Divertors of impurities and helium ash from burning plasma in fusion reactors are one of the components in nuclear fusion reactors that experience extremely high heat fluxes (You et al., 2021). In addition to high-melting-point plasma facing components (PFCs), divertors require high-strength, creep-resistant, high thermal conductivity materials to efficiently draw heat from PFCs (Asakura et al., 2023). The International Thermonuclear Experimental Reactor (ITER) is a scientific research reactor currently under construction in France which is intended to achieve burning plasma, test the safety and functionality of integrated fusion power plant designs, and demonstrate the generation of fusion power at levels ten times the input heating power needed to sustain plasma (Siccinio et al., 2020; ITER Organization, 2024). The European demonstration power plant, EU-DEMO, is ITER’s planned successor and will be a larger reactor intended to produce a net electrical power output (×25 input) on a continuous basis (EUROfusion, 2024). EU-DEMO will use and build upon knowledge gained from ITER, but its larger size and broader power generation goals require materials with operational limits significantly higher than required for ITER operation. High heat flux materials for divertors may face an order of magnitude greater total radiation damage, an estimated 6–13 dpa greater in EU-DEMO than in ITER, (You et al., 2021) and experience temperatures 300–400 °C (Zinkle and Ghoniem, 2000; Zinkle, 2016).

The high electrical conductivity (and thus high thermal conductivity, due to the Wiedemann–Franz law) of Cu makes it attractive for high heat flux components; however, pure Cu lacks the mechanical properties needed in high temperature reactor conditions. Alloys such as GlidCop Al25 (Fabritsiev and Pokrovsky, 2007), GRCop-84 (Groh et al., 2008), CuNiBe (Zinkle, 2014), Cu–Al₂O₃, and CuCrNiSi (Wang et al., 2015) have been developed in the hopes of producing stronger alloys without degradation of the thermal conductivity. Cu–Cr–Zr (CCZ) alloys have been identified as a promising candidate for commercial reactor heat exchange components. CCZ was selected for use in the divertors, first wall, and blanket components of the ITER fusion facility (Barabash et al., 2007) due to its relatively high thermal conductivity and good mechanical properties up to 300 °C, with an ultimate tensile strength of 325 MPa at 300 °C in peak aged condition (Edwards et al., 2007). Although CCZ has been studied and developed based on ITER specifications (Barabash et al., 2011; Gusarov et al., 2018; Fabritsiev and Pokrovsky, 2005), the greater dose and temperature demands of the DEMO reactor are beyond the optimal operating limits achieved with ternary CCZ alloys (Zinkle, 2016). In particular, based on limited testing, the creep strength of CCZ falls rapidly with increasing temperatures above 300 °C, with the design allowable stress for ITER-grade CCZ falling below 50 MPa at temperatures above ∼360 °C due to thermal creep. (Asakura et al., 2023; Terentyev et al., 2023; Zinkle and Ghoniem, 2000; Zinkle, 2016; Zinkle and Fabritsiev, 1994; Barabash et al., 2011; You, 2015; Groh et al., 2008).

New Cu–Cr–Nb–Zr (CCNZ) alloys were designed to improve the strength and creep life of CCZ alloys by combining the fine precipitate strengthening of CCZ with the Laves phase strengthening of Cu–Cr–Nb (GRCop-84) alloys (Yang et al., 2018; Wang et al., 2020; Cheng et al., 2021; Krishna et al., 2022; Anderson et al., 1995; Groh et al., 2008). This was achieved through a bimodal distribution of precipitates with coarse Cr₂(Nb,Zr) Laves phase precipitates at grain boundaries (GB) as well as nanoscale Cr and Cu₅Zr nanoprecipitates in the matrix. The Cr₂(Nb,Zr) GB precipitates can effectively pin grain boundaries and mitigate grain boundary sliding-induced creep. The measured creep rupture times for these improved CCNZ alloys is about ten times higher than the ITER reference CCZ alloys at 90–125 MPa at 500 °C (Gusarov et al., 2018; Wang et al., 2020). The tensile properties and electrical conductivity of the advanced CCNZ and ITER CCZ alloys are comparable. However, despite the need for a clear understanding of the neutron irradiation effects on alloys designed for use in cooling channels of ITER and DEMO divertors, only limited information is available on these material properties (Barabash et al., 2011; Kalinin et al., 2011; Li and Zinkle, 2020).

Given this knowledge gap, this study focuses on the room temperature electrical and mechanical properties of neutron irradiated CCNZ alloys and compares them to CCZ alloys irradiated under the same conditions. Through subsequent microstructural evaluation and transmutation calculations, we correlated this data with the various radiation-induced changes in these alloys.

Cu alloys with a nominal composition of Cu-2Cr-1.35Nb-0.15Zr (wt%) were arc-melted from pre-alloyed bulk Cu-21.3wt%Nb, Cr, and Zr and drop cast to form bars. The cast bars were subjected to multiple passes of cold rolling followed by high temperature (970 °C) solution annealing and subsequent thermal aging at 475 °C for 3 h, resulting in a microstructure containing a high density of nanoscale Cr precipitates and larger Cr₂Nb precipitates. The detailed experimental procedure for Cu alloy casting and processing was described in detail by Yang et al. (2018) and Wang et al. (2020). ITER-grade CCZ specimens with a nominal composition of Cu-0.81Cr-0.07Zr were fabricated in a similar process, with a cold rolled sample solution annealed at 980 °C before thermal aging at 475 °C for 3 h to precipitate nanoscale Cr precipitates. Wire electrical discharge machining (EDM) was used to cut samples into SSJ3 tensile specimens. CCNZ and CCZ specimens were irradiated for three cycles in the HFIR flux trap facility to target a damage dose of 5dpa (based on the stainless steel equivalent) and irradiation temperatures (T_ir) of 250–350°C. It completed up to ∼6454 MWD (at nominal operating power of 85W) and was exposed to a total fast fluence of 6.8–7.6 (×10²⁵) n/m² for E > 0.1 MeV. This resulted in displacement damage of 7.8e14 n/cm^2-s (E > 0.1 MeV), which is roughly 7.2e-7 dpa/s in Cu, assuming E_d = 30 eV. The corresponding thermal neutron flux is 2.2e15 n/cm^2-s. The actual temperatures were between 241 °C and 359 °C. Exact irradiations temperatures of specimens were derived from a combination of SiC thermometry placed around specimens in each capsule and finite element analysis of capsule heat flow; experimental irradiation temperatures can be found in Table 1 (Howard and Smith, 2018; Field et al., 2019; Campbell et al., 2016).

We studied six tensile specimens of CCNZ. Since thermometry and finite element analysis calculations revealed that the irradiation temperatures of each specimen were fairly evenly spaced at intervals of 10–14 °C, each specimen was considered a unique data point; arbitrarily grouping the points would obscure some of the temperature-dependent trends observed in this study. Post-irradiation experiments were performed primarily in the Low Activation Materials Development and Analysis Laboratory (LAMDA) at Oak Ridge National Laboratory. The examination of a limited number of ITER-grade CCZ samples was also performed in order to obtain reference comparison points. Electrical resistivity measurements were performed with a custom built system for performing four-point measurements (Echols et al., 2023). Resistivity was measured across the area of each tensile specimen gauge (5 mm × 1.2 mm, 0.75 mm thick). Five readings of voltage and current were taken for each specimen after 5 s and averaged; a measure of the voltage and current after a 30 s hold was also taken for each sample. Vickers hardness measurements were taken at room temperature with a Buehler Tukon 3100. Each indentation had an applied load of 500 gf and a dwell time of 15 s; the reported hardness for each specimen was the average of five indentations. Tensile testing was performed at room temperature with an MTS Insight Electromechanical Test Frame at a constant strain rate of 10⁻³ 1/s; standard deviations are not available for this data as only one specimen was available for each irradiation temperature. Experimentally measured resistivity, hardness, and tensile testing data points for each sample can be found in Table 1.

Microstructural characterization consisted of scanning electron microscopy (SEM) and atom probe tomography (APT). Samples were polished for SEM with colloidal silica down to a final polish of 0.5 µm. SEM imaging was performed with a TESCAN MIRA3 GMH microscope in backscatter electron detection (BSE) mode at 10 kV, and energy dispersive X-ray spectroscopy (EDS) was used to identify phases and estimate compositions in the Laves phase and Cu matrix. ImageJ was used to analyze microstructural feature sizes and thresholding of void area fractions. APT nanotips were prepared using a dual-beam FEI Versa 3D focused ion beam (FIB). A Cameca Instruments LEAP4000X HR tomograph was used to analyze APT tips in laser mode at the Center for Nanoscale Materials Science (CNMS) at ORNL. A detailed description of the APT tip preparation and characterization process can be found in Bahl et al. (2021). Although tensile, hardness, and resistivity data were obtained for all samples, SEM was limited to three samples at varying irradiation temperatures. APT analysis was limited to one specimen, the sample with T_ir = 318 °C, due to the time-intensive nature of radioactive sample preparation for both techniques.

Transmutation products were calculated using ORIGEN, part of the SCALE code system (Wieselquist and Lefebvre, 2023). One-group transition coefficients used by ORIGEN were calculated using a 252-group HFIR spectrum for the flux trap regions (Xoubi and Primm, 2004). This model has been previously validated against historic flux wire activation data to confirm its accuracy (Skutnik and Austin, 2023). Based on the sample positions relative to the reactor midplane within the flux trap region, the total effective flux for each sample as adjusted per the empirical fit functions was derived from historic flux wire activation data, assuming a nominal midplane scalar flux of 3.98 *10¹⁵ n/cm²⋅s (ibid). In general, flux corrections for the axial locations of the samples were less than 4% of the nominal flux. Depletion modeling assumed that the reactor was at full nominal (85 MW) power over the entirety of each irradiation interval, with inter-cycle outages explicitly accounted for.

Previous research on neutron irradiated copper and CCZ alloys has found that significant radiation hardening (mainly associated with formation of nanoscale stacking fault tetrahedra (SFT) and dislocation loops) occurs for irradiation temperatures up to ∼275 °C, whereas slight softening in CCZ has been reported for irradiation temperatures >275 °C due to reduced SFT and loop formation along with potential precipitate coarsening (Li and Zinkle, 2020; Barabash et al., 2011; Zinkle and Fabritsiev, 1994; Fenici et al., 1994; Fabritsiev et al., 2004; Fabritsiev et al., 1996). Irradiation generally causes an increase in electrical resistivity due to increased electron scattering from transmutant solutes and radiation-induced defect clusters (SFTs, loops, cavities) (Chaplin and Coltman, 1982; Li and Zinkle, 2020; Zinkle and Fabritsiev, 1994; Fabritsiev et al., 1996; Fabritsiev and Pokrovsky, 2005; Edwards et al., 2007; Frost and Kennedy, 1986; Singh et al., 1997; Edward et al., 1994; Fabritsiev et al., 1998; Fabritsiev and Pokrovsky, 1997). The electrical resistivity increase associated with the creation of nanoscale defect clusters (SFTs, loops) during the irradiation of Cu near room temperature saturates after ∼0.1 dpa at a value of ∼0.1 µΩ-cm and therefore only contributes in a minor way to the observed electrical resistivity changes for the current irradiation conditions (Chaplin and Coltman, 1982; Birtcher and Blewitt, 1981).

Multiple temperature-sensitive competing effects in the changing microstructure may operate during irradiation, such as ballistic mixing and dissolution of precipitates, solute diffusion, transmutation, and defect formation. These can contribute to changes in the strengthening contributions from precipitates, solute in solid solution, and defect formation. The detection of higher levels of Ni and Zr in the matrix at lower irradiation temperatures is likely due to the dominance of ballistic mixing at lower temperatures compared to thermal diffusion (Nelson et al., 1972; Zhao et al., 2022). The resulting solid solution strengthening from Ni and Zr at lower temperatures may partly explain why greater radiation hardening is observed at lower temperatures. Although changes to the volume fraction and particle size of the Laves phase were not observed, defect consolidation around the Laves phase and the observed embrittlement of the Laves phase affected its efficacy for strengthening. Likewise, the competing effect of solid solution strengthening in the Cu matrix may have improved the yield strength. The introduction of nanoscale SFT and dislocation loop defect clusters that can be sheared by gliding dislocations, also with embrittlement of the Laves phase and void formation at Laves/matrix interfaces, are likely the primary contributors to the observed drop in elongation. The greater levels of solute in the matrix at lower temperatures is also likely a contributing factor to the higher electrical resistivity of the CCNZ samples irradiated at lower temperatures, peaking at 4.47 μΩ·cm at 268 °C (an increase of 87%) compared to the 55% increase to 3.7 μΩ·cm observed at 294 °C.

The detailed local concentration and distribution data obtained from APT provides total levels of solute in solid solution in the Cu matrix for the sample irradiated at 318 °C. The effect of each of these solutes in solid solution on the conductivity of pure Cu is well-documented in the literature (Davis and ASM International, 2001; MaterialsAmerican Society for Testing, 1991; Simon et al., 1992). Table 4 summarizes the resistivity increase per 1wt% increase, the wt% of that solute in the irradiated alloy matrix (calculated by ORIGEN), and the corresponding calculated net resistivity increase caused by each solute in this system. These values assume a linear increase in resistivity with solute additions, which is only a valid assumption for small additions of each solute and when the distance between impurities is large (Žnidarič, 2022). The concentrations of each solute observed are near or below their individual solubility limits in Cu (MaterialsAmerican Society for Testing, 1991) and can reasonably be considered within this linear regime. With these values, the contributions from each solute can be added according to Matthiessen’s rule, which states that the total resistivity of a metal is the sum of contributions due to imperfections in the crystal as well as thermal vibrations of the lattice. Although the net contributions for each solute total 2.08 μΩ·cm are based on the composition of the Cu matrix (Table 4), the experimental change in resistivity, ∆ ρ exp , for CCNZ alloys only amounts to 1.55 μΩ·cm. The overestimation from individual solute contributions is likely due to a lack of published data on the mechanisms by which the non-irradiated sample’s resistivity is increased from pure Cu, and due to any non-linearity in solute contributions. However, it is still useful to use the known solute data for these alloys to assess the relative importance of different irradiation effects on the change in resistivity, such as transmutation, ballistic mixing, and radiation- or thermally induced segregation and partitioning.

The total change in the resistivity of each alloy, ∆ ρ , after irradiation can be broken down into several contributions: ∆ ρ = ∆ ρ t r + ∆ ρ b a l + ∆ ρ s s p + ∆ ρ def , (3) where ∆ ρ t r is the total change in resistivity due to transmutation products in solid solution in the matrix phase, ∆ ρ b a l is the resistivity change due to ballistic mixing of Zr and Cr precipitates back into the matrix (into solid solution), ∆ ρ s p is the change in resistivity due to solute partitioning (original solutes or transmutation products) to precipitates or nanoscale voids, and ∆ ρ def is the contribution from nanoscale defect clusters and dislocation loops. Although microstructural features such as grain boundaries can impact electrical resistivity, the similarity of the Laves phase size and morphology, as well as the grain sizes and grain size distribution, between irradiated and unirradiated CCNZ suggests that this contribution remains relatively constant during irradiation.

To quantify the effect of solute partitioning, maximum possible contributions of each solute are calculated in the case of zero partitioning to precipitates. Maximum contributions from ballistic mixing ( ∆ ρ b a l _ t o t a l ) will comprise the contributions from Cr and Zr in matrix, with the assumption that the non-irradiated matrix content of each is at the solubility limit for the aging temperature of CCNZ, 475 °C, which is 0.0wt% for Zr and 0.045wt% for Cr (Kang and Jung, 2010; Alexander, 1939). It also requires consideration of the maximum possible concentration of Zr in the matrix in the case of no Zr partitioning to Cr precipitates (Zr = 0.15 wt% in this case). Because Cr in the irradiated matrix is below this limit, the ballistic contributions from Cr solute are 0, while ∆ ρ b a l _ Z r = 1.2 µ Ω · cm . A maximum contribution from transmutation ( ∆ ρ t r _ t o t a l ) will assume that all transmutation products which originated in the Cu matrix have stayed in the matrix—that is, Ni and Zn are both present in equal amounts, up to 1.824wt%—resulting in contributions of ∆ ρ t r _ Z n = 0.52 µ Ω · cm and ∆ ρ t r _ N i = 2.19 µ Ω · cm and also including minor contributions from transmutant solutes Co, Fe, and Si. Contributions from solute partitioning ( ∆ ρ s p _ t o t a l ) of Ni and Zr, then, will be negative, as this considers the difference in the experimental matrix composition, and these maximum concentrations used in calculating ballistic and transmutation contributions. ∆ ρ s p _ N i = − 0.94 µ Ω · cm and ∆ ρ s p _ Z r = − 1.11 µ Ω · cm . Additionally, as the concentration of Cr in the matrix (0.034wt%) is lower than the aging temperature solubility, we can consider the depletion of Cr from the matrix through re-precipitation under the contribution ∆ ρ s p _ C r = − 0.05 µ Ω · cm . Under these assumptions, ∆ ρ t r _ t o t a l = 2.94   µ Ω · cm  ∆ ρ b a l _ t o t a l = 1.2   µ Ω · cm  ∆ ρ s p _ t o t a l = − 2.1   µ Ω · cm .

When these contributions are used to calculate [∆ρ _total ] in Equation 3, they sum approximately to the original predicted change from solutes (as each calculation introduced rounding error), ∆ ρ t o t a l = 2.04 µΩ·cm. For the moderate-dose (∼5 dpa) irradiation conditions being investigated, defect clusters make a relatively minor contribution to resistivity ( ∆ ρ def ∼0.1 μΩ·cm) compared with the solute-relevant effects (Chaplin and Coltman, 1982; Birtcher and Blewitt, 1981). Comparison of these contributions finds that the primary influence on the resistivity in these alloys will be transmutation, followed by solute partitioning to precipitates, ballistic mixing, and then defect clusters. These results highlight the importance of a high-volume density of Cr precipitates in the matrix, not only for strengthening but as a means of mitigating resistivity change under irradiation by providing preferential segregation sites for Ni and Zr solutes, depleting their matrix concentrations; this is more effective at higher temperatures when greater thermal diffusion rates will increase their mobility through the matrix.

Although the presence of Nb is prohibitive for fusion energy systems in terms of acceptable activity limits even in scant quantities (Fetter et al, 1988), the enhanced creep rupture life afforded by the addition of the Laves phase is an attractive design feature of the CCNZ alloys given the relatively similar post-irradiation tensile and electrical properties of CCNZ and CCZ. We suggest that redesign with a different Laves-forming solute could be a promising design strategy for the future development of commercially viable creep-resistant Cu alloys for nuclear fusion.

Cu–Cr–Nb–Zr (CCNZ) alloys designed to improve the creep resistance of ITER-grade Cu–Cr–Zr (CCZ) alloys were neutron irradiated to 5dpa at 250–325 °C. Post-irradiation experiments found radiation hardening to be more significant in CCNZ (36% increase at 250 °C) than CCZ (17% increase at 250 °C), but the trend of a decrease in hardening with increasing irradiation temperature was observed for both. The ultimate tensile strength of irradiated CCNZ was found to be similar to non-irradiated CCNZ and remained relatively stable with increasing T_ir—within 10% of non-irradiated values—but the yield strength, elongation, hardness, and electrical resistivity data all suggest the influence of competing factors such as solute partitioning and transmutation which have temperature-dependent influence. Evidence for embrittlement in the Cr₂Nb Laves phase and pore formation at the matrix–Laves interfaces were the most significant microstructure changes observed. Transmutation calculations found that CCNZ and CCZ alloys both produced ∼1wt% each of Ni and Zn through Cu transmutation, and these products were the most significant contributors to the increase in electrical resistivity. APT revealed substantial partitioning of Zr and Ni to the nanoscale Cr precipitates, reducing the resistivity contributions of each solute. These results suggest that the addition of a Laves phase is a promising design concept for improving the creep properties of CCZ alloys for fusion environments without severely compromising electrical and thermal conductivity.

This manuscript has been authored by UT-Battelle, LLC, under contract DE-AC05-00OR22725 with the US Department of Energy (DOE). The US government retains and the publisher, by accepting the article for publication, acknowledges that the US government retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. DOE will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (https://www.energy.gov/doe-public-access-plan).