topics of CuSc alloys focus on higher aging temperatures and on as-cast quenched and excessively deformed specimens for precipitation treatment. With respect to industrial production, moderate degrees of cold deformation are of particular interest, leading this study to concentrate on the correlating processing.

1.3. Copper Alloys Containing Hafnium

Thermodynamic equilibrium considerations indicate a maximum solubility of 0.84 wt.% hafnium in copper at 963 °C [28], which is the highest solubility compared to the other alloying systems and visualized in Figure A3 of the Appendix A. The mechanical properties of CuHf alloys are advantageous with improved mechanical properties [2,29], compared to the benchmark CuCr alloy with minor sacrifices in electrical conductivity [29,30,31,32]. Due to their similar thermomechanical behavior, CuHf alloys are compared directly with CuZr and CuTi alloys [33,34,35,36].

The precipitate phase is considered to be the stable intermetallic compound Cu₅Hf [37,38,39,40], which has AuBe₅, known from the low-alloyed copper alloy CuZr [41,42], as a cubic prototype [36,41]. Based on this similarity, it is concluded that the precipitation process of CuZr and CuHf alloys is comparable [38,42,43,44,45]. The precipitation process follows rod-like structures that grow, for example, at 450 °C within 10 min, 60 min to 300 min from 5 nm to 10 nm and 20 nm [29].

Small additions of the grain refining [29,46] alloying element raise the recrystallization temperature [29,47] and a weakening occurs at higher temperatures [40] while the hafnium concentration increases [29,32]. Accordingly, studies of ultrafine-grained (UFG) microstructures are conducted using hafnium as an alloying element, mainly in ternary combinations with chromium [42,43,44,46].

In summary, research topics on CuHf alloys focus on aging temperatures above 400 °C and excessive deformation processes. With respect to industrial processing methods, moderate degrees of cold deformation are more realistic in their application, which leads this study to focus on those. In this scope, the behavior of binary CuHf was not compared with other low alloyed copper alloys directly.

1.4. Analyzing Thermomechanical Processing of Metals Using DSC Measurements

Differential scanning calorimetry (DSC) studies are a powerful thermoanalytical tool for analyzing the exothermic and endothermic effects of a heated, cooled, or isothermally treated sample. Therefore, DSC is used to investigate the heat released from clustering processes, nucleation, and precipitate growth [48,49,50], recovery, or recrystallization [51,52]. On the contrary, endothermic effects can, for instance, be related to a dissolution of precipitates that consumes energy from the surrounding [49,53]. For low alloyed CuCrZr [50,52], CuBe [54], CuNiSi [55,56], and CuCrHf [43] some DSC measurements have been published. Regarding the benchmark alloy of this research, the available literature focuses mainly on methods to generate UFG materials [43,50,51,52], including their recrystallization behavior.

For academic and industrial research, there is a strong interest in the influence and interaction of cold deformation with a finalizing aging treatment [57]. The cold-worked state of a quenched solid solution, accompanied by dislocation hardening, is a favorable condition [58], which lowers the onset and peak temperatures of the following reactions [59,60,61,62,63,64] due to the facilitated diffusion of alloying elements along dislocations [59,64]. More crystallographic defects and heterogeneous nucleation sites lower the necessary activation energy [63,64,65]. During the heating process in DSC measurement, different effects occur and can overlap, which requires sufficient knowledge of the material based on other investigations [57]. For cold worked specimen, exothermic peaks, in addition to the expected precipitation, at lowest temperatures can be related to recovery [61], whereas effects at highest temperatures can usually be related to the recrystallization [58].

The purpose of this work is to gain a deeper understanding of the kinetics of mechanisms during a thermomechanical treatment of low-alloyed copper alloys. Therefore, this publication discusses the DSC measurements with the background of electrical and mechanical properties accompanied by metallographic investigations. The identical processing enables a direct comparison of alloying concepts. Subsequently, CuSc0.3, CuHf0.7, and CuCr0.7, in an as-casted and a cold worked condition, are the focus. In addition to precipitation analysis, investigations are conducted on thermal stability and recrystallization. Especially for applications at higher operating temperatures, the alternative alloying concepts highlight their benefits. This work differs from available literature as it compares binary alloying concepts and emphasizes the behavior at lower aging temperatures and moderate cold deformation. The analysis of CuSc0.3 with identical processing was recently published [26,66] and is the starting point of this investigation.

2. Materials and Methods

The three identically processed binary copper alloys CuCr0.7, CuHf0.7, and CuSc0.3 were analyzed and compared near their maximum solubility in copper. The gravity die casting process of the specimen with a 500 g casting weight was performed using a VC400 casting machine (Indutherm Blue Power Casting Systems, Walzbachtal, Germany). The raw material Cu-OFE and the master alloys (CuHf60, CuSc23, and CuCr10) were melted in a boron nitride-covered graphite crucible and casted at 1300 °C in a graphite die (5 mm thick bar) under vacuum conditions. The alloy composition and maximum derivation of 0.05 wt.% from the target composition was proven with a calibrated optical emission spark spectrometer (Spectrotest, SPECTRO Analytical Instruments GmbH, Kleve, Germany).

To enable precipitation hardening, a 120 min solution treatment in a preheated furnace (ME65/13, Helmut ROHDE GmbH, Prutting, Germany) was followed by water quenching. The solution annealing temperatures correspond to the maximum solubility of the individual element. Therefore, CuCr0.7 was annealed at 1000 °C, CuHf0.7 at 960 °C, and CuSc0.3 at 870 °C.

An optional cold deformation of 25%, 50%, and 75%, which correlates with a feed rate of 0.25 mm in equidistant longitudinal rolling steps, was performed on a duo roll stand (roll diameter 110 mm, rotation speed 27 min⁻¹) (Bühler, Pforzheim, Germany). The casted bars containing the most alloying elements were used for the 0%, and those containing the least alloying element were used for the 75% deformation route to not amplify differences between the processing routes related initially to differences in the alloys’ composition. Therefore, the sample preparation corresponds to the manufacturing of Dölling et al. [26] and is visualized in Figure 1.

During isothermal aging experiments (for hardness and electrical conductivity measurements), the specimens were aged up to 48 h at 350 °C, 375 °C, 400 °C, 425 °C, 450 °C, and 500 °C (Figure 2). Measurements of electrical conductivity were conducted with an eddy current test (Sigmascope SMP10, with TF100A for temperature compensation, Helmut Fischer GmbH, Sindelfingen, Germany) and hardness tests with a microhardness tester (HV_0.1) ((NEXUS 412A equipped for DIN EN ISO 6507-1:2018 Vickers hardness test, Innovatest GmbH, Selfkant-Heilder, Germany). Due to the huge amount of experimental data, only the peak values are visualized. Each data point (red) on the corresponding surface plots is based on the peak of an isothermal aging curve. The average standard derivation for electrical conductivity measurements was 0.83 MS/m (n = 7) and 5.18 HV_0.1 (n = 5) for hardness measurements.

The precipitation behavior was substantiated by utilizing differential scanning calorimetry (DSC) (STA 449 F3 Jupiter equipped with a platinum oven for usage up to 1500 °C, NETZSCH-Gerätebau GmbH, Selb, Germany). Within the chamber, two Al₂O₃ crucibles with lids were positioned with high repeatability. One of them was permanently the empty reference. The temperature and enthalpy were calibrated by melting pure In, Sn, Bi, Zn, and Al multiple times. During the whole experiment, the temperature was regulated using the sample temperature controller (STC). A reproducible baseline with empty crucibles was checked initially and regularly during the experiments to ensure repeatability. The heat flow signal of the baseline was subtracted from the alloy measurements to correct the signal. To perform calibration and tests in an atmosphere of argon, the protective (20 mL/min) and purge inlets (50 mL/min) were used during heating with 10 K/min from room temperature (RT) to 600 °C and 800 °C. After holding the final temperature for 1 min, the cooling process followed, as visualized in Figure 2.

The solution annealed and quenched specimens, with optional cold deformation, were machined to cylinders with 4 mm diameter, followed by a surface grinding (up to P2500). All DSC specimens had a comparable mass of 117 mg, which was adjusted by their height. A reference of pure copper (Cu-OFE) with identical processing was measured for comparison. The onset temperatures were conducted according to DIN EN ISO 11357-1:2016, visualized with a (*) in the corresponding Figures 4, 7 and 11.

Heating rates oriented on experiments for aluminum, magnesium, and copper alloys, which mainly used 10 K/min [48,53,61,67,68], to ensure the visibility of relevant thermodynamic effects and comparability. Using an enhanced heating rate results in a shift of peaks to higher temperatures [59,64,69,70], which underlines the time and temperature dependence of these thermally activated and kinetically controlled solid-state reactions [60,64,71].

All DSC measurements were accompanied by metallographic analysis in an optical and scanning electron (SEM) microscope. To obtain better contrast in the optical microscope, the specimens were etched, corresponding to Klemm III [72].

3. Results

A brief comparison of the three alloys CuCr0.7, CuSc0.3, and CuHf0.7 was the primary goal. First, a basic understanding of the alloys’ mechanical properties and electrical conductivity, representing the outcome of precipitation reactions, was generated during isothermal aging experiments. Afterward, independent DSC experiments were conducted to directly measure thermal effects of energetic transformations with accompanying metallographic investigations.

The analysis showed similarities for the recrystallization inhibiting alloys CuSc0.3 and CuHf0.7, depending greatly on the thermomechanical treatment, and contrasted the differences of the mainly temperature sensitive CuCr0.7. Generally, the electrical conductivity measurement was very sensitive to changes in the solute amount of alloying element in the copper matrix and, therefore, the precipitating volume fraction.

3.1. Results for Copper-Chromium Alloy CuCr0.7

Regarding the resulting properties, CuCr0.7 showed high electric conductivity (Figure 3a) and remarkable hardness benefits (Figure 3b). The maximum hardness increased up to an aging temperature of 450 °C for a 0% and 25% cross section reduction prior to the aging process and up to 425 °C for the further deformed options. Lower aging temperatures did not appear to have a significant impact on the cold rolled specimens. Only the 0% specimen needed an isothermal aging temperature of at least 400 °C for beneficial mechanical properties. Overall, the best mechanical properties were measured for 50% and 75% cold-rolled specimens after 4 h aging at 425 °C with 180.8 HV_0.1 and 176.2 HV_0.1. Increasing the aging temperature to 450 °C accelerated the reaction and reduced the required time to 2 h (50%) and 1 h (75%), resulting in a peak hardness of approximately 175 HV_0.1. On the other hand, specimens without cold rolling reached 159.2 HV_0.1 after aging for 4 h at 450 °C as well, which is only 21.6 h.

CuCr0.7 with a cross-section reduction of 75% enabled the best electrical conductivity above 55 MS/m using at least 425 °C for isothermal aging. The maximum electrical conductivity was reliably reached after the maximum aging time of 48 h. Nevertheless, specimens without cold deformation prior to the aging treatment above 425 °C ended up at about 50 MS/m to 55 MS/m as well.

Overall, the alloy CuCr0.7 reacted homogenously and showed similar results for all processing options but appeared to be sensitive to the aging temperature. Generally, the aging treatment at 425 °C provided beneficial mechanical and physical optimum properties.

In order to characterize the precise temperature positions of the reactions, the DSC analysis in Figure 4 illustrates onset and peak temperatures up to conditions 1 and 2 (referring to Figure 2), emphasizing the experiments’ high reproducibility. Generally, the DSC signals of CuCr0.7 appeared with a large exothermic peak at lower temperatures. Cold rolling, prior to the in-situ aging during the DSC measurement, did not promote solid-state reactions with huge impacts. The average peak temperature of specimens with 75% cross section reduction was 454.7 °C (455.9 °C and 453.5 °C) and only 16 K lower compared to the average 470.8 °C (468.1 °C and 473.5 °C) for specimens without cold rolling prior to the measurements. This investigation correlates well with developments in hardness and electrical conductivity (Figure 3), confirming that the alloy reacted more sensitively to temperature than to the introduction of crystallographic defects.

Following the prominent peak, more minor exothermic effects with reproducibly measurable peak temperatures were detectable for all specimens. For specimens without cold deformation, prior to the in-situ aging during the DSC measurements, a remarkable peak at an average 547.8 °C was observed. The specimens with 75% cross section reduction due to cold rolling showed this second peak at average 516.4 °C. The average distance between the first and second peak was 77 K for the 0% option, and 61.7 K for the 75% cold rolled one. Regarding the already described peak shift of the first exothermic effect (16 K to lower temperatures), these differences are consistent and can be transferred belonging to the same microstructural changes in both processing methods. Following the same path of analysis, a third peak at about 589 °C for the 0% option and 549 °C for the 75% option finalized this picture.

Developing insight into the effects taking place in the DSC aging treatment is mainly supported by the consequent metallographic analysis in Figure 5, as well as Appendix B, and Figure A5 and Figure A6, characterizing conditions 1 and 2 of DSC measurements. For specimens without cold rolling prior to the aging process, the precipitation of chromium was exclusively responsible for the induction of exothermic effects during the measurements.

Cold rolling before the aging process introduced a texture in the material that induced recrystallization (arrows 2, Figure 5) because of heating up to 600 °C. Further heating to 800 °C resulted in comparably larger grains (arrows 3, Figure 5), also visible in Appendix B Figure A6. Subsequently, on primary recrystallization, further grain growth in secondary recrystallization occurred between conditions 1 and 2. Considering the correlating DSC curve, this exothermic effect can be identified to start at approximately 700 °C, visualized in Figure 4 within the second area of recrystallization. Secondary recrystallization is not finished, as the grains appear to grow further and the DSC effect did not reach its peak temperature up to 800 °C.

Regarding the primary recrystallization (Figure 4), a difference between the DSC signals of 0% and 75% cold rolling options needs to be the measurable basis. Subsequently, the peak at the highest temperature, which cannot directly be related to a matching equivalent in the 0% DSC curve for condition 1, needs to refer to the primary recrystallization phenomenon. Therefore, the recrystallization appears to be located at an average peak temperature of 576.9 °C, visualized by the corresponding areas of primary recrystallization in Figure 4.

3.2. Results for Copper-Scandium Alloy CuSc0.3

The aging experiments of the binary CuSc0.3 alloy showed remarkable changes in mechanical properties (Figure 6b), as well as considerable increases in the maximum electrical conductivity (Figure 6a). CuSc0.3 showed its best mechanical properties at high degrees of cold rolling and lower aging temperatures, utilizing work hardening and precipitation formation. The highest hardness with 197.6 HV_0.1 was reached in a specimen with 75% cold rolling aged at 350 °C for 4 h, whereas values, such as 190 HV_0.1, were possible after 1 to 2 h at an aging temperature of 375 to 425 °C (Figure 6a). Higher temperatures above 450 °C, combined with the provoked cold rolled state of quenched solid solution, decreased the measurements on maximum hardness.

Specimen aged directly after solution annealing and quenching showed clear outcomes of precipitation through a sensible reaction of the material’s electrical conductivity. Starting at 400 °C isothermal aging, the specimens reached more than 40 MS/m (Figure 6a), but the maximum electrical conductivity above 50 MS/m was obtained at 75% cross section reduction by cold rolling and further increased aging temperatures.

Combining the results of these measurements allows for the conclusion that cold deformation appears to be essential and results in significantly enhanced properties of this material. Overall, the most remarkable influence was present between the non-deformed specimens and all other options (Figure 6a,b), but the trends continued as the degree of cold rolling increased.

To further describe the alloy’s reaction to a continuous heating process, the DSC offered information on the thermal effects analyzed in conditions 1 and 2. The first exothermic effects of CuSc0.3 are shown in Figure 7 for specimens with 0%, 25%, 50%, or 75% cold rolling prior to the DSC measurements, at the beginning onset of 513.8 °C, 430.1 °C, 430.1 °C, or 429.1 °C and a peak temperature at 568.9 °C, 466.3 °C, 461.1 °C, or 470.4 °C. Therefore, the peak was shifted by 102.6 K and the onset temperature by 83.7 K to lower temperatures due to the introduction of 25% cold deformation. This shifting continued with a significantly smaller increment. The increase in electrical conductivity, as a sensitive marker for the transfer of alloying elements from the copper matrix to precipitation phases, showed the same behavior. Furthermore, the measurements were reproducible for all experiments conducted up to conditions 1 and 2.

The specimens were analyzed in conditions 1 and 2 of the DSC measurements in Figure 8 to directly relate visible effects to notable changes in the microstructure. Figure 8a visualizes the formation of small lamellar structures (magnification 10,000) homogeneously distributed in the grains (arrows 1).

The darker structures that appeared periodically in the backscatter detector were directly related to an enhanced scandium content in the EDS (Appendix B Figure A4) and visible in optical microscopy as well (Appendix B Figure A7).

With further heating to 800 °C, the visible number of precipitates in Figure 8b decreased and the intermetallic phases coarsened (magnification 3000) to larger needle-shaped and plate-like precipitates (arrows 4). Furthermore, alloyed scandium migrates in the grain boundaries and builds intermetallic phases in the intergranular space as well (arrows 3).

The visible peak of the DSC at 568.9 °C with its onset temperature at 513.8 °C can be directly related to the precipitation formation and growth process of this alloy.

For the cold rolled specimens, the DSC curve showed how the precipitation peak was shifted to lower temperatures. Furthermore, additional exothermic effects related to microstructural changes occurred.

Figure 9a (and Figure A8b in Appendix B) shows how fine precipitates of the intermetallic CuSc-phase formed on slip bands (arrows 1) and grew next to crystallographic defects, such as dislocations. Further heating to 800 °C affected the microstructure of all cold rolled specimens with recrystallization. The options with 25% and 50% cross-section reduction also showed primary recrystallization (Appendix B Figure A11b).

Primary recrystallization, visible in the partly new-built grains in Figure 9a (arrows 2), of the 75% cold rolled specimen, at 648.5 °C in the corresponding DSC curve, was followed by secondary recrystallization, visible due to the comparably large newly built grains in Figure 9b. This microstructural change, which did not appear for specimens with a lower degree of cold rolling, can be related to the second visible peak in the DSC signal at 719.4 °C.

For a direct comparison with Cu-OFE, Appendix C documents the results of DSC (Figure A12) and metallographic analysis (Figure A13 and Figure A14). In the case of the same cross-section reduction after cold rolling, Cu-OFE showed its primary recrystallization peak at 350 °C (Appendix C Figure A12), which is 299 K lower than the one at CuSc0.3.

3.3. Results for Copper-Hafnium Alloy CuHf0.7

Regarding the aging experiments, the binary CuHf0.7 alloy showed beneficial mechanical properties (Figure 10b), as well as good maximum electrical conductivity (Figure 10a).

For an excellent maximum hardness, this alloy needed an increased degree of cold rolling before aging. The highest hardness of 204 HV_0.1 was reached by a 75% cold rolled specimen aged at 350 °C for 12 h, whereas 195 HV_0.1 to 199 HV_0.1 were realistic after 2 h at 375 °C to 400 °C isothermal treatment (Figure 10b). As already investigated for CuSc0.3, higher aging temperatures combined with the provoked cold rolled state of the quenched solid solution resulted in decreasing maximum hardness, due to the conglomerate of hardening mechanisms and the resulting interaction of dislocations and precipitates. For specimens without cold rolling, a higher aging temperature helped introduce precipitation and increase mechanical (Figure 10b) and physical properties (Figure 10a). Still, it was not possible to reach comparable properties, with respect to the cold rolled specimens, at all.

Higher aging temperatures were constantly helpful for the electrical conductivity, regardless of previous material processing (Figure 10a). With 75% cold rolling and the following aging at 500 °C, the electrical conductivity reached a maximum of 52 MS/m.

Again, this points out the importance of careful process parameter adjustment for individual requirements. The behavior trends in CuHf0.7 appear comparable to the already described CuSc0.3, differing from CuCr0.7.

The DSC measurements of CuHf0.7 in Figure 11 confirmed and precisely measured how significant cold rolling influenced the precipitation process and how the peak and onset temperature of the exothermic precipitation process were affected. Cold rolling inducing a 25%, 50% or 75% cross section reduction shifted the start of the exothermic effect to lower temperatures of 473.5 °C, 463.0 °C, or 460.1 °C with a peak at 513.4 °C, 497.8 °C, or 491.1 °C. Therefore, the maximum derivation between the three options was 13 K for the onset temperature and 22 K for the peak temperature. In contrast, the 0% specimen showed a precipitation effect at 509.6 °C, 36 K above the onset temperature of the 25% cold rolled option. The onset is followed by a 48 K higher peak temperature, compared to the 25% option.

In summary, cold rolling prior to the DSC experiments and aging reproducibly shifted the effects in CuHf0.7 to significantly lower temperatures. There was a remarkably larger difference between specimens aged directly after solution annealing and quenching and the other cold rolled solid solution material options.

The measurements were reproducible in direct comparison to all experiments conducted up to conditions 1 and 2 during the DSC measurements.

The accompanying metallographic investigations in SEM (Figure 12) and optical microscope (Appendix B Figure A9) contribute essential information to assign the thermal effects to the occurring changes in the alloy’s microstructure.

For specimens without mechanical treatment, the precipitation process can be directly related to the exothermic effect that occurs. Figure 12a shows the resulting microstructure in condition 1 and Figure 12b in condition 2, referring to the DSC treatment. In the case of condition 1, the lamellar precipitates were small (magnification 10,000) and homogenously distributed in the grains (arrows 2), which matches the location of the corresponding peak temperature at 561.7 °C. The bright and periodically appearing structures in the backscatter detector were directly related to an enhanced hafnium content in EDS (Appendix B Figure A4). Close to the grain boundaries, the zones with enhanced hafnium concentration were smaller (arrows 3) and migrated partially in the intergranular space (arrows 4).

With further heating to 800 °C, a coarsening of precipitates (arrows 5) is visible in Figure 12b (magnification 3000). Due to the sample preparation, only the precipitation sequence can possibly induce exothermic effects in the analyzed temperature range and be responsible for the peak at 561 °C, as well as 603 °C. For the cold rolled specimens, the covered temperature ranges of exothermic precipitation processes are tighter, which characterizes an accelerated and, therefore, steady stimulation of the precipitation process.

Figure 13 shows microstructural changes, including precipitation and recrystallization for a specimen with 75% cross-section reduction due to cold rolling. The investigation of condition 1 shows how precipitates grew preferentially at crystallographic defects, such as dislocations, which is why they are clearly visible along the slip bands in Figure 13a (magnification 7000). Therefore, the effect at 491.1 °C can be related to the precipitation process. Heating up to 800 °C showed its results in Figure 13b (magnification 3000) with a noticeable growth and accumulation of precipitates (arrows 2). Furthermore, many newly formed small grains became visible during primary recrystallization (arrow 3), allocated to the visualized area of recrystallization having its peak exothermic effect at 718.4 °C. Specimens with a lower degree of cold deformation matched this behavior at higher temperatures (recrystallization areas in Figure 11 and the corresponding metallography in Appendix B Figure A10).

4. Discussion

A primary objective of this investigation was to use different experimental methods to visualize and validate the behavior trends of CuSc0.3, CuHf0.7, and the industrially used CuCr0.7 during heat treatment and to establish the active mechanisms step by step. Supersaturated solid solutions were the driving force for nucleation and precipitation growth [73,74], which enhanced the mechanical properties and electrical conductivity at the same time.

4.1. Influence of Cold Deformation Prior to an Aging Treatment of Low-Alloyed Binary Copper Alloys

This experimental study focuses on the influence of moderate degrees of cold rolling, prior to aging treatments. Figure 14, as a synopsis of the observed microstructural changes, shows schematically how the evolution of the materials occurred in CuSc0.3 and CuHf0.7.

Generally, all alloys showed precipitation-promoting reactions in the case of a cold rolled state of supersaturated solid solution. Especially CuSc0.3 and CuHf0.7 appeared to be highly impacted by paths of higher diffusion and the amount of potent nucleation sites [1,75]. The use of cold deformation and, therefore, the introduction of dislocations resulted in plenty of small precipitates. Nearby, an enhanced density of crystallographic defects, such as dislocations, or slip bands, were beneficial positions for precipitation nucleation and growth (Figure 9a and Figure 13a). In specimen, directly solution annealed, quenched, and aged from the as-cast state, the precipitates appeared to be significantly larger. After excessive heating up to 800 °C in the DSC measurement, the resulting microstructure in CuSc and CuHf specimens without the utilization of cold rolling contained rod-like and needle-like structures (Figure 8b and Figure 12b). The 75% cold rolled specimen contained more spherical intermetallic phases (Figure 9b and Figure 13b).

Regarding the hardness development during the aging experiments, a lower aging temperature and, respectively, lower diffusion in the material, combined with many potential nucleation sites for precipitates, appeared to provide the most advantageous combination of all strengthening influences. At higher temperatures, the precipitates coarsened because the minimization of the total interfacial energy is the driving force [1]. Therefore, the amount of visible precipitates in optical microscopy, especially noticeable for CuHf0.7 and CuSc0.3, decreased, but their size increased (Figure 8b, Figure 9b, Figure 12b and Figure 13b). For the electrical conductivity the amount of solute elements in the copper matrix is the most crucial factor due to a noticeable scattering of conduction electrons. Therefore, the precipitating volume fraction, which increased continuously over time, resulted in a steady improvement of the electrical conductivity (Figure 3a, Figure 6a and Figure 10a). Enhanced aging temperatures and cold rolling prior to the aging treatment promoted this effect (Figure 3a, Figure 6a and Figure 10a).

During continuously heated measurements in a DSC experiment, several thermic effects can occur and might overlay the signal. Therefore, a careful investigation, including other experimental setups, was essential to describe the materials’ behavior in-depth. The interacting effects of dislocation annihilation, material regeneration, precipitation formation, and growth, as well as recrystallization, need to be considered for the isothermal aging and the constantly heated DSC experiments. In none of the experiments, a solution of alloying elements after precipitation appeared up to the maximum temperature of 800 °C (condition 2 of DSC measurements).

To draw an exothermic energy release back on the fraction transformed to further analyze the influence of cold deformation prior to the aging process requires highly comparable specimens [76]. Since nucleation and precipitation growth are driven by supersaturation [9,75,77], the identical alloy composition for comparison is mandatory. It was necessary to set acceptable ranges of deviation for the alloying contents (of 0.05 wt.%), which remained limiting. Especially for low alloyed copper alloys, a small variation resulted in remarkable differences. Further analysis with a focus on exothermic energy release and a calculation of activation energies will include specimens of different degrees of deformation located in the cast product next to each other.

4.2. Discussion of the Different Alloying Concepts

In conclusion, Figure 15 directly compares outcomes of precipitation in the three alloys. CuSc0.3, without cold rolling prior to the aging process, began to precipitate at 514 °C and peaked at approximately 568 °C. This effect was shifted by 75% cold rolling to an onset temperature of 429 °C and a peak temperature of 470 °C. For direct comparison, CuHf0.7 started the reaction at 492 °C and culminated at 563 °C without mechanical processing steps, whereas 75% cold rolling shifted the peak to an onset temperature of 456 °C and a peak temperature of 491 °C.

CuCr0.7 appeared with remarkably different behavior. For this alloy, the aging temperature is the most important factor in controlling the atom movement. An optional cold rolling prior to the aging process was beneficial but did not have a huge impact on the characterizing temperatures.

The hardness and electrical conductivity measurements continuously matched the investigated thermal analysis with DSC and visualized the effects of the characteristic behavior of all three alloys.

The influence of time and temperature on the thermally activated and kinetically controlled solid-state reactions [60,71], reported within the peak shifting influence of different heating rates in DSC measurements [59,69], became visible in the comparison between the comparably high reaction temperatures during the DSC measurements and the isothermal aging experiments. The highest mechanical properties, characterized by hardness monitoring, were identified at lower aging temperatures and more extended aging durations than the occurrence temperature of effects in DSC. A further investigation of time-dependent alloy optimization was not the focus of this research but can be reproducibly related to isothermal aging experiments [26].

Furthermore, the characteristic aging experiments measure the properties as a conglomerate of effects leading to a certain result. Therefore, not only the precipitates strengthened the material. Other hardening mechanisms, especially dislocation or solid solution hardening, change their contribution due to thermal treatments. Consequently, the aging treatment of CuSc0.3 and CuHf0.7 showed the maximum hardness for high degrees of deformation and low aging temperatures.

Regarding mechanical properties, CuHf0.7 and CuSc0.3 provided the most beneficial results. In the case of 75% cold rolling prior to aging, both alloys reached hardness measurements of approximately 200 HV_0.1 (isothermal aging at 350 °C or 375 °C for 12 h or 2 h). On the contrary, CuCr0.7 reached its optimum (75% cold rolling) of 180 HV_0.1 at 425 °C after 4 h but provided optimum electrical conductivities of about 56 MS/m to 57 MS/m if aged at temperatures above 425 °C (Figure 3a). Comparable specimens of CuHf0.7 and CuSc0.3 reached at least 52 MS/m (Figure 6a and Figure 10a). All alloys achieved the highest conductivities for high degrees of cold rolling prior to the aging process and elevated aging temperatures after 48 h. Continuing precipitation reduced the content of alloying elements in the copper matrix, and precipitation ripening and coarsening also helped to reduce the amount of crystallographic defects. Therefore, the scattering of conduction electrons and the specific electrical resistance decreased, respectively, the electrical conductivity increased.

Due to the comparatively high atomic radius of hafnium, its diffusion in the copper lattice is lower, resulting in elevated temperatures for precipitation reactions [2]. Paths of facilitated diffusion, introduced during cold rolling [1,75], strongly promote the movement of scandium and hafnium atoms in the copper matrix. On the other hand, chromium atoms can move more easily because of a similar atomic radius to the copper matrix atoms. It can be concluded that the scandium and hafnium atoms need diffusion facilitation along high diffusivity paths [1,78] more than chromium. Especially changes in electrical conductivity show how solute alloying elements segregate and confirm the DSC results.

Furthermore, X-ray diffraction (XRD) measurements were conducted to analyze the crystal structure in the alternative alloying systems of copper with scandium and hafnium, which are rarely discussed in the literature. The limited resolution in the XRD measurement and hardly definable peaks in these low-alloyed copper alloys made the identification of the phases difficult. Referring to Hao et al., the Cu₄Sc-phase should be present in aged CuSc-alloys due to the strong correlation between processing, properties, and the EDS results (Appendix B Figure A4).

Regarding CuHf, the precipitation sequence has not been analyzed in-depth in literature yet but is consistently related to the process in CuZr-alloys. In the CuZr alloy system Watanabe et al. refer to early coherent fcc precipitates as a Guinier–Preston (GP) zone or an intermediate phase [79]. Peng et al. and Zhang et al. report the presence of rod-shaped Cu₅Zr precipitates, which form from the supersaturated solid solution to clusters and semi-coherent structures afterwards [80,81]. Nakashima et al. add disk-shaped Cu₅Zr phases to be considered [82]. Further research, including TEM analysis, should help to confirm the assumed phase of Cu₅Hf and the corresponding precipitation sequence to build a basis for further discussions. Regarding the DSC measurements, Bochvar et al. observed a peak at 572 °C, in a comparable solution annealed and quenched binary CuHf0.9 alloy [43], which is very close to the corresponding peak of 561 °C in this work (identical heating rate). In direct comparison, this measured exothermic effect of the 0% option covered an extended temperature range and included a second exothermal effect itself. Due to the specimens’ processing history, for this effect, a change in the precipitates’ structure appears to be the most probable mechanism.

Taking into account the other DSC curves for specimens with cold deformation prior to the aging process, this effect in the precipitation sequence is not visible. Due to the accelerated solid-state reaction, these effects could be below the resolution limit of this heating rate. Therefore, a reduced heating rate could be used. Nevertheless, the cold rolled solid solution facilitated the precipitation process and shifted the DSC peak to considerably lower temperatures (Figure 15).

In Bochvar et al., equal channel angular pressing (ECAP) processing of CuHf0.9 six times shifted this precipitation peak to 402 °C [43]. The discussed influence of deformation to provoke the reaction continues consistently.

Dalan et al. reported a precipitation peak at 441 °C after solution annealing at 1020 °C for specimen of CuCrZr (0.81 wt.% Cr and 0.08 wt.% Zr) without any influence of mechanical deformation, or 493 °C after solution annealing at 930 °C (10 K/min heating rate as well) [75]. In this comparative study, the binary CuCr0.7 showed a peak temperature of 468 °C after solution annealing at 1000 °C, which fits into the picture and underlines the high comparability of the results.

According to Dalan et al., one ECAP processing reduced the peak temperature of CuCrZr to 438 °C and 461 °C, respectively. Further ECAP passes reduced the peak temperature in several investigations for CuCrZr [50,75] and binary CuCr0.7 [43], but this shift’s magnitudes turned out to be comparatively small. Therefore, the investigated trend in the results of this experimental study continues in further excessive processing procedures.

For CuCrZr, which is of great industrial interest, Bourezg et al. reported on the reaction kinetics after ECAP processing. Several studies observe three peaks at comparable temperatures and refer them to the clustering of chromium, followed by nanoprecipitation of the Cu₅₁Zr₁₄ [52], respectively Cu₃Zr [50] and recrystallization. Measurements of binary CuCr0.7 in this research, as well as the results of Bochvar et al. [43], showed secondary peaks after the main precipitation occurred as well, which were attributed to recovery effects without further justification [43]. Due to the high comparability regarding the first-occurring peak, a deeper analysis would be informative, but is beyond the scope of the present studies. For this work, CuCr0.7 is only a reference for comparison with the two alternative alloying systems and validation of the experimental setup.

In the case of the deformed specimens, the recrystallization peaks partly appeared less intense during the DSC measurements which could be improved using an optimized higher heating rate for the experiment. Nevertheless, the visibility of thermal effects needs to be ensured by changing the heating rate. With respect to recrystallization, the benefits of the alternative alloying approaches should be emphasized. For CuHf0.7, the peak attributable to recrystallization was observed at about 718 °C and for CuSc0.3 at approximately 648 °C (75% cold rolling). In contrast, the benchmark alloy CuCr0.7 with 75% cold rolling was already recrystallized in metallographic investigations of condition 1, which matched the recrystallization peak at average 577 °C.

All results, including hardness, electrical conductivity, DSC measurements, and the accompanying metallographic investigations, contributed to a consistent understanding of different low-alloyed copper alloys to optimize the targeted manipulation of applied precipitation processes. Therefore, the interacting use of thermomechanical treatments and different binary alloying concepts showed favorable results on low alloyed copper alloys with optimized mechanical properties and good conductivities.

5. Conclusions

Thermomechanical treatment had an essential influence on precipitation processes in supersaturated solid solutions of copper containing chromium, scandium, and hafnium. Measuring different outcomes of the heat treatments enabled to draw a consistent conclusion of active processes in the CuCr0.7, CuHf0.7, and CuSc0.3 alloys:

• All alloys, CuCr0.7, CuSc0.3, and CuHf0.7, showed substantial increases in hardness and electrical conductivity due to precipitation effects. A lower aging temperature and a higher degree of cold deformation resulted in an excellent interaction of strengthening effects. Regarding electrical conductivity, which was highly sensitive to the amount of alloying elements in the copper matrix, higher degrees of cold deformation, long aging durations, and enhanced temperatures were favorable due to a maximized segregation of alloying elements from the copper matrix.
• The time and temperature-dependent solid-state reactions resulted in different onset, peak temperatures of the continuously heated DSC measurements, and optimal aging temperatures in isothermal aging experiments. The enhanced heating rate during the DSC method shifted the effects to higher temperatures.
• The alloys CuSc0.3 and CuHf0.7 showed comparable precipitation behavior, including the onset and peak temperatures which reacted susceptible to cold deformation prior to the aging process. A cold rolled supersaturated solid solution promoted the precipitation of fine homogenously distributed precipitates. The difference between as-cast temperature treated specimens and all other cold rolled options was the most substantial. The degree of cross-section reduction due to cold rolling did not have a huge impact but continuously followed the trend to promote the precipitation reaction.
• CuCr0.7 appeared to be mainly influenced by the aging temperature. An initial cold rolling shifted the solid-state reactions to lower temperatures, but in a considerably lower degree compared to CuSc0.3 and CuHf0.7.
• The alternative alloying concepts containing scandium and hafnium show excellent benefits due to recrystallization appearing at conspicuously higher temperatures, compared to the benchmark alloy CuCr0.7.

The thorough discussion in this research work has contributed to generating an in-depth view of which effects impact a copper-based alloy during heat treatment. Especially, moderate degrees of deformation have been emphasized to scoop the precipitation potential of the alternative alloying concepts further, which had not been discussed and directly compared in literature so far. Hafnium and scandium containing low alloyed high-performance copper alloys appeared to be well-equipped for applications at enhanced operating conditions. Future studies will determine the influence of combining different alloying elements to use the potential of alternative alloying concepts, including scandium and hafnium.