Development of Y₂O₃ Dispersion-Strengthened Copper Alloy by Sol-Gel Method

Abstract

In this study, oxide dispersion-strengthened Cu alloy with a Y₂O₃ content of 1 wt.% was fabricated through citric acid sol-gel synthesis and spark plasma sintering (SPS). The citric acid sol-gel method provides molecular mixing for the preparation of precursor powders, which produces nanoscale and uniformly distributed Y₂O₃ particles in an ultrafine-grained Cu matrix. The effects of nanoscale Y₂O₃ particles on the microstructure, mechanical properties and thermal conductivity of the Cu-1wt.%Y₂O₃ alloy were investigated. The average grain size of the Cu-1wt.%Y₂O₃ alloy is 0.42 μm, while the average particle size of Y₂O₃ is 16.4 nm. The unique microstructure provides excellent mechanical properties with a tensile strength of 572 MPa and a total elongation of 6.4%. After annealing at 800 °C for 1 h, the strength of the alloy does not decrease obviously, showing excellent thermal stability. The thermal conductivity of Cu-1wt.%Y₂O₃ alloy is about 308 Wm⁻¹K⁻¹ at room temperature and it decreases with increasing temperature. The refined grain size, high strength and excellent thermal stability of Cu-1wt.%Y₂O₃ alloys can be ascribed to the pinning effects of nanoscale Y₂O₃ particles dispersed in the Cu matrix. The Cu-Y₂O₃ alloys with high strength and high thermal conductivity have potential applications in high thermal load components of fusion reactors.

1. Introduction

Copper alloys are the main candidate materials for the heat sink of the water-cooled target of a divertor in nuclear fusion reactors, due to its high thermal conductivity, strength and good radiation resistance [1,2,3,4,5]. Pure copper has a very high thermal conductivity but its strength is relatively low especially at high temperatures. Furthermore, the service life of pure copper is limited because of its high creep, swelling rate and irradiation hardening [6,7]. Therefore, it is necessary to develop copper alloys with both high strength and good thermal conductivity.

CuCrZr alloys are the primary candidates for heat sink materials in the International Thermonuclear Experimental Reactor (ITER) due to its high thermal conductivity and high strength at medium and low temperature [8,9,10]. However, long-term thermal exposure of CuCrZr alloy to excessive temperature (>400 °C) leads to loss of strength due to over-ageing where the precipitates undergo Ostwald ripening losing their effective resistance as a barrier against dislocation glide [11]. Furthermore, neutron irradiation will cause degradation of thermal conductivity and mechanical properties of the alloy [12]. These drawbacks limit the operative temperature window of CuCrZr alloys. Developing a high-performance copper alloy to provide the best combination of excellent mechanical strength and good thermal conductivity at high temperatures is the main trend. Dispersion-strengthened copper alloys are an effective method to improve the high temperature strength of Cu-based materials. Highly stable ceramic particles (such as Al₂O₃ [11,13], TiC [14,15], NbC [16] and Y₂O₃ [17,18,19]) dispersed in the copper matrix can effectively hinder the motion of dislocations and migration of the grain boundary and extend the operating temperature range of copper alloys [20]. Among them, the Y₂O₃ dispersion-strengthened Cu alloys attract intense attention due to their excellent mechanical strength with higher thermal conductivity. The Cu-0.46wt.%Y₂O₃ alloy (Glidcop Al25) has been one of the Cu alloys considered for ITER [5,21].

Recently, Y₂O₃ dispersion-strengthened copper alloys have attracted increased attention. Compared with alumina, Y₂O₃ is relatively more suitable because of the low solubility of element Y in copper (less than 0.05 wt.% at 300 °C), which helps to maintain a high thermal conductivity [17]. In addition, the high thermodynamic stability of Y₂O₃ enhances the coarsening resistance at high temperatures [22]. Many efforts have been devoted to develop Y₂O₃ dispersion-strengthened copper alloys by Mechanical Alloying (MA) methods [6,7,17,18,23,24,25,26]. For example, G. Carro et al. [6] investigated the microstructural and mechanical properties of Cu-0.8wt.%Y produced by a powder metallurgy route and subsequent consolidation by hot isostatic pressing. The resultant equiaxed grain size distribution ranged from 0.5 μm to 50 μm, which greatly weakens the effect of fine grain strengthening. S.M.S. Aghamiri et al. successfully prepared a Cu-0.42wt.%Y₂O₃ alloy with comparable strength to the Glidcop-Al25 alloy by adding stearic acid as a process control agent and using MA and spark plasma sintering (SPS) [23]. Bing MA et al. explored the influence of Cu-Y compound content on the microstructure of Cu-Y₂O₃ alloys synthesized by MA and HIP process [26]. However, Y₂O₃ particles are prone to aggregation and growing up in the MA synthesized Cu-Y₂O₃ alloys, and impurities are easily introduced in the MA process. As the mechanical properties of dispersion-strengthened alloys are closely linked to the size and distribution of dispersoids, the segregation and large size of Y₂O₃ in copper alloys would inevitably reduce the mechanical properties.

In this work, a citric acid sol-gel method [27] involving molecular level mixing was adopted to synthesize nanoscale Y₂O₃ particles homogeneously dispersed copper powders, and ultrafine-grained Cu-1wt.%Y₂O₃ alloy with high strength, high thermal conductivity and excellent thermal stability was obtained through spark plasma sintering (SPS) from sol-gel synthesized powders. The influence of Y₂O₃ nanoparticles on the mechanical properties, microstructure and thermal conductivity of the alloys was investigated and the corresponding mechanisms were discussed.

2. Materials and Methods

Cu-1wt.%Y₂O₃ alloy powders were synthesized using a citric acid sol-gel method. Raw materials for the synthesis of Cu-1wt.%Y₂O₃ powders are listed in

Table 1. Raw materials for the synthesis of Cu-1wt.%Y₂O₃ powders.

Raw Materials	Molecular Formula	Purity
Citric acid monohydrate	C6H8O7·H2O	≥99.5%
Copper nitrate trihydrate	Cu(NO3)2·3H2O	≥99%
Yttrium nitrate hexahydrate	YN3O9·6H2O	≥99.99%
Polyethylene glycol	HO(CH2CH2O)nH	-

. The specific experimental process is as follows: C₆H₈O₇·H₂O was firstly dissolved in deionized water to create an acidic environment, then Cu(NO₃)₂·3H₂O and YN₃O₉·6H₂O were added into the solution. Citric acid is employed as chelating agent, which can provide the mixing of cations at the molecular level in the sol-gel process. The solution was heated at 80 °C with continuous stirring. An appropriate amount of polyethylene glycol (PEG, molecular weight 20,000) was added as surfactant after stirring at 80 °C for 2 h. The temperature was maintained at 80 °C with continuous stirring to remove excess water until a gel formed. The resulting gel was dried at 140 °C for 12 h and then was calcined at 550 °C for 6 h in air to remove the organic compounds, leading to a black mixture of CuO and Y₂O₃ powders. The calcined oxide powders were further ground to crush large particles, and subsequently, the oxides mixture was reduced in flowing hydrogen at 400 °C for 1 h, resulting in Cu-1wt.%Y₂O₃ powders.

The sol-gel synthesized Cu-1wt.%Y₂O₃ powders were consolidated at 900 °C for 5 min under a pressure of 50 MPa using an SPS method to obtain bulk samples and the specific process can be seen in the literature [28]. The size of the sintered samples was 30 mm in diameter and about 3 mm in thickness. The densities of the samples were determined by the Archimedes principle using distilled water as the immersion liquid. The theoretical density of Cu-1wt.%Y₂O₃ alloy is 8.85 g/cm³ according to the rule of mixtures and the relative density of the sintered samples is about 97.6%.

The samples for the tensile tests were machined to a gauge dimension of about 5 × 1.5 × 0.75 mm³ using an electrical discharge machine and slightly polished. The tensile tests were carried out at room temperature (RT) with an Instron 5967 testing machine at a constant speed of 0.3 mm/min (corresponding to a nominal strain rate of 1 × 10⁻³ s⁻¹) in air.

The thermal conductivity (γ) was calculated from the thermal diffusivity (α), density (ρ) and specific heat (C_p) according to the relation: γ = αC_pρ. The α and C_p was determined using the laser flash diffusivity system (LFA457 Micro flash, NETZSCH). The sample was cut into cylinders with a diameter of about 12.6 mm and a thickness of about 2.5 mm and polished smooth with sandpaper. The thermal conductivity was measured at the Engineering Materials Science Experiment Center of University of Science and Technology of China (USTC, Hefei, China).

The sol-gel synthesized oxides precursor and reduced Cu-1wt.%Y₂O₃ powders were characterized by X-ray diffraction (XRD, X’Pert) with Cu Kα radiation. The morphology of the powders and tensile fracture of bulk samples was characterized by a field emission scanning electron microscope (FE-SEM, SU8020). Electron backscatter diffraction pattern (EBSD) mappings were collected using a Zeiss SIGMA field emission scanning electron microscope (SEM) equipped with a CRYSTAL detector (Oxford Instruments, Oxfordshire, UK) and the acceleration voltage was 13 kV. A misorientation angle of θ > 10° was used to distinguish the grain boundaries. The EBSD images were slightly filtered through the software (HKL Tango) to eliminate the influences of second phase particles on Cu grain observations. Microstructure characterization was carried out by a transmission electron microscopy (TEM, Tecnai G2 F20) at 200 kV to determine the size and distribution of Y₂O₃ particles in Cu-1wt.%Y₂O₃ alloys. Energy-dispersive X-ray spectroscopy (EDS, INCA) installed in the TEM was used for elemental analysis. The TEM samples were prepared by cutting flaked samples with diameters greater than 3 mm from the SPS sintered sample. First, the flakes were ground to about 50 μm in thickness. Subsequently, a Gatan 659 disk punch was used to cut a 3-mm-diameter disk from the thin wafer and then fixed on a molybdenum ring with an outer diameter of 3 mm. Then the specimen was transferred to Gatan 623 and further thinned to about 30 μm. Finally, the precision ion milling system (Gatan model 691) was used to polish the disk specimens at 4.3 keV to obtain electron transparency in the middle part. Note that the final angle and voltage of ion thinning for the TEM samples preparation in this study were 3° and 3.2 keV, respectively.

3. Results and Discussion

3.1. Characteristics of Powders

Figure 1 shows the XRD patterns of the sol-gel synthesized precursors before and after hydrogen reduction. The powder composition before hydrogen reduction is mainly composed of CuO and Y₂O₃ through software (Highscore plus 3.0) analysis. The XRD diffraction peaks at 35.55°, 38.67° and 48.84° correspond to the (11-1), (111) and (20-2) plane of CuO (PDF: 98-004-3179). After hydrogen reduction at 400 °C for 1 h, the XRD peaks at 43.33°, 50.46° and 74.15° corresponding to the (111), (002) and (022) plane of Cu (PDF: 98-062-7113) indicates that the CuO powders were completely reduced. The XRD diffraction peaks at 20.49°, 29.14° and 33.77° correspond to the (112), (222) and (004) plane of Y₂O₃ (PDF: 98-001-6394). However, no diffraction peak of yttrium oxide was found in XRD. This may be due to the low content and poor crystallinity of yttrium oxide, which will be further analyzed later in the powder TEM. Figure 2a shows the SEM images of the sol-gel synthesized precursor powders, consisting of CuO and Y₂O₃. The particle size distribution of the precursor powders is 20~200 nm with an average particle size of about 80 nm. After hydrogen reduction at 400 °C for 1 h, Cu particles grow up and sintering necks are formed between some Cu particles, as shown in Figure 2b,c. Figure 2d shows the high-magnification image of the rectangular region marked with a red dotted line in Figure 2c. Some nanoscale particles were evenly distributed on the surface and inside of the copper particles and the uniform distribution of these nanoparticles in copper powders is the prerequisite for obtaining high performance Cu-Y₂O₃ alloys.

Figure 3a,b shows the TEM images of the powders reduced at 400 °C for 1 h. Most of the particles are spherical, and some sintering necks were formed between the particles. Figure 3c shows the EDS results of Figure 3a, and the results show that there are three elements Cu, O and Y in the powders. Selected area electron diffraction (SAED) of Figure 3a is shown in Figure 3d; the result shows that the powders are composed of Cu and Y₂O₃. Elemental mapping analysis was performed on the rectangular region in Figure 3b to determine the composition of these particles, as shown in Figure 3e. The results show that most of these particles are Cu, while the distribution of Y and O elements and SAED results further confirm the existence of Y₂O₃. Figure 3f shows the high resolution TEM (HRTEM) image of the particles indicated by the rectangle in Figure 3e. The measured interplanar spacings of the two particles are 0.306 nm and 0.433 nm, respectively, corresponding to the (222) and (112) crystal planes of Y₂O₃ (PDF: 98-001-6394).

3.2. Microstructure Characterization

The mechanical properties of dispersion-strengthened materials is closely linked to the grain structure and the size and distribution of the strengthening phases. Figure 4a,d shows the inverse pole figure (IPF) maps of as-sintered Cu-1wt.%Y₂O₃ alloy and the corresponding grain size distribution. The majority of Cu grains in the as-sintered Cu-1wt.%Y₂O₃ alloys are in the submicron range with an average grain size of 0.42 μm. Compared with the Cu-5vol%Y₂O₃ alloy (average grain size 1.15 μm) prepared by mechanical alloying [24], the grain size of the Cu-1wt.%Y₂O₃ alloy is obviously refined. After annealing at 800 °C, the grain growth in the Cu-1wt.%Y₂O₃ alloy is not obvious with an average grain size of 0.45 μm, as shown in Figure 4b,e. As the annealing temperature increases to 900 °C, the average grain size of the Cu-1wt.%Y₂O₃ alloy increases to 0.54 μm, and the annealed sample consist of two different types of grains: larger recrystallized grains and smaller submicron grains, as shown in Figure 4c,f. This phenomenon is attributed to the recrystallization of copper during high-temperature annealing. The recrystallized grains nucleate from the grain boundary and then many small submicron grains grow into recrystallized grains. Meanwhile, the Y₂O₃ particles distributed in the Cu matrix can effectively inhibit the growth of the grain and many grains are still smaller than 0.5 μm in size even after 900 °C annealing. Overall, the recrystallization temperature of Cu-1wt.%Y₂O₃ alloy samples is significantly higher than that of the Cu samples, and Y₂O₃ particles plays a significant role in the improvement of grain refinement and thermal stability.

In addition to the effect of grain size, the size and distribution of Y₂O₃ particles in the Cu-1wt.%Y₂O₃ alloy play an important role in improving the strength of the material. Figure 5 shows the dark field TEM (DF-TEM) images, particles size distribution of Y₂O₃ and elemental distribution mapping of the Cu-1wt.%Y₂O₃ alloy. It can be seen from the DF-TEM image in Figure 5a that there are many twins, most of which have grain sizes below 500 nm, and some larger Y₂O₃ particles are distributed at the grain boundaries. The specific distribution and size of Y₂O₃ particles on Cu are shown in Figure 5b,c. Nanoscale Y₂O₃ particles with an average size of about 16.4 nm are uniformly distributed in the interior and grain boundary of copper grains. Many Y₂O₃ particles with a particle size smaller than 10 nm are evenly distributed in the grains. Dispersing nanosized particles in the grain interior is an effective method to improve the strength and ductility, because the intragranular particles can generate, pin down and thus accumulate dislocations within the grains [29]. Figure 5d shows the high resolution transmission electron microscope (HRTEM) image of the interface between Y₂O₃ particles and the copper matrix. The measured interplanar distance of the Y₂O₃ particle and Cu matrix is 0.306 nm and 0.209 nm, respectively, corresponding to the values of the Y₂O₃ (222) and Cu (111) planes from the Powder Diffraction File (PDF) card of 98-001-6394 and 98-006-4699. In addition, the local area of particle distribution is analyzed through elemental distribution mapping, as shown in Figure 5e,f. Through the elemental distribution mapping of Cu, Y and O and analysis results of HR-TEM, it can be determined that the particles uniformly distributed in Cu matrix are Y₂O₃ particles. In fact, the uniform distribution and small size are attributed to the Cu-1wt.%Y₂O₃ alloy powders prepared using the citric acid sol-gel method and the short sintering time by SPS, which effectively reduces the number and size of Y₂O₃ particles at the grain boundaries. These Y₂O₃ nanoparticles with a high melting point and high hardness greatly improve the strength and thermal stability of the alloy.

Figure 6 shows HAADF-STEM images and Y₂O₃ particles size distribution of the Cu-1wt.%Y₂O₃ samples annealed at 800 °C and 900 °C. After annealing at 800 °C, the average particle size of Y₂O₃ is 17.0 nm and the size and distribution of Y₂O₃ particles do not change significantly. After annealing at 900 °C, the grain size increased slightly to 17.6 nm. This phenomenon is consistent with the change of mechanical properties, which further explains the high thermal stability of the copper alloy.

3.3. Mechanical Properties

Figure 7 shows the engineering stress-strain curves of the Cu-1wt.%Y₂O₃ alloy and pure Cu annealed at 800 °C and 900 °C. The ultimate tensile strength (UTS) and total elongation (TE) of the alloy at RT are 572 MPa and 6.4%, respectively (Figure 7a). The strength is significantly improved compared with the pure Cu sample prepared by the same process (Figure 7b). In addition, compared with the Y₂O₃ dispersion strengthened copper alloy prepared by MA, the strength of the alloy has been greatly increased. For example, Zhou et al. reported Cu-5vol%Y₂O₃ alloys prepared by mechanical alloying, high temperature heat treatment and powder compact extrusion, with UTS and TE of 389 MPa and 8.9%, respectively [24]. The UTS of the present Cu-1wt.%Y₂O₃ alloys is similar to that of the Cu-60wt.%W composite prepared using the melt-infiltration method [30]. Compared with precipitation-strengthened CuCrZr alloy and dispersion-strengthened CuAl-25 alloy, the strength of the Cu-1wt.%Y₂O₃ alloy is obviously improved, while the ductility is decreased [31]. The low ductility of the Cu-1wt.%Y₂O₃ alloy may arise from the relatively high porosity (2.4%) and the particles distributed at the grain boundaries. The pores and intergranular Y₂O₃ particles would act as stress concentrators and cause localized plastic deformation [30]. After annealing at 800 °C, the UTS of Cu-1wt.%Y₂O₃ alloys had no obvious change with a high value of 563MPa and the TE was 5.2%. After annealing at 900 °C, the UTS decreased to 532 MPa, which may be caused by partial recrystallization at this temperature. In a word, Cu-1wt.%Y₂O₃ alloys prepared using sol-gel method have a high strength and good thermal stability.

Figure 8 shows the tensile fracture morphologies of the as-sintered Cu-1wt.%Y₂O₃ and the ones annealed at 800 °C and 900 °C. Many tiny dimples can be seen in the fracture morphology of the SPSed Cu-1wt.%Y₂O₃ alloy, indicating ductile fracture. The size and distribution of dimples in the sample annealed at 800 °C is similar to that of the SPSed one (Figure 8b). For the sample annealed at 900 °C (Figure 8c), the dimples appear to be larger and deeper, which may be caused by the recrystallization of the alloy.

3.4. Thermal Conductivity

The thermal conductivity of copper alloys is one of the most important parameters that determines whether they can be used as heat sink material for future fusion reactors. However, there are few mentions in the literature about the thermal conductivity of Y₂O₃ dispersion strengthened copper alloy. Figure 9 shows the thermal conductivity of the Cu-1wt.%Y₂O₃ alloy was measured at different temperatures from RT to 800 °C. The thermal conductivity of the Cu-1wt.%Y₂O₃ alloy at room temperature is 308 Wm⁻¹K⁻¹, which is lower than that of pure copper. The decrease in thermal conductivity after Y₂O₃ addition can be ascribed to the relatively lower thermal conductivity of Y₂O₃, and more defects such as the grain boundaries and Cu/Y₂O₃ phase interfaces than in pure Cu. These defects would scatter electrons and therefore reduced the thermal conductivity [28]. The thermal conductivity of the alloy decreases with the increasing temperature. Even so, the thermal conductivity at 800 °C of the Cu-1wt.%Y₂O₃ alloy is above 250 Wm⁻¹K⁻¹, which is still superior to the value of the W-Cu composite [32].

4. Conclusions

Nanostructured Cu-1wt.%Y₂O₃ alloy with high strength, high thermal conductivity and good thermal stability was fabricated by sol-gel synthesis and the SPS method. The sol-gel process involving molecular mixing can be used to prepare nanoscale Y₂O₃ homogenously dispersed copper powders. The average particle size of Y₂O₃ in Cu-1wt.%Y₂O₃ alloy is only 16.4 nm, which can effectively refine grains and improve the strength and thermal stability of the material by pinning dislocations and grain boundaries. The average grain size of a Cu-1wt.%Y₂O₃ alloy is about 0.42 μm. The tensile strength and thermal conductivity of the nanostructured Cu-1wt.%Y₂O₃ is 572 MPa and 308 Wm⁻¹K⁻¹, respectively. The strength and microstructure of the alloy do not change significantly after annealing at 800 °C, indicating excellent thermal stability. This work provides a promising method for fabricating high performance nanoparticle dispersion-strengthened Cu alloys.