Imagine a world where steel could be molded as easily as plastic. In fact, it may be possible to do so using metallic glasses—materials that are both strong and elastic, making them ideal for producing golf heads, for example. And like most glasses, the metallic form softens when heated to modest temperatures. Unfortunately, they are brittle and tend to fracture under tensile strain, when shear bands form in unpredictable locations. However, recent work by Hua Guo and colleagues at Institute of Metal Research, Chinese Academy of Sciences, Shenyang and En Ma from John Hopkins University, Maryland, shows that vastly improved tensile ductility is possible.1

Instead of oxides, as for window glass, metallic glasses contain metal atoms. Upon cooling, most metals form crystals, but if cooled fast enough, atoms freeze in place before they can form a crystalline lattice so the resulting solid is amorphous. It is not always practical, or possible, to rapidly cool such materials. Hence a novel form of alchemy is necessary to avoid crystallization: use of appropriate mixtures of atoms of different sizes to ensure the absence of crystalline order.

In this study, Guo and co-workers used Zr52.5Cu17.9Al10Ni14.6Ti5; a bulk metallic glass whose properties are well known. The difference in this group’s approach was that the authors used transmission electron microscopy to examine the deformation properties of the samples on the scale of 100 nm. Notably, they found qualitatively different behavior in small-volume metallic glasses—in other words, size matters.

Fig. 1: Elongation under strain captured by a transmission electron microscope. The dashed white lines represent the strain gauge section (average strain 5x10-4 s-1). a, The sample before strain. b, Up to a strain of 15% the sample elongates uniformly, then non-uniform deformation occurs, with a necked region marked by a white arrow (d). In total, the tensile strain reached 45%.© 2008 Nature Materials

Straining the samples resulted in deformation at ‘necked’ regions with the deformation gradually spreading during elongation (Fig.1). This type of gradual necking is normally found in ductile metals. But even in the heavily deformed areas, both before and after fracture, there was no trace of crystallization. Thus the authors concluded that the plasticity originates in the flow of monolithic glass, rather than through the formation of shear bands or the presence of nanocrystals.

Why is there a size-effect? It may be that nanometer sized samples contain fewer flaws, which reduces the probability of localized shear bands forming and thus enables multiple atomic-level shear events to occur throughout the sample. Moreover, catastrophic crack propagation cannot take place, as there is an associated critical length-scale for brittle failure that is much larger than the sample.

The fact that metallic glasses can, in principle, deform plastically like their crystalline counterparts without catastrophic failure, is an important fundamental advance, which could lead to applications in coatings and micro-gears.