Direct measurement of nanostructural change during in situ deformation of a bulk metallic glass

2019-06-30 13:36:12 SUTOR EYEWEAR


To date, there has not yet been a direct observation of the initiation and propagation of individual defects in metallic glasses during deformation at the nanoscale. Here, we show through a combination of in situ nanobeam electron diffraction and large-scale molecular dynamics simulations that we can directly observe changes to the local short to medium range atomic ordering during the formation of a shear band. We observe experimentally a spatially resolved reduction of order prior to shear banding due to increased strain. We compare this to molecular dynamics simulations, in which a similar reduction in local order is seen, and caused by shear transformation zone activation, providing direct experimental evidence for this proposed nucleation mechanism for shear bands in amorphous solids. Our observation serves as a link between the atomistic molecular dynamics simulation and the bulk mechanical properties, providing insight into how one could increase ductility in glassy materials.


Bulk metallic glasses (BMGs) are an interesting class of materials noted for their wide variety of mechanical properties, associated most notably with their lack of long-range crystallographic order1. BMGs include alloys that exhibit extremely high strength in excess of most engineering materials2,3, as well as low stiffness and high elastic strain limits4,5. Because of these wide-ranging properties, BMGs are attractive alloys for future applications as they offer a potential for the development of stronger and tougher structural materials6,7,8,9.

One of the main impedances to the adoption of high-strength BMGs is their limited ductility, which can be restricted by single shear band formation and rapid propagation at low strains, which often results in catastrophic failure10,11,12,13,14. Accordingly, of key importance to further alloy development is understanding how such shear bands originate at the nanoscale because, although single shear-band formation can cause BMGs to fail at near-zero tensile ductilities, multiple shear-band formation represents the fundamental essence of plasticity in these alloys. As BMGs invariably display high strength, the creation of tensile ductility — via multiple shear banding — is thus essential to their fracture toughness, and hence damage-tolerance, in terms of their potential role as future structural materials.

Several mechanisms have been proposed for the initiation and propagation for shear bands, the predominant hypotheses being free volume softening15,16,17,18,19,20, adiabatic heating softening21,22,23, and shear transformation zones (STZs)20,24,25. Recently, with advancements in both modeling26,27,28 and experimentation29, STZ formation has emerged as the prevailing mechanism by which shear bands form and propagate10,11. In this theory, a STZ is a cluster of atoms which plastically rearranges under mechanical stress. As the stress to transform many STZs homogeneously is very high, in a real material this is hypothesized to preferentially occur at stress concentrations10,30. Once a high enough density of activated STZs have formed, a shear band develops and can propagate10,11.

To date, observing the mechanisms of BMG shear band formation, while possible in molecular dynamics (MD) simulations, has been experimentally challenging due to the high rate of the catastrophic shear band propagation and the current experimental limits of electron microscopy. However, observing shear band nucleation and dynamics at the scales possible in transmission electron microscopy (TEM) is crucial to linking our understanding of deformation mechanisms provided by MD simulations to the macroscale mechanical behavior. Previous TEM experiments in bulk metallic glasses have largely been limited to ex situ qualitative imaging studies with high enough resolution to resolve shear bands, but have difficulties in quantitative interpretation31,32,33,34, or more quantitative fluctuation electron microscopy (FEM) studies on the structure of BMGs35,36,37,38,39 that fall below the local spatial resolution needed for individual shear band characterization. In situ experiments to date have been qualitative, too slow during acquisition, hard to interpret due to a lack of understanding of the contrast mechanisms in shear bands, or at too low of a spatial resolution to be comparable to MD models40. Recent advancements in techniques and hardware have, however, allowed for the observation of strain41 and as we will show here, the evolution of locally resolved atomic short and medium range order, with nanometer resolution during in situ deformation, providing much more comparable information to the significant modeling efforts which have been performed.

In this study, we design an in situ sample to study the coupling of local atomic order and strain during tensile deformation. The BMG used in this study is a member of the model glass family CuxZryAl100−(x+y)42,43, which has been extensively studied for its high glass-forming ability44, and relative ease of computational modeling. These glasses have local clusters of atoms that pack into icosohedral structures45,46,47,48,49, which due to their two-, three- and five-fold symmetry axes in projection, have characteristic symmetric diffraction patterns50. We directly observe a change in structural order correlated with strain as measured from the NBED patterns acquired during in situ deformation.


In situ nanobeam electron diffraction

Specifically, the sample used was Cu46Zr46Al8, which was thinned to electron transparency (~80–90 nm) and then milled using a focused ion beam (FIB) into an in situ tensile bar specimen. The annular dark field images (ADF) of the resulting sample and subsequent deformation can be seen in Fig. 1a. Unique to this in situ experiment, after each 10 nm increase in deformation, deformation was paused and a nanobeam electron diffraction (NBED) dataset was acquired, in which a full diffraction pattern was acquired for each ADF image pixel at 400 frames per second, for a total of 167,440 diffraction patterns. The full experimental procedures can be found in the attached methods section. The diffraction patterns were then used to measure the spatially resolved evolution of both strain41 and short and medium range order48,49,50,51,52 at every probe position over a large area as the sample was mechanically deformed, with a spatial resolution (probe position step size) of 2.5 nm. It should be noted that these diffraction patterns arise through an interaction of the electron beam with a finite sized (1.47 nm FWHM) volume projected through the sample thickness, and therefore symmetry elements in the patterns can arise from interactions with multiple oriented clusters, making it impossible in this experiment to distinguish between singular oriented clusters (short range order) and cluster networks (medium range order). In addition, it has been shown that as the sample thickness increases, higher order symmetries are extinguished in the diffraction patterns51.