Achieving Large Uniform Tensile Ductility in Nanocrystalline Metals Y. M. Wang, 1, * R. T. Ott, 2,† A. V. Hamza, 1 M. F. Besser, 2 J. Almer, 3 and M. J. Kramer 2 1 Physical and Life Sciences Directorate, Lawrence Livermore National Laboratory, Livermore, California 94550, USA 2 Division of Materials Science and Engineering, Ames Laboratory (USDOE), Ames, Iowa 50011, USA 3 Advanced Photon Source, Argonne National Laboratory, Argonne, Illinois 60439, USA (Received 20 January 2010; published 17 November 2010) Synchrotron x-ray diffraction and high-resolution electron microscopy revealed the origin of different strain hardening behaviors (and dissimilar tensile ductility) in nanocrystalline Ni and nanocrystalline Co. Planar defect accumulations and texture evolution were observed in Co but not in Ni, suggesting that interfacial defects are an effective passage to promote strain hardening in truly nanograins. Twinning becomes less significant in Co when grain sizes reduce to below 15 nm. This study offers insights into achieving excellent mechanical properties in nanocrystalline materials. DOI: 10.1103/PhysRevLett.105.215502 PACS numbers: 62.20.F, 61.46.Hk Although inspiring progress has been made in achieving extraordinary mechanical properties in nanocrystalline (NC) solids (grain sizes D< 100 nm), the lack of tensile ductility in NC metals remains a long-standing problem [1]. This obvious shortcoming is widely attributed to their lack of strain hardening; i.e., a large amount of dislocation storage has not been realistic in most NC metals. This is especially true when D decreases below 30 nm—a regime where the grain boundaries (GBs) become dominant sources or sinks for dislocations. In some other cases, however, large strain hardening and uniform tensile elongation (" unif ) have been observed in NC metals [2–4]. The physical origin of these experimental observations unfortunately remains specula- tive. Recently, strain hardening and dislocation accumula- tions through a Lomer-Cottrell lock mechanism were reported in cryogenic-rolled NC-Ni (D  20 nm)[5]. The general implications of these results in enhancing the " unif of NC metals remain uncertain however, as NC-Ni exhibits a persistently low " unif ð<2%Þ even at cryogenic temperatures [6]. Collectively, these ongoing investigations suggest that various deformation mechanisms (DMs) reported so far under nontensile conditions may bear less implication to tensile plasticity due to clearly different achievable strain levels and possible tension-compression asymmetry. The loading mode becomes more problematic in hexagonal close-packed (hcp) metals, where different slip/twinning systems are operative under tension or compression. These load-dependent DMs in NC metals underscore the impor- tance of using in situ tensile experiments to investigate the relevant DMs. Here we study the DMs of NC-Ni and NC-Co through synchrotron x-ray diffraction (SXRD) experiments with the key aim to uncover the physical origin of their intrinsic differences in strain hardening and resultant tensile plas- ticity. Both metals are electrodeposited with clean grain interiors. The face-centered cubic (fcc) NC-Ni is from the same batch as reported in Ref. [6]. Transmission electron microscopy (TEM) examinations indicate that it has a D distribution (29  11 nm; 25 nm from SXRD data). The as-synthesized hcp NC-Co has an average D of 20  9 nm, and contains a negligible amount (< 0:5%) of fcc phase evidenced by the small ð200Þ fcc peak in the SXRD data shown in Fig. 1(a). This low-intensity peak was not revealed previously by conventional Cu K  XRD [3], but is consistent with a few growth twins observable in TEM images. The NC-Ni has a weak (111) in-plane peak based on the (111) intensity ratio relative to other peaks. Room- temperature quasistatic tensile tests at a strain rate of 1:1  10 4 s 1 displayed in Fig. 1(b) indicate that both NC metals exhibit appreciable strain hardening immediately following macroscopic yielding, but head off to clearly different trends as the plastic strain progresses, leading to an eminent disparity in tensile ductility; i.e., the achievable " unif of NC-Co ( 8:4%) is about 3 times larger than that of NC-Ni ( 2:7%). The inset of Fig. 1(b) summarizes the yield strength ( y ) vs " unif for various NC metals with an average D< 30 nm, showing a few examples of large-" unif (> 5%) metals, including NC-Cu [2], NC-Co [3], and NC- Ni-Fe [4]. Earlier work attributes this intriguing large " unif to various reasons such as dislocation-trapping [2], twin- ning [3], and/or sample quality [4], but no direct evi- dence was obtained. The high-energy (E ¼ 80:715 keV, 0.153 589 A ˚ wavelength) x-rays at sector 1 of the Advanced Photon Source enables real-time tracking of the lattice strain and peak intensity in NC metals during tensile load- ing (the sample gauge section is 4:5 mm long  2:2 mm wide  0:13 mm thick). The samples were loaded in a MTS 858 frame, while the diffracted intensity from the samples was collected on an amorphous Si detector (41  41 cm 2 in area) positioned 1103 mm from the NC-Ni and 970 mm from the NC-Co. The pixel size of the camera is 200  200 m 2 . The Q-space resolution for both camera lengths is better than 0:01  A 1 . For SXRD studies, a useful parameter for understanding DMs is the in-plane lattice strain, which is defined as " hkl ¼ðd hkl  d hkl 0 Þ=d hkl 0 , where hkl denotes grains with PRL 105, 215502 (2010) PHYSICAL REVIEW LETTERS week ending 19 NOVEMBER 2010 0031-9007= 10=105(21)=215502(4) 215502-1 Ó 2010 The American Physical Society