Multi-principal element grain boundaries: Stabilizing nanocrystalline grains with thick amorphous complexions
1 Multi-principal element grain boundaries: Stabilizing nanocrystalline grains with thick amorphous complexions
Charlette M. Grigorian a , Timothy J. Rupert a, b, c, * a Department of Chemical and Biomolecular Engineering, University of California, Irvine, California 92697, USA b Department of Mechanical and Aerospace Engineering, University of California, Irvine, California 92697, USA c Department of Materials Science and Engineering, University of California, Irvine, California 92697, USA * Email address: [email protected]
Abstract
Amorphous complexions have recently been demonstrated to simultaneously enhance the ductility and stability of certain nanocrystalline alloys. In this study, three quinary alloys (Cu-Zr-Hf-Mo-Nb, Cu-Zr-Hf-Nb-Ti, and Cu-Zr-Hf-Mo-W) are studied to compare the influence of increased chemical complexity on thermal stability of the nanocrystalline microstructure, in addition to grain boundary structure. Significant boundary segregation is observed for Zr, Nb, and Ti in the Cu-Zr-Hf-Nb-Ti alloy, creating a quaternary interfacial composition which limits grain growth even after 1 week at ~97% of the melting temperature. This level of thermal stability is attributed to the high levels of chemical complexity at the grain boundary that is a consequence of the multi-component segregation. High resolution electron microscopy of grain boundary structure reveals that the complex grain boundary chemistry in the Cu-Zr-Hf-Nb-Ti alloy is associated with a 44% and 32% increase in the average amorphous complexion thicknesses found in previously studied Cu-Zr and Cu-Zr-Hf alloys. Keywords: nanostructure, metal, grain boundaries 2
Introduction
The recent discovery of grain boundary complexions allows for a deeper understanding of previously unexplained interfacial phenomena in materials. Grain boundary complexions can be defined as phase-like grain boundary states which can only exist in equilibrium with their abutting phases [1, 2]. Six distinct complexion types in doped alumina samples were classified by Dillon and Harmer, listed here in order of increasing thickness and structural disorder: (I) sub-monolayer segregation, (II) clean grain boundaries, (III) bilayer segregation, (IV) multilayer segregation, (V) nanoscale intergranular films, and (VI) wetting films [1]. Classification schemes for grain boundary complexions are not limited to such thickness-based descriptions either; several other classification schemes include complexion composition or geometry [2, 3]. Complexions may be described as intrinsic or extrinsic when they are found in pure systems or a system with dopants/impurities, respectively. Extrinsic complexions can be further described by whether the atoms are arranged in an ordered, periodic fashion within the phase, or if they are chemically disordered. Local structure can also be used to characterize different complexions, as even a boundary in a pure system can be either ordered or disordered (i.e., amorphous). The type of complexion which exists at a particular grain boundary is dependent on a variety of factors, including temperature, pressure, chemical composition, and boundary misorientation, which can impact the structure and energy of the grain boundary [4, 5]. Finally, transitions between different complexion types can occur analogously to bulk phase transitions as a thermodynamic parameter is varied [5-7]. Absent long range structural order, amorphous complexions are of particular interest in the enhancing the thermal stability and mechanical properties of nanocrystalline metals. Nanocrystalline microstructures exhibit high strength compared to bulk metals, but are notorious 3 for microstructural instability [8] and poor ductility [9, 10]. Unique thermal stability behavior attributable to amorphous complexions has been reported in Ni-W alloys, which experience significant grain growth at intermediate temperatures, but demonstrate exceptional thermal stability when annealed above 1000 ° C, once amorphous complexions form at the grain boundaries [11]. A similar stability was recently demonstrated in a Ni-Mo alloy, where increasing the average amorphous complexions thickness helped to significantly hinder coarsening and inhibit abnormal grain growth compared to the same alloy with thinner amorphous complexions [12]. Concerning mechanical properties, Zhao et al. showed that the formation of crystalline/amorphous Cu/Cu-Zr interfaces upon increasing the Zr concentration of sputtered Cu-Zr films led to significantly increased alloy strength [13]. Amorphous complexions are also known to enable efficient absorption of dislocations, with especially thick films slowing the process of crack nucleation and crack growth as the material is deformed [14]. This behavior enhances the ductility and toughness in nanocrystalline metals, as demonstrated in Cu-Zr alloys by Khalajhedayati et al. [15]. Structurally disordered complexions have also been connected to other behaviors which affect materials processing and degradation. Activated sintering behavior has been observed in a variety of materials containing amorphous complexions [16-20]. Liquid-like intergranular films were found in TiO doped with CuO, with the improved mass transport in the amorphous complexion believed to be the cause of activated sintering in this system [21]. Similar behavior was found by Donaldson and Rupert [22], who showed that amorphous complexions assisted in the densification of a bulk nanocrystalline Cu-Zr alloy sample from its constituent ball milled powder form. Such findings have significant implications for the creation of bulk nanostructured materials and also promise simultaneous unique properties imparted by the amorphous grain boundary phase. Last but not least, prior computational and experimental studies of Cu-Zr alloys 4 containing amorphous complexions showed that these features were more resistant to radiation damage than their counterparts with ordered grain boundaries [23, 24]. While the majority of studies characterizing amorphous complexions in metallic systems have focused on their formation in relatively binary systems, ternary and multicomponent systems of broad applicability in engineering applications are of particular interest for their ability to sustain different complexion states. Preliminary work suggests that increasing the chemical complexity of the grain boundary region may alter complexion thickness and stability. For example, Hf additions to a Cu-Zr alloy were found to increase the average thickness of amorphous complexions by ~9% [25]. Amorphous complexions in the Cu-Zr-Hf alloy were also observed to be more stable than those in Cu-Zr against transformation back to an ordered grain boundary during cooling from the high temperature needed for boundary premelting, with the critical cooling rate necessary to retain amorphous complexions reduced by roughly three orders of magnitude [26]. A more complex grain boundary chemistry also has the potential to improve important properties. Zhou et al. [27] showed that nanocrystalline alloys with complex, “high-entropy” compositions are very thermally stable; a grain size of 32 nm was measured in a Ni-Mo-Ti-Nb-Ta alloy after annealing at 900 ° C for 5 hours, while the binary Ni-Zr alloy used for comparison experienced significant grain growth and had an average grain size of 189 nm after the same heat treatment. The increase in thermal stability was attributed to the complex alloy chemistry that the authors hypothesized reduced grain boundary energy at high temperatures as compared to their binary Ni-based counterparts. However, it is important to note that the chemical composition and structure of the grain boundary complexions in this multicomponent alloy were not directly characterized [27], so these conclusions were based on indirect evidence. Mechanical properties 5 of materials also stand to benefit from chemically complex interfacial phases. Wu et al. reported that the presence of nano-sized metallic glass shells surrounding grains in a nanocrystalline Al-Ni-Y alloy significantly improves the strength of the material, while also allowing for the achievement of plastic strains up to 76% in a microcompression experiment [28]. These authors also demonstrated the benefits that the inclusion of nanoscale amorphous metallic glass phases surrounding the grains of a multi-principal element alloy, demonstrating the achievement of near theoretical yield strength in a Cu-Co-Ni alloy surrounded by amorphous Fe-Si-B interfaces [29]. Similarly, Yang et al. recently developed a Ni Co Fe Al Ti B superlattice alloy with nanoscale disordered interfaces rich in Fe, Co, and B that exhibited high thermal stability, strength, and ductility [30]. Adding increasing amounts of B to a base Ni-Co-Fe-Al-Ti alloy was found to increase the enrichment of Fe and Co at the grain boundaries, allowing for the formation of the disordered interfacial film. As a whole, these studies highlight the potential for improvement of properties in multicomponent alloys if complex grain boundary chemistries and unique boundary complexions achieved. Grain boundary complexion diagrams have been developed to predict interfacial structure and composition in binary and ternary alloys, but compositionally complex alloys are not as well characterized [31]. Of particular interest is the influence of chemical complexity on the formation and stability of amorphous complexions. In this study, the dopant distribution, interfacial structure, and thermal stability of quinary alloys are investigated in materials designed to achieve multi-principal element grain boundary compositions. Cu-Zr-Hf alloys were doped with 2 at.% each of Mo, W, Nb, and Ti in various combinations in order to observe the contribution of increased compositional complexity on thermal stability, with a focus on whether each element is able to segregate out of the Cu matrix and whether they stabilize the nanocrystalline grain structure . The Cu-Zr-Hf-Nb-Ti exhibits the 6 greatest thermal stability of the three alloys, due to co-segregation of Nb, Ti, and Zr to the grain boundaries and the subsequent formation of thick amorphous complexions with multi-principal element compositions. The presence of multiple dopant elements was found to have significant effects on previously characterized segregation behavior of Hf, with instead this element observed to be leached away into carbide particles, suggesting that the other alloying elements are preferred segregants. The average thickness of amorphous complexions found in the Cu-Zr-Hf-Nb-Ti alloy are roughly 44% thicker than those measured in a binary Cu-Zr alloy and 32% thicker than those measured in a ternary Cu-Zr-Hf alloy, with a clear trend that higher chemical complexity leads to thicker amorphous complexions for the same processing conditions. The results of this study demonstrate the capacity of thick amorphous complexions formed in multicomponent alloys to impart significant thermal stability to nanocrystalline microstructures, with fine grain sizes maintained even after 1 week at 97% of the alloy melting temperature. Results and Discussion
First, the as-milled structure of each quinary alloy, as well as their evolution during annealing, were investigated. X-ray diffraction (XRD) patterns of each of the three alloys after ball milling for 10 h, as well as after annealing for 5 h at 500 ° C and then followed by annealing treatments at 950 ° C for 5 min, 1 h, and 1 week, are shown in Figure 1. During mechanical alloying, HfC forms in all three alloys investigated in this study. The formation of carbide phases in the alloy during ball milling results from the addition of stearic acid as a process control agent, which is necessary to resist cold welding of the particles that prevents mechanical mixing [32]. In addition, in the patterns of the as-milled Cu-Zr-Hf-Mo-W (Figure 1(a)) and Cu-Zr-Hf-Mo-Nb (Figure 1(b)) powders, peaks for W and Mo phases are also seen, respectively, indicating that these 7 alloying elements did not completely mix to form a solid solution with Cu during the ball milling process. In the Cu-Zr-Hf-Nb-Ti sample, a very small amount of NbZr intermetallic phase (~1 vol.%) is seen in the as-milled powder. In order to attempt to force all dopant atoms into the Cu matrix to form a homogeneous solid solution, the three alloys were mechanically alloyed for much longer times of 20 h and 40 h (2 and 4 times longer than the original treatment, respectively). However, additional milling time did not lead to more complete mixing of these samples and similar secondary phases were observed, so we focus on the original alloys that were milled for 10 h for the rest of this study. The second phase volume percentages and XRD grain sizes of the second phases for the three alloys as a function of annealing time at 950 ° C are presented in Figure 2. During heat treatment, the volume percent of unmixed W in the Cu-Zr-Hf-Mo-W is greatly reduced after 5 min at 950 ° C and goes to zero after 1 h at 950 ° C, while Mo C precipitates form during annealing. In the Cu-Zr-Hf-Mo-Nb alloy, the peaks for unincorporated Mo disappear while Mo C peaks appear, in addition to the formation of a new NbC phase. After annealing, the NbZr phase found in the as-milled Cu-Zr-Hf-Nb-Ti microstructure goes away, but a small amount of NbC appears. The amount of the NbC phase present stays constant at 3.5 vol.%, with a slight increase in grain size as the material is annealed longer. After annealing for 1 week, a small amount of Ti-rich hexagonal close packed phase also begins to precipitate in the Cu-Zr-Hf-Nb-Ti alloy. The XRD grain sizes of the face centered cubic Cu-rich phase (the majority or matrix phase) are plotted as a function of annealing time at 950 ° C in Figure 3(a). After 5 min of annealing at 950 ° C, the grain sizes of all three alloys are very similar (~40 nm), within error for XRD measurements. However, after only 1 h of annealing, the difference in grain sizes highlight the relative thermal stability of the alloys; the Cu-Zr-Hf-Nb-Ti alloy has an average grain size of 43 8 nm, while the Cu-Zr-Hf-Mo-W and Cu-Zr-Hf-Mo-Nb alloys both coarsen noticeably to average grain sizes of 95 and 71 nm, respectively. The high thermal stability of the Cu-Zr-Hf-Nb-Ti alloy is further demonstrated by the relatively small grain size of 63 nm even after 1 week (168 h) of annealing, while the average grain sizes of the Cu-Zr-Hf-Mo-W and Cu-Zr-Hf-Mo-Nb alloys continue to grow to 98 nm and 102 nm, respectively. It is important to note that these last two quoted grain sizes are approaching the resolution limit of roughly 100 nm for XRD measurements of crystallite size, so it is possible that these samples no longer have a nanocrystalline microstructure after 1 week of annealing. The decreased thermal stability in the Cu-Zr-Hf-Mo-W and Cu-Zr-Hf-Mo-Nb may be attributed to the early precipitation of second phases during annealing. Figure 2(a) shows that the overall volume of second phases throughout the microstructure of the Cu-Zr-Hf-Mo-Nb alloy increases from roughly 5.4% in the as-milled state to 8.9% after only 1 h of annealing. For the Cu-Zr-Hf-Mo-W alloy, the rate of increase of the total second phase volume percent is even greater, growing from 6% in the as-milled state to 10.8% after 1 h of annealing, with this value staying relatively constant with further annealing time. With the precipitation and growth of the second phases which are rich in the alloying elements, there will be much fewer dopant atoms available to segregate to grain boundaries in the Cu-Zr-Hf-Mo-W and Cu-Zr-Hf-Mo-Nb alloys, potentially reducing the stability of the microstructure and allowing for grain growth. Additionally, the average sizes of the second phase particles within the Mo-W alloy are too large to maintain kinetic stabilization of a nanocrystalline microstructure via Zener pinning of the grains [33, 34]. The Cu-Zr-Hf-Mo-Nb coarsens more gradually than the Cu-Zr-Hf-Mo-W alloy, which is likely related to the greater availability of grain boundary segregating species. In contrast, the Cu-Zr-Hf-Nb-Ti alloy maintains the smallest total second phase volume (< 7.9%) even after one week of annealing at 950 ° C, and is therefore able to retain a grain size of 9 63 nm. The heat treatment temperature used here is estimated to be above 97% of the melting temperature for this alloy, based on the average of the melting temperatures of the binary alloy counterparts (Cu-Zr, Cu-Hf, Cu-Nb, and Cu-Ti), underscoring its exceptional thermal stability. The correlation of grain size stability with the amount of dopant content available to segregate to the boundaries (or, conversely, the inverse correlation between grain size stability and the second phase volume percent) amongst the three alloys suggests that boundary segregation is the dominant stabilizing mechanisms in these materials. In contrast, if kinetic stabilization through Zener pinner were dominant, the specimens with the most second phase particles would be stabilized the most, which is not consistent with the data shown here. Due to its superior thermal stability, the Cu-Zr-Hf-Nb-Ti sample was chosen for further investigation via transmission electron microscopy (TEM), in order to study the distribution of each dopant element within the microstructure. The grain boundary structure of this alloy is also investigated in detail to determine whether the addition of multiple dopant elements is conducive to the formation of thick amorphous complexions. The sample which was annealed for 5 min at 950 °C was chosen for this investigation, as denoted by the red arrow in Figure 3(a). A selected area electron diffraction (SAED) pattern and bright field TEM image of the sample region from which the pattern was taken are shown in Figures 3(b) and 3(c), respectively. The SAED pattern confirms the presence of the NbC and HfC phases previously detected with XRD, while the TEM image confirms the nanocrystalline grain size. Figure 3(d) presents the grain size distribution measured in this sample from which an average grain size of 38 nm was measured, which is in agreement with the previously measured XRD average grain size of 43 nm (XRD measurements typically are assumed to have up to 25% error). This sample was also viewed in high-angle annular dark field scanning transmission electron microscopy (HAADF STEM) conditions, where bright 10 contrast at grain boundaries indicate regions of segregation of higher Z atoms, such as Nb, Zr, and/or Hf, denoted by the yellow arrows in Figure 4. As seen in Figure 4, the majority of grain boundaries appear to be decorated with higher Z dopant elements, as evidenced by the bright contrast. Energy dispersive X-ray spectroscopy (EDS) line scans were performed across these regions of interest to confirm the presence of dopant segregation. The yellow lines in Figures 5(a) and 5(c) indicate the locations where the line scans shown in Figures 5(b) and (d) were taken, respectively. Both EDS scans show heavy segregation to the boundaries, but the dopants which are present at the grain boundary are not the same. In Figure 5(b), the segregation of Nb, Ti, and Zr is observed to the grain boundary, which combined with the base Cu shows that four elements are present for a quaternary composition in the grain boundary region. In contrast, the line scan in Figure 5(d) shows that only Nb and Ti dopants are present, with no obvious Zr segregation found, giving a ternary composition when combined with the base Cu. In both line scans shown in Figure 5, Hf was not observed in either the grain boundary or crystalline regions. These observations are representative examples of over 50 such measurements of the composition of grain boundaries within this sample. While Zr was sometimes observed at certain grain boundaries and found to be depleted at others, co-segregation of Nb and Ti atoms was consistently found across all boundaries characterized. However, segregation of Hf was not found at any grain boundaries investigated in this study. We hypothesize that the lack of Hf at the boundaries can be attributed to the formation of HfC particles, which serves to leach the potential dopant from the matrix phase. We note that there is a reduction in the volume percent of HfC in the sample after annealing for 1 week, as shown in Figure 2(c). Future work should explore adjustments to the processing route or annealing treatments, in order to prevent carbide formation and enhance Hf segregation to grain boundaries. Generally, the variation in grain boundary chemistries observed here is reminiscent of that reported 11 in our prior study of ternary Cu-rich alloys, where co-segregation of the two dopants occurred at times while other boundaries were only doped with one of the dopant species [25]. For this study, we emphasize that the grain boundaries could have up to four elements present, with the important caveat that boundary-to-boundary variations do occur. High resolution TEM micrographs of amorphous complexions in the Cu-Zr-Hf-Nb-Ti sample are shown in Figure 6. Yellow dashed lines are included to highlight where the amorphous complexion meets the crystalline region on either side, which could be denoted as an amorphous-crystalline interface (see, e.g., [35]). While amorphous complexions were found at many grain boundaries throughout the microstructure of this sample, it is important to note that these features do not exist at every single grain boundary, as some boundaries remained with a typical ordered structure. Figures 7(a) and 7(b) show the population of amorphous complexion thicknesses measured from 46 amorphous complexions in cumulative distribution function and histogram form, respectively. All amorphous complexions in this study were observed and measured in an edge-on viewing condition, to enable accurate measurements of complexion thickness. In addition, thickness measurements were taken from the thinnest observable region of the complexion to ensure consistency. The mean thickness of amorphous complexions in this sample was 2.44 nm, with a standard deviation of 1.36 nm. In order to compare the thicknesses of amorphous complexions in the Cu-Zr-Hf-Nb-Ti sample with multi-principal element grain boundary compositions with what could be considered baseline binary (Cu-Zr) and ternary (Cu-Zr-Hf) alloys [25], the complexion thickness distributions for these samples are plotted in Figure 7(a). Increasing the chemical complexity of the grain boundary region is found to result in significant increases to the thickness of the amorphous complexions in the material, as going from two to three and then to four elements at the grain boundary results in mean amorphous complexion 12 thicknesses of 1.70 nm, 1.85 nm, and 2.44 nm, respectively. Framed as direct comparisons, the mean complexion thickness in the Cu-Zr-Hf-Nb-Ti alloy is 44% greater than that found in the binary alloy, and 32% greater than that in the ternary alloy. While the range of complexion thicknesses measured in the ternary and quinary samples are the same, similar minimum and maximum values, there are a larger number of very thick amorphous complexions in the quinary alloy. For example, roughly 30% of the amorphous complexions measured in the Cu-Zr-Hf-Nb-Ti sample have a thickness greater than 3 nm, compared to only ~12% of those same types of thick amorphous complexion having been measured in the binary and ternary alloys. This can be attributed to the addition of multiple dopant elements which segregate to the grain boundaries in the quinary alloy. We note that while the global alloy compositions are not exactly the same, with the binary and ternary alloys from Ref. [25] targeting 5 at.% dopant and the quinary alloys studied here targeting 8 at.% dopant, these values are actually quite close once one considered that the Hf was never found to segregate to the grain boundaries in the Cu-Zr-Hf-Nb-Ti alloy. Removing the Hf from consideration, roughly 6 at.% dopant was available to segregate to the grain boundary and create amorphous complexions in the quinary alloy. It is important to consider the effects associated with having multiple dopant elements present at the grain boundaries in order to stabilize the amorphous grain boundary phase, as well as the influence of second phase formation on the availability of potential segregating dopants. For example, a prior study demonstrated amorphous complexion formation in a Cu-4 at.% Hf alloy, but increasing the dopant concentration to 11 at.% Hf actually proved to be detrimental to the thermal stability of the alloy. For the higher dopant concentration, the formation of a large volume of HfC particles throughout the microstructure was observed and hypothesized to have decreased the amount of dopant available to segregate to grain boundaries, inhibiting the formation 13 of amorphous complexions and therefore rampant grain growth was observed for this alloy. A similar phenomenon can explain the lack of Hf observed at the grain boundaries of the Cu-Zr-Hf-Nb-Ti alloy in the present study. The HfC that forms during the mechanical alloying process actually serves to rob the grain boundaries of Hf as a potential dopant species. This type of behavior was observed in nanocrystalline W-Cr alloys by Donaldson et al., where the precipitation of a Cr-rich phase reduced the amount of Cr present at grain boundaries, leading to an unstable grain structure in the alloy [36]. In contrast, a Cu-10 at.% Zr alloy investigated in a prior study proved to be more stable against grain growth than a Cu-5 at.% Zr alloy, implying that Zr atoms may either have a stronger tendency to segregate to grain boundaries to facilitate amorphous complexion formation or a weaker tendency to form carbide particles, as these two microstructural features can be thought of as competing for the available dopants. As a whole, this discussion highlights the fact that the formation of thick amorphous complexions is not always as simple as increasing the concentration of dopant atoms in the alloy. The addition of multiple dopant elements to an alloy creates even more complications for grain boundary segregation, where attractive or repulsive forces between dopant atoms themselves may encourage co-segregation of particular elements or cause the depletion of others due to site competition effects. Prior studies by Xing et al. have investigated the interactions between dopants and their influence on grain boundary segregation in ternary alloys [37, 38]. It was found that induced co-segregation of dopant atoms can occur, even when a particular dopant element is predicted to desegregate from grain boundaries when by itself in the binary counterpart, as was the case in the Pt-Au-Pd alloy that was investigated in detail in that work [38]. The authors were able to develop a model to predict the segregation behavior of dopant elements in a ternary alloy, with a focus on identifying whether there would be dopant depletion, site competition, or induced co- 14 segregation of the dopants to grain boundaries. Such complex interplay between different dopant species has been observed in a variety of alloys, where the addition of one alloying has been proven to either inhibit or enhance the segregation behavior of the other added elements, with the mechanical behavior of steels being a notable example. For example, Mo and P co-segregate to grain boundaries in steels, where the concentration of P at grain boundaries increases with that of Mo [39]. In contrast, the addition of C to Fe-P alloys decreases the amount of P segregating to grain boundaries, and in fact can suppress the grain boundary embrittlement caused by P segregation [40-42]. As an example of even more complicated dopant interplay, Cr addition increases the grain boundary segregation of P in Fe-C-P alloys, but has no effect in Fe-P alloys [40]. Segregation competition may provide an explanation as to why no segregation of Hf was observed in the Cu-Zr-Hf-Nb-Ti sample, while it was previously observed in binary Cu-Hf and ternary Cu-Zr-Hf alloys. In the binary and ternary alloys, the Hf found energetically favorable grain boundary sites, while in the quinary alloy, the grain boundary segregation sites preferred to be populated by the Zr, Nb, and Ti atoms. This may also explain the relative second phase evolution observed in the Cu-Zr-Hf-Mo-W and Cu-Zr-Hf-Mo-Nb samples investigated in this study, with examination of the volume percentages of second phases containing Mo being of importance. During annealing, the maximum amount of Mo C which forms in the Cu-Zr-Hf-Mo-Nb alloy is only 3 vol.%, while a much greater maximum 8.4 vol.% forms in the Cu-Zr-Hf-Mo-W alloy, implying that there may be either a repulsive interaction between Mo and W atoms or an attractive interaction between Mo and Nb (or both effects simultaneously), with the net effect being that Mo atoms prefer to leave the boundaries and form carbides more in one of the quinary systems. Co-segregation of dopants may also have unexpected effects on complexion formation. Materials selection rules for the formation of amorphous complexions in binary alloys suggest 15 choosing dopant elements which have a positive ∆ H seg , a negative ∆ H mix , and an atomic size mismatch of greater than 12% [43], fundamentally suggesting that (1) pronounced grain boundary segregation and (2) a good glass forming composition are needed. By these standards, one would be expected that amorphous complexions would not form in a binary Cu-Nb alloy, due to the positive ∆ H mix value calculated, indicating that it would be preferable for the precipitation of a Nb-rich phase to occur, a prediction which is supported by experimental observations [44, 45]. In fact, Pan and Rupert [46] observed such precipitation of a body centered cubic, Nb-rich phase at the interfaces during atomistic simulations of grain boundary segregation in Cu-Nb. However, co-segregation of Nb to the grain boundaries led to thicker amorphous complexions in the Cu-Zr-Hf-Nb-Ti alloy, while no other grain boundary phases were observed. While Nb is not useful for forming amorphous complexions by itself, increasing chemical complexity in the interfacial region can increase amorphous complexion thickness when used in combination with a grain boundary dopant such as Zr which is very efficient at driving grain boundary premelting. In the bulk metallic glass literature, materials design guidelines, such as those from Inoue and coworkers [47, 48], are utilized in order to maximize their glass forming ability, with a particularly well known rule proposing the use of three or more elements in order to frustrate the crystallization of the amorphous phase. Glass forming ability is typically characterized by the critical cooling rate required to retain the amorphous structure for a given sample size or the critical sample thickness at a given cooling rate, with either essentially describing the resistance of the glassy phase to transformation back to an ordered phase [49-51]. Several studies have demonstrated enhanced glass forming ability associated with the addition of multiple alloying elements in bulk metallic glasses [50, 52-56]. Analogously, the partial substitution of shift from a binary to a ternary composition in nanocrystalline alloys was recently shown to reduce the critical cooling rate 16 necessary to retain amorphous complexions during quenching [26]. Viewed from this perspective, the increased thickness of the complexions in the Cu-Zr-Hf-Nb-Ti alloy here represents a greater resistance to transformation back to ordered grain boundary complexions, due to the increased chemical and structural disorder induced by the presence of multiple dopant elements. These findings have profound implications for the development of bulk nanocrystalline alloys containing amorphous complexions, as a reduction in the cooling rate necessary to retain amorphous complexions would simplify the formation of these materials, enabling their utilization in a variety of appropriate lower temperature applications. Conclusions
In this study, three quinary nanocrystalline alloys were formed via mechanical alloying, then subsequently annealed and characterized in order to probe the thermal stability and potential for amorphous complexion formation. Various dopant elements (Mo, W, Nb, and Ti) were added to a Cu-Zr-Hf base material in different combinations but with a consistent total doping level. The following conclusions can be drawn: • Second phase formation was observed in the three alloys investigated in this study, primarily in the form of carbide phases but also with some secondary metallic phases appearing. The grain size stability of the bulk face centered cubic phase observed in each alloy was inversely correlated with the second phase volume percent, suggesting that precipitation “robbed” the grain boundaries of potential dopant species so there were fewer atoms available to segregate and reduce excess grain boundary energy. Kinetic stabilization of the microstructure by Zener pinning was ruled out as a potential dominant effect in this study. 17 • The Cu-Zr-Hf-Nb-Ti alloy had the greatest thermal stability, retaining a grain size of 63 nm after a week of annealing at 950 ° C, more than 97% of the melting temperature of the alloy. The superior thermal stability of this alloy is attributed to the segregation of Zr, Nb, and Ti atoms to grain boundaries, enabling the formation of thick amorphous complexions with chemically complex compositions of up to four elements. As a result, the interfacial complexions in this alloy can be termed multi-principal element complexions . The amorphous complexions measured in this study were on average 44% thicker than those in a previously studied binary Cu-Zr alloy, and 32% thicker than those in a previously studied ternary Cu-Zr-Hf alloy. The increased complexity of the grain boundary composition is expected to enhance the stability of the amorphous complexion, allowing for the retention of thicker complexions upon quenching from high temperatures. • The addition of multiple dopant elements significantly affects the segregation behavior. Hf atoms, which were previously found to segregate to grain boundaries in Cu-Hf and Cu-Zr-Hf alloys, did not segregate in the Cu-Zr-Hf-Nb-Ti alloy investigated in the present study. Instead, the Hf atoms were accommodated into the carbide second phases. The complex balance between the attractive and repulsive interactions between different elements leading to co-segregation or depletion of particular elements at grain boundaries, and second phase formation consuming the amount of dopant available to form amorphous complexions, remains a fruitful area for future research. The results of this work demonstrate the improvement of amorphous intergranular film thickness through the utilization of multi-principal element grain boundary compositions. This work also highlights the complications associated with grain boundary segregation and amorphous 18 complexion formation in multicomponent alloys, due to the variety of interactions that can occur between different dopant elements. This work motivates experimental and computational studies to further the understanding of alloy selection for the formation of amorphous complexions beyond simple alloy chemistries, as well as characterizing the arrangement of the various dopant atoms within the amorphous complexions themselves.
Materials and Methods
Alloy compositions were selected in order to maximize dopant segregation to grain boundaries, creating a high-entropy grain boundary composition. In contrast to high-entropy alloys, which are composed of multiple principal elements in high proportions, the alloys in this study are doped with multiple elements in small amounts. These alloys are then heat treated to induce segregation of these dopant elements to the grain boundaries, creating high-entropy grain boundary complexions with high proportions of these dopant elements when only the grain boundary concentration is considered. Schuler and Rupert [43] have outlined specific alloy selection criteria for binary alloys capable of forming amorphous complexions at grain boundaries, including a positive enthalpy of segregation, a negative enthalpy of mixing, and a large atomic size mismatch with the bulk phase. The formation of thick amorphous complexions has been previously demonstrated in a ternary Cu-Zr-Hf alloy, due to the co-segregation of Zr and Hf dopants to grain boundaries [25]. It is anticipated that the addition of multiple dopants which are able to segregate to grain boundaries will lead to the formation of even thicker and more stable amorphous complexions, by further frustrating the crystallization of the complexion upon cooling. This is a commonly referenced guideline concerning the stability of bulk metallic glasses (BMGs), as many are composed of quaternary or quinary compositions in order to maximize their glass- 19 forming ability [57-59]. In a study of Mg-based BMGs, Yin et al. reported maximum casting diameters of ternary Mg-Ni-Gd and Mg-Ni-Nd to be 2 mm, while BMGs with quaternary compositions (Mg-Ni-Gd-Nd) had a significantly greater glass-forming ability and were able to be cast to diameters up to 5 mm [56]. In order to investigate the effects of adding multiple dopant elements on the formation of amorphous complexions with high-entropy compositions, Cu doped with 2 at.% Zr and 2 at.% Hf was used as a base alloy. Two additional dopant elements were added to the base Cu-2Zr-2Hf alloy in the amount of ~2 at.% each to investigate how different dopants influence grain boundary segregation and thermal stability. Mo, Nb, Ti, and W were chosen as the additional dopant elements, due to the large atomic size mismatches with Cu of 39%, 38%, 28%, and 40%, respectively. In addition, the segregation of Mo [43], Nb [46, 60], Ti [61], and W [62] to grain boundaries in Cu-based alloys has been previously demonstrated.
Cu-Zr-Hf-Mo-Nb, Cu-Zr-Hf-Nb-Ti, and Cu-Zr-Hf-Mo-W alloys were formed via mechanical alloying in a SPEX SamplePrep 8000M Mixer/Mill. Cu (Alfa Aesar, 99.99%, -170 + 440 mesh), Zr (Micron Metals, 99.7%, -50 mesh), Hf (Alfa Aesar, 99.8%, -100 mesh), Mo (Alfa Aesar, 99.95%, -100 mesh), Nb (Alfa Aesar, 99.8%, -325 mesh), Ti (Alfa Aesar, 99.5%, -325 mesh), and W (Alfa Aesar, 99.9%, -100 mesh) powders were alloyed in an Ar atmosphere with a hardened steel mixing vial and milling media. A 10:1 ball-to-powder mass ratio was used, and 1 wt.% of stearic acid was added as a process control agent to prevent excessive cold welding. Prior to milling each of the desired alloys, pure Cu powder and 1 wt.% stearic acid was milled for 2 h to coat the vial walls and milling media with a thin layer of Cu and then subsequently disposed of, limiting the contamination of the alloy with Fe from collisions with the hardened steel during milling. Confirmation of the chemical composition of the powders was performed using EDS in 20 a Tescan GAIA3 scanning electron microscope (SEM). Measured compositions for each sample from EDS are shown in Table 1. While minor deviations of the measured compositions and the target values are found, these are within the ~1 at.% error expected of EDS measurements. Alloy powders were sealed inside of quartz tubes under vacuum prior to annealing, in order to limit exposure to atmosphere which could cause the formation of additional phases such as oxides during heat treatment. The powders were first annealed at 500 ° C for 5 h in order to encourage the segregation of the various dopant elements to the grain boundaries. This was followed by an additional annealing treatment at 950 ° C for 5 min (0.083 h), 1 h, and 1 week (168 h) to allow for the formation of amorphous complexions. We only quote the second annealing step when describing the powder microstructures in subsequent text, but remind the reader that all samples were first exposed to the lower temperature annealing treatment. The high temperature annealing step should provide the correct conditions for grain boundary premelting while also allowing for the investigation of the thermal stability of the microstructure against coarsening. The powder samples were quenched at the end of the anneal in order to freeze in the equilibrium grain boundary structure at high temperatures, including any amorphous complexions that were able to form. The as-milled and annealed powders were then characterized using a Rigaku SmartLab X-ray diffractometer with Cu K α radiation and a 1D D/teX Ultra 250 detector. Phase identification, grain size measurements, and volume percent determination of all phases were extrapolated from XRD patterns via Rietveld analysis using MAUD software [63]. The mean grain sizes of the Cu-rich, face centered cubic phase in all as-milled alloy powders were in the range of 19-32 nm. An electron transparent lamella of the Cu-Zr-Hf-Nb-Ti sample that was annealed at 950 ° C for 5 min was created for inspection with TEM via the focused ion beam (FIB) lift-out method using a Tescan GAIA SEM/FIB. Excess Ga + ion beam damage of the lamella caused during 21 sample preparation was removed with a final 5 kV polish. Detailed characterization of the microstructure of this sample was performed in the TEM, including bright field TEM, SAED, high resolution TEM of the grain boundary structure, HAADF STEM, and EDS were performed using a JEOL JEM-2800 microscope equipped with dual EDS detectors. The presence of second phases extracted from the XRD pattern was confirmed with SAED. The average grain size obtained from the XRD pattern was also confirmed by measuring >160 grains imaged using bright field TEM. The distribution of alloying elements throughout the sample microstructure was observed using EDS. Line scans allowed for the observation of dopant segregation or depletion at grain boundaries where amorphous complexions are expected to form. The influence of alloy chemistry on the thickness of amorphous complexions is investigated by measuring the thickness of 46 amorphous complexions imaged with high resolution TEM in an edge-on condition. The edge-on condition of each grain boundary was confirmed by verifying that no change in amorphous complexion thickness is observed when viewing the boundary in under- and over-focused conditions. Acknowledgements
This study was supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Materials Science and Engineering Division under Award No. DE-SC0021224. SEM, FIB, TEM, and XRD work was performed at the UC Irvine Materials Research Institute (IMRI) using instrumentation funded in part by the National Science Foundation Center for Chemistry at the Space-Time Limit (CHE-0802913).
Conflict of Interest Statement
22 On behalf of all authors, the corresponding author states that there is no conflict of interest.
Data Availability
Data will be made available by the corresponding author, upon reasonable request. References
1. S.J. Dillon, M. Tang, W.C. Carter and M.P. Harmer: Complexion: A new concept for kinetic engineering in materials science.
Acta Mater. , 6208-6218 (2007). 2. P.R. Cantwell, M. Tang, S.J. Dillon, J. Luo, G.S. Rohrer and M.P. Harmer: Grain boundary complexions. Acta Mater. , 1-48 (2014). 3. T.J. Rupert: The role of complexions in metallic nano-grain stability and deformation. Curr. Opin. Solid State Mater. Sci. , 257-267 (2016). 4. S.J. Dillon and M.P. Harmer: Demystifying the role of sintering additives with “complexion”. J. Eur. Ceram. Soc. , 1485-1493 (2008). 5. P.R. Cantwell, T. Frolov, T.J. Rupert, A.R. Krause, C.J. Marvel, G.S. Rohrer, J.M. Rickman and M.P. Harmer: Grain Boundary Complexion Transitions. Annu. Rev. Mater. Res. , 465-492 (2020). 6. Y. Hu and T.J. Rupert: Atomistic modeling of interfacial segregation and structural transitions in ternary alloys. J. Mater. Sci. , 3975-3993 (2019). 7. M. Tang, W.C. Carter and R.M. Cannon: Grain Boundary Transitions in Binary Alloys. Phys. Rev. Lett. , 075502 (2006). 8. H.V. Atkinson: Overview no. 65: Theories of normal grain growth in pure single phase systems. Acta Metall. , 469-491 (1988). 9. H. Li and F. Ebrahimi: Ductile-to-Brittle Transition in Nanocrystalline Metals. Adv. Mater. , 1969-1972 (2005). 10. C.C. Koch, D.G. Morris, K. Lu and A. Inoue: Ductility of Nanostructured Materials. MRS Bull. , 54-58 (1999). 11. J.D. Schuler, O.K. Donaldson and T.J. Rupert: Amorphous complexions enable a new region of high temperature stability in nanocrystalline Ni-W. Scr. Mater. , 49-53 (2018). 12. J. Ding, Z. Shang, Y.F. Zhang, R. Su, J. Li, H. Wang and X. Zhang: Tailoring the thermal stability of nanocrystalline Ni alloy by thick grain boundaries.
Scr. Mater. , 21-26 (2020). 13. J.T. Zhao, J.Y. Zhang, L.F. Cao, Y.Q. Wang, P. Zhang, K. Wu, G. Liu and J. Sun: Zr alloying effect on the microstructure evolution and plastic deformation of nanostructured Cu thin films.
Acta Mater. , 550-564 (2017). 14. Z. Pan and T.J. Rupert: Amorphous intergranular films as toughening structural features.
Acta Mater. , 205-214 (2015). 15. A. Khalajhedayati, Z. Pan and T.J. Rupert: Manipulating the interfacial structure of nanomaterials to achieve a unique combination of strength and ductility. Nat. Comm. , 1-8 (2016). 16. J. Luo, H. Wang and Y.-M. Chiang: Origin of Solid-State Activated Sintering in Bi2O3-Doped ZnO. J. Am. Ceram. Soc. , 916-920 (1999). 17. E. Jud, C.B. Huwiler and L.J. Gauckler: Sintering Analysis of Undoped and Cobalt Oxide Doped Ceria Solid Solutions. J. Am. Ceram. Soc. , 3013-3019 (2005). 18. J. Luo: Liquid-like interface complexion: From activated sintering to grain boundary diagrams. Curr. Opin. Solid State Mater. Sci. , 81-88 (2008). 19. V.K. Gupta, D.-H. Yoon, H.M. Meyer and J. Luo: Thin intergranular films and solid-state activated sintering in nickel-doped tungsten. Acta Mater. , 3131-3142 (2007). 24 20. X. Shi and J. Luo: Grain boundary wetting and prewetting in Ni-doped Mo. Appl. Phys. Lett. , 251908 (2009). 21. J. Nie, J.M. Chan, M. Qin, N. Zhou and J. Luo: Liquid-like grain boundary complexion and sub-eutectic activated sintering in CuO-doped TiO2. Acta Mater. , 329-338 (2017). 22. O.K. Donaldson and T.J. Rupert: Amorphous Intergranular Films Enable the Creation of Bulk Nanocrystalline Cu–Zr with Full Density.
Adv. Eng. Mater. , 1900333 (2019). 23. J.D. Schuler, C.M. Grigorian, C.M. Barr, B.L. Boyce, K. Hattar and T.J. Rupert: Amorphous intergranular films mitigate radiation damage in nanocrystalline Cu-Zr. Acta Mater. , 341-354 (2020). 24. J.E. Ludy and T.J. Rupert: Amorphous intergranular films act as ultra-efficient point defect sinks during collision cascades arXiv e-prints . 25. C.M. Grigorian and T.J. Rupert: Thick amorphous complexion formation and extreme thermal stability in ternary nanocrystalline Cu-Zr-Hf alloys.
Acta Mater. , 172-182 (2019). 26. C.M. Grigorian and T.J. Rupert: Critical cooling rates for amorphous-to-ordered complexion transitions in Cu-rich nanocrystalline alloys.
Acta Mater. , 116650 (2021). 27. N. Zhou, T. Hu, J. Huang and J. Luo: Stabilization of nanocrystalline alloys at high temperatures via utilizing high-entropy grain boundary complexions.
Scr. Mater. , 160-163 (2016). 28. G. Wu, C. Liu, L. Sun, Q. Wang, B. Sun, B. Han, J.-J. Kai, J. Luan, C.T. Liu, K. Cao, Y. Lu, L. Cheng and J. Lu: Hierarchical nanostructured aluminum alloy with ultrahigh strength and large plasticity.
Nat. Comm. , 5099 (2019). 29. G. Wu, S. Balachandran, B. Gault, W. Xia, C. Liu, Z. Rao, Y. Wei, S. Liu, J. Lu, M. Herbig, W. Lu, G. Dehm, Z. Li and D. Raabe: Crystal–Glass High-Entropy Nanocomposites with Near Theoretical Compressive Strength and Large Deformability. Adv. Mater. , 2002619 (2020). 30. T. Yang, Y.L. Zhao, W.P. Li, C.Y. Yu, J.H. Luan, D.Y. Lin, L. Fan, Z.B. Jiao, W.H. Liu, X.J. Liu, J.J. Kai, J.C. Huang and C.T. Liu: Ultrahigh-strength and ductile superlattice alloys with nanoscale disordered interfaces. Science , 427 (2020). 31. N. Zhou, T. Hu and J. Luo: Grain boundary complexions in multicomponent alloys: Challenges and opportunities.
Curr. Opin. Solid State Mater. Sci. , 268-277 (2016). 32. C. Suryanarayana: Mechanical alloying and milling. Prog. Mater Sci. , 1-184 (2001). 33. G.S. Rohrer: “Introduction to Grains, Phases, and Interfaces—an Interpretation of Microstructure,” Trans. AIME, 1948, vol. 175, pp. 15–51, by C.S. Smith. Metall. Mater. Trans. A , 1063-1100 (2010). 34. K. Okada and T. Sakuma: The Role of Zener's Pinning Effect on the Grain Growth in Al2O3-ZrO2. J. Ceram. Soc. Jpn. , 382-386 (1992). 35. C. Brandl, T.C. Germann and A. Misra: Structure and shear deformation of metallic crystalline–amorphous interfaces.
Acta Mater. , 3600-3611 (2013). 36. O.K. Donaldson, K. Hattar, T. Kaub, G.B. Thompson and J.R. Trelewicz: Solute stabilization of nanocrystalline tungsten against abnormal grain growth. J. Mater. Res. , 1-13 (2017). 37. W. Xing, A.R. Kalidindi and C.A. Schuh: Preferred nanocrystalline configurations in ternary and multicomponent alloys.
Scr. Mater. , 136-140 (2017). 25 38. W. Xing, A. Kalidindi, D. Amram and C. Schuh: Solute Interaction Effects on Grain Boundary Segregation in Ternary Alloys.
Acta Mater. , (2018). 39. L. Li, E.-B. Liu, Q.-F. Li and Z. Li: Non-equilibrium grain boundary cosegregation of Mo and P.
Appl. Surf. Sci. , 3989-3992 (2006). 40. H. Erhart and H.J. Grabke: Equilibrium segregation of phosphorus at grain boundaries of Fe–P, Fe–C–P, Fe–Cr–P, and Fe–Cr–C–P alloys.
Metal Science , 401-408 (1981). 41. W. Yu-Qing and C.J. McMahon: Interaction of phosphorus, carbon, manganese, and chromium in intergranular embrittlement of iron. Mater. Sci. Technol. , 207-216 (1987). 42. J.R. Rice and J.-S. Wang: Embrittlement of interfaces by solute segregation. Materials Science and Engineering: A , 23-40 (1989). 43. J.D. Schuler and T.J. Rupert: Materials selection rules for amorphous complexion formation in binary metallic alloys.
Acta Mater. , 196-205 (2017). 44. M.A. Atwater and K.A. Darling: A Visual Library of Stability in Binary Metallic Systems: The Stabilization of Nanocrystalline Grain Size by Solute Addition: Part 1, (City, 2012). 45. K.E.E. D.A. Porter, M. Sherif: Phase Transformations in Metals and Alloys (Revised Reprint) (CRC Press, City, 2009). 46. Z. Pan and T.J. Rupert: Formation of ordered and disordered interfacial films in immiscible metal alloys.
Scr. Mater. , 91-95 (2017). 47. A. Inoue: Stabilization of metallic supercooled liquid and bulk amorphous alloys.
Acta Mater. , 279-306 (2000). 48. A. Inoue: Stabilization of Supercooled Liquid and Opening-up of Bulk Glassy Alloys. Proceedings of the Japan Academy, Series B , 19-24 (1997). 49. Z.P. Lu and C. Liu: A New Glass-Forming Ability Criterion for Bulk Metallic Glasses. Acta Mater. , 3501-3512 (2002). 50. Y.C. Kim, J.C. Lee, P.R. Cha, J.P. Ahn and E. Fleury: Enhanced glass forming ability and mechanical properties of new Cu-based bulk metallic glasses. Materials Science and Engineering: A , 248-253 (2006). 51. W.B. Sheng: Correlations between critical section thickness and glass-forming ability criteria of Ti-based bulk amorphous alloys.
J. Non-Cryst. Solids , 3081-3086 (2005). 52. H. Men, S.J. Pang and T. Zhang: Glass-forming ability and mechanical properties of
Cu50Zr50−xTix alloys.
Materials Science and Engineering: A , 326-329 (2005). 53. W.L. Johnson: Bulk Glass-Forming Metallic Alloys: Science and Technology.
MRS Bull. , 42-56 (1999). 54. X.F. Zhang, H.X. Li, Z.Q. Jiang and S.H. Yi: Effect of Minor Si and Ag Additions on Glass-Forming Ability of Ti-Cu-Co-Zr-Sn-Be Bulk Metallic Glass. Mater. Sci. Forum , 400-406 (2011). 55. T.A. Baser and M. Baricco: Fe-based bulk metallic glasses with Y addition.
J. Alloys Compd. , 176-179 (2007). 56. J. Yin, G.Y. Yuan, Z.H. Chu, J. Zhang and W.J. Ding: Mg–Ni–(Gd,Nd) bulk metallic glasses with improved glass-forming ability and mechanical properties.
J. Mater. Res. , 2130-2140 (2009). 57. S.F. Zhao, G.N. Yang, H.Y. Ding and K.F. Yao: A quinary Ti–Zr–Hf–Be–Cu high entropy bulk metallic glass with a critical size of 12 mm. Intermetallics , 47-50 (2015). 58. A.I. C. Suryanarayana: Bulk Metallic Glasses, 22 November 2017 ed., City, 2017). 59. J.F. Löffler: Bulk metallic glasses. Intermetallics , 529-540 (2003). 26 60. M. Kapoor, T. Kaub, K.A. Darling, B.L. Boyce and G.B. Thompson: An atom probe study on Nb solute partitioning and nanocrystalline grain stabilization in mechanically alloyed Cu-Nb. Acta Mater. , 564-575 (2017). 61. M. Yoshitake and K. Yoshihara: Surface segregation of substrate element on metal films in film/substrate combinations with Nb, Ti and Cu.
Surf. Interface Anal. , 509-513 (1992). 62. N.Q. Vo, S.W. Chee, D. Schwen, X. Zhang, P. Bellon and R.S. Averback: Microstructural stability of nanostructured Cu alloys during high-temperature irradiation. Scr. Mater. , 929-932 (2010). 63. L. Lutterotti, M. Bortolotti, G. Ischia, I. Lonardelli and H.R. Wenk: Rietveld texture analysis from diffraction images Tenth European Powder Diffraction Conference, (De Gruyter, City, 2007), pp. 125-130. 27 Alloy Composition (at. %) EDS Concentration (at. %) Cu Zr Hf Mo Nb W Ti Cu-Zr-Hf-Mo-Nb 89.84 2.39 2.20 2.99 2.57 - - Cu-Zr-Hf-Mo-W 91.14 2.78 2.19 2.30 - 1.58 - Cu-Zr-Hf-Nb-Ti 90.27 2.24 2.40 - 2.87 - 2.22 Table 1.
Chemical concentrations of each alloy investigated in this study, as obtained with EDS. Dopant concentrations obtained with EDS are typically subject to measurement error of ~1 at.%. Figure 1.
X-ray diffraction patterns of the (a) Cu-Zr-Hf-Mo-W, (b) Cu-Zr-Hf-Mo-Nb, and (c) Cu-Zr-Hf-Nb-Ti alloy samples. Patterns are shown of the as-milled powders, as well as samples subject to an initial heat treatment at 500 ° C for 5 h followed by an anneal at 950 ° C for 5 min (0.083 h), 1 h, and 1 week (168 h). Phase identification shows the presence of carbide and intermetallic phases which either formed during mechanical alloying or during annealing. Figure 2.
The phase amounts, expressed as volume percentages, of second phases present in the (a) Cu-Zr-Hf-Mo-W, (b) Cu-Zr-Hf-Mo-Nb, and (c) Cu-Zr-Hf-Nb-Ti alloys are plotted as a function of annealing time at 950 ° C. Grain sizes of these second phases for each of the three alloys are shown in parts (d), (e), and (f), respectively. The Cu-Zr-Hf-Mo-W alloy has the greatest total volume percent of second phases, while the Cu-Zr-Hf-Nb-Ti has the lowest volume percent and remains >90% face centered cubic for all annealing treatments. Figure 3. (a) Grain size of each of the three alloys plotted as a function of annealing time at 950 ° C. The Cu-Zr-Hf-Nb-Ti alloy is the most stable against grain growth in this study, retaining a grain size of ~60 nm after 1 week of annealing at this high temperature. (b) SAED pattern confirming the presence of NbC and HfC phases in the Cu-Zr-Hf-Nb-Ti alloy annealed for 5 min at 950 ° C. (c) Bright field TEM micrograph of the microstructure of the Cu-Zr-Hf-Nb-Ti sample annealed for 5 min. The grain size distribution measured for this sample is shown in (d), after measuring 167 grains and obtaining a mean grain size of 38 nm. Figure 4.
HAADF STEM micrograph of the Cu-Zr-Hf-Nb-Ti alloy. Yellow arrows point to regions of bright contrast at grain boundaries, associated with dopant segregation and the possible formation of amorphous complexions. Figure 5. (a) HAADF STEM micrograph and (b) corresponding line scan across a grain boundary in Cu-Zr-Hf-Nb-Ti. The segregation of Nb, Ti, and Zr atoms to this grain boundary is observed for this boundary, while no Hf is measured either at the boundary or inside the grains. (b) A HAADF STEM micrograph and (c) corresponding line scan demonstrate the variation of possible grain boundary segregation behavior in this alloy, where the boundary is enriched with only Nb and Ti for this specific boundary. Figure 6.