Abstract: The design of high‐performance structural thin films consistently seeks to achieve a delicate equilibrium by balancing outstanding mechanical properties like yield strength, ductility, and substrate adhesion, which are often mutually exclusive. Metallic glasses (MGs) with their amorphous structure have superior strength, but usually poor ductility with catastrophic failure induced by shear bands (SBs) formation. Herein, we introduce an innovative approach by synthesizing MGs characterized by large and tunable mechanical properties, pioneering a nanoengineering design based on the control of nanoscale chemical/structural heterogeneities. This is realized through a simplified model Zr 24 Cu 76 /Zr 61 Cu 39 , fully amorphous nanocomposite with controlled nanoscale periodicity ( Λ , from 400 down to 5 nm), local chemistry, and glass–glass interfaces, while focusing in‐depth on the SB nucleation/propagation processes. The nanolaminates enable a fine control of the mechanical properties, and an onset of crack formation/percolation (>1.9 and 3.3%, respectively) far above the monolithic counterparts. Moreover, we show that SB propagation induces large chemical intermixing, enabling a brittle‐to‐ductile transition when Λ  ≤ 50 nm, reaching remarkably large plastic deformation of 16% in compression and yield strength ≈2 GPa. Overall, the nanoengineered control of local heterogeneities leads to ultimate and tunable mechanical properties opening up a new approach for strong and ductile materials.

Abstract: Strain has a strong effect on the properties of materials and the performance of electronic devices. Their ever shrinking size translates into a constant demand for accurate and precise measurement methods with a very high spatial resolution. In this regard, transmission electron microscopes are key instruments thanks to their ability to map strain with a subnanometer resolution. Here, we present a method to measure strain at the nanometer scale based on the diffraction of electron Bessel beams. We demonstrate that our method offers a strain sensitivity better than 2.5 × 10−4 and an accuracy of 1.5 × 10−3, competing with, or outperforming, the best existing methods with a simple and easy to use experimental setup.