Magnesium (Mg), a lightweight metal, holds promise as a structural material for energy-saving applications. However, its susceptibility to twinning during deformation leads to plastic anisotropy and low ductility. Despite extensive research, the mechanisms governing the nucleation and early-stage growth of deformation twins remain unclear. A key debate centers on whether {1012}<1011> twinning (common in hexagonal close-packed (HCP) metals) follows a shear-shuffle or pure-shuffle mechanism. Direct visualization of twin nucleation has been challenging due to the localized, ultrafast nature of the process. While twin formation under stress concentrations has been predicted, the actual sequence of events leading to visible twins in Mg hasn't been directly tracked. This study aims to address this gap by combining nanomechanical deformation, in-situ transmission electron microscopy (TEM), and atomic-scale simulations to unravel the nucleation and early-stage growth of {1012} deformation twins in Mg. The researchers hypothesized that by strategically designing the geometry of the Mg crystals, they could isolate and visualize these processes, which are essential for understanding deformation at both a fundamental and technological level. Understanding these mechanisms is crucial for developing strategies to control twinning and enhance the performance of magnesium alloys in structural applications.

The literature extensively explores deformation twinning in Mg, examining the influence of deformation conditions, microstructure, and alloying on twin formation. However, the underlying mechanisms remain debated. Models involving shear-shuffle nucleation, such as the pole mechanism, and pure-shuffle nucleation (a classical phase transformation mechanism) have been proposed. Past studies, primarily relying on theoretical analysis and atomic-scale simulations, have provided insights but lacked direct experimental visualization of the early stages of twin formation. In-situ TEM studies on conventional pillars have captured twin propagation, but the speed of this process has obscured the nucleation event. Previous simulations have predicted twin formation under nanoindentation stress concentrations, but they didn't track the initial stages of twin growth. The current study builds upon these prior works by combining experimental observation with computational modeling to provide a more comprehensive understanding of the process.

The researchers employed strategically designed truncated wedge-shaped pillars (TWPs) fabricated from single-crystal Mg using focused ion beam (FIB) milling. These pillars, with varying top widths (100, 250, and 400 nm), were designed to generate a steep stress field under compression, promoting twin nucleation while inhibiting rapid propagation. Finite element analysis (FEA) was used to predict the stress distribution within the pillars, confirming the localized high-stress regions at the pillar tops. In-situ TEM compression experiments were conducted on these pillars, visualizing twin formation and growth under controlled conditions. The compression speed varied depending on the pillar geometry, allowing for detailed observation of the processes. Post-mortem atomic-resolution TEM and scanning TEM (STEM) were used to analyze the crystallographic structure of the resulting twins. Molecular dynamics (MD) simulations using an empirical interatomic potential were performed to complement the experimental observations and gain atomic-level insights into the nucleation and growth mechanisms. The simulations employed a wedge-shaped model under compressive loading to simulate the experimental conditions. The simulations tracked the atomic rearrangements during twin formation and growth, providing validation for the experimental findings.

The study revealed significant differences in twinning behavior depending on pillar geometry. In conventional, uniform-stress pillars, large twins appeared abruptly, precluding the observation of nucleation. In contrast, the TWPs allowed for the distinction between nucleation and growth. In TWPs with a 400 nm top, two twins nucleated at the corners and expanded until they intersected. In pillars with 250 nm tops, a triangular twin initially formed and then expanded along the basal plane, without significant shearing along the twinning plane. In the narrowest TWPs (100 nm top), the formation and movement of a much smaller twin nucleus (~20 nm initially) were tracked, showing that it expanded primarily along the basal plane. TEM characterization of the twin boundaries in all the TWPs consistently revealed the presence of basal-prismatic (BP) and prismatic-basal (PB) interfaces, not the typical {1012} coherent twin boundaries (CTBs) of mature twins. This finding directly supports a pure-shuffle nucleation mechanism, involving prismatic-basal transformations and atomic shuffling without significant shear along the twinning plane. MD simulations confirmed this mechanism, showing the formation of twin nuclei with BP and PB interfaces through atomic shuffling. The simulations further demonstrated that the initial growth of the twin nuclei occurs through the migration of these interfaces, with basal <a> dislocation emission playing a role in stress relaxation. The study also found a crystal geometry-dependent twin growth mechanism; in regular pillars, the early-stage growth is dominated by fast glide-shuffle along the twinning plane, leading to unstable plastic flow and strain bursts. However, in TWPs, the slower migration of PB and BP interfaces leads to more stable plastic flow and reduced strain bursts, suggesting a potential strategy for enhancing the plasticity of Mg through geometric control.

The findings directly address the long-standing debate regarding the twinning mechanism in Mg, providing compelling evidence for the pure-shuffle mechanism. The use of strategically designed geometries was instrumental in isolating and visualizing the early stages of twinning, overcoming previous limitations associated with the high speed of twin propagation. The observation of BP and PB interfaces, along with the supporting MD simulations, strongly supports the pure-shuffle mechanism over conventional shear-shuffle models. The geometry-dependent growth mechanism suggests that tailoring the microstructure, particularly the grain geometry in polycrystalline Mg, could lead to enhanced plasticity and control over the anisotropic response of the material. This study demonstrates a powerful approach to investigate nucleation and growth phenomena in materials, opening new avenues for optimizing material properties.

This study elucidated the nucleation and early-stage growth of {1012} twins in Mg, demonstrating a pure-shuffle mechanism based on prismatic-basal transformations and atomic shuffling. The strategically designed TWPs enabled the visualization of these previously elusive stages. The observed geometry-dependent growth mechanism offers potential strategies for controlling twinning and enhancing plasticity in Mg. This work represents a significant advance in the fundamental understanding of deformation twinning and highlights the power of combining in-situ TEM with atomic-scale simulations.

The study primarily focused on single-crystal Mg pillars, potentially limiting the direct applicability to polycrystalline materials. While the samples were designed to minimize FIB-induced artifacts, some residual effects might remain. The MD simulations, while validating the experimental findings, were limited by computational constraints, requiring relatively small model sizes. The study focuses mainly on the initial growth of the twin nucleus; future studies could explore the transition from the initial nucleus to a mature twin with fully developed CTBs.