Non-Classical Crystallization Pathways in Eutectic-Forming Systems

Abstract

Crystallization is the central process of synthesizing materials across length scales, with ubiquitous examples in synthetic, biogenic, and geologic environments. During crystallization a continuum of patterns could emerge due to the interplay of growth kinetics, material or solution chemistry, and crystallographic defects. In particular, solidification of eutectic alloys, characterized by the proximity of their compositions to a nonvariant point in the phase diagram, produces multi-phased micro- and nanostructures with diverse morphologies. This spontaneous pattern formation lies among the broader self-organization strategies that can be easily scaled to large areas, potentially enabling higher throughput and lower cost than serial processes.

This dissertation sheds new light on non-classical pathways for eutectic crystallization, perplexing characteristics that cannot be satisfactorily explained nor predicted by classical nucleation and growth models. The scope of this work entails a platform combining advanced experimental techniques – precise synthesis along with multiscale, three-dimensional, and time-resolved measurements – and computational methods – computer vision and machine learning – for tracking eutectic formation at temperature and their structural evolution under external stimuli. The first thrust of this dissertation focuses on crystallization in the presence of chemical modifiers, and the second thrust on the emergence of two-phase metastable spirals and their response in extreme environments.

Thrust one demonstrates cases in which the interaction of the modifier with the growing crystal is either synergistic, illustrated in a case study of Al-Si and Al-Ge eutectics, or antagonistic, shown in the growth of primary Si crystals. Thrust two focuses on spiral growth in the Zn-Mg system, and their behavior at elevated temperatures. These spirals are thermodynamically metastable, so their successful synthesis requires steering the system down certain kinetic pathways on intermediate time-scales. Collectively, our multi-modal characterization studies provide the necessary benchmark data for simulations of complex self‐organization patterns, thus expanding the horizon for the design of next‑generation alloys with superior properties.