Bulk metallic glasses (BMGs) have been intensively investigated because of their special mechanical properties as amorphous materials and their unique glass transition state. However, the atomic-scale origins of their behavior have not been unequivocally clarified. To explain their properties, structural models of BMGs and their small-scale deformation behavior have been proposed but not yet confirmed due to the inability of conventional measurement approaches to characterize samples at the relevant scale. For example, local structural analysis of glasses at the atomic scale using methods such as transmission electron microscopy or neutron/x-ray scattering is challenging due to the material’s disordered nature. In contrast, scanning probe microscopes and nanoindenters possess the potential of direct nanometer-scale observation local glass structure and mechanical properties. But even with these approaches, extraction of meaningful data is challenging due to the difficulty to prepare clean, atomically flat surfaces of BMG. This is because a surface roughness of some nanometers, standard with most sample preparation techniques, may alter the results of local testing if the volumes probed are nanometer-sized as well.

In this talk, we will be reviewing our recent progress in developing novel imprinting and fabrication methods of metallic glasses that can produce both atomically flat surfaces with sub-nanometer-scale features and samples with well-defined nanometer- and micron-sized total volumes as well as their subsequent use for the study of their nanometer-scale structural and mechanical properties. Imprinting is realized via thermoplastic forming of BMGs [1,2] and, alternately, by magnetron sputtering of general metallic glasses [3]. The capability of imprinting at an atomic scale enriches the range of applications of BMGs and brings a new way to directly characterize heterogeneity, relaxation, and crystallization in BMGs [4, 5]. It also allows to study onset of yielding and the local plastic flow mechanisms of BMGs in the limit of very small activation volumes (about 1000 atoms). The experiments revealed a much higher yield stress compared to the value obtained by conventional nanoindentation testing, followed by homogeneous plastic flow [6]. These atomic-scale results are contrasted to the larger-scale model that explains plastic deformation of BMG as originating from the finite STZs activation. Finally, current work is aimed at producing large numbers (>1000) of well defined, uniform micron- or nanometer-scaled pillars that can be used to explore the deformation behavior of BMGs under compression as a function of sample volume and compression rate in a statistically relevant manner.