Manipulating the interfacial structure of nanomaterials to achieve a unique combination of strength and ductility

Abstract

The control of interfaces in engineered nanostructured materials has met limited success compared with that which has evolved in natural materials, where hierarchical structures with distinct interfacial states are often found. Such interface control could mitigate common limitations of engineering nanomaterials. For example, nanostructured metals exhibit extremely high strength, but this benefit comes at the expense of other important properties like ductility. Here, we report a technique for combining nanostructuring with recent advances capable of tuning interface structure, a complementary materials design strategy that allows for unprecedented property combinations. Copper-based alloys with both grain sizes in the nanometre range and distinct grain boundary structural features are created, using segregating dopants and a processing route that favours the formation of amorphous intergranular films. The mechanical behaviour of these alloys shows that the trade-off between strength and ductility typically observed for metallic materials is successfully avoided here.

Figure 1. Alloy design strategy for adding segregating dopants.

Pure Cu (a) can be converted to an alloy by adding Zr during ball milling (b). Although the Zr is mixed throughout the grain structure, annealing treatments can then be used to induce preferential Zr segregation to the grain boundaries (c). Both the (d) pure Cu and (e) Cu-Zr as-milled samples have an average grain size of 30 nm (scale bars, 100 nm). (f) Annealing the Cu-Zr sample at 950 °C for 1 h allows for segregation and only causes coarsening to a grain size of 45 nm (scale bar, 100 nm). The same level of Zr segregation was found in both the quenched and air-cooled samples.

Figure 2. High-resolution TEM images of grain boundary structure in nanocrystalline Cu-Zr alloys.

(a) An amorphous intergranular film with thickness of 5.7 nm was observed at a grain boundary after quickly quenching from 950 °C (scale bar, 5 nm). (b) In contrast, grain boundaries in a slowly cooled sample, with structures that are in equilibrium near ambient temperatures, are all ordered interfaces (scale bar, 2 nm). Insets are fast Fourier transform patterns, highlighting the disordered nature of the interface in a. (c,d) Additional examples of amorphous complexions in the quenched sample (scale bars, 2 nm), whereas e summarizes the measurements from the 28 interfacial films found here.

Figure 3. Mechanical testing results from nanocrystalline Cu and Cu-Zr samples.

Failed micropillars after compression, as well as yield strength measurements, are shown for (a) pure Cu with ordered interfaces, (d) Cu-Zr with ordered interfaces and (g) Cu-Zr with amorphous intergranular films (scale bars, 5 μm). Failed micropillars after bending, as well as strain-to-failure measurements, are shown for (b,c) pure Cu with ordered interfaces, (e,f) Cu-Zr with ordered interfaces and (h,i) Cu-Zr with amorphous intergranular films (scale bars, 2 μm). A nanocrystalline alloy with AIFs can be both stronger and more ductile than its traditional, pure metal counterpart.

Figure 4. Connection between interfacial structure and damage tolerance.

Molecular dynamics simulations of dislocation absorption at (a) an ordered grain boundary and (b) an amorphous intergranular film show the formation and propagation of crack damage. The ordered interface quickly fractures, whereas the amorphous interface diffuses the strain concentration brought by dislocation absorption and fracture is delayed. The delay of failure explains why (d) nanocrystalline Cu-Zr with AIFs has ductility reminiscent of (e) coarse-grained Cu while retaining the high strength of (c) nanocrystalline Cu (scale bars, 5 μm).

Figure 5. Strain-to-failure and yield strength for Cu and Cu-based alloys.

The vast majority of data fall within the grey envelope, with recent advanced alloys pushing slightly outside this limit. Our Cu-Zr alloy breaks the expected trend, with both higher strength and ductility than the pure Cu sample. All strain-to-failure data comes from material that failed under tensile stresses, either under uniaxial tension (literature data) or from the tensile side of a bending experiment (this study). Literature data are taken from refs , , , , , , , .