Olga Wodo

Establishing microstructure–property correlations that generalize across diverse microstructural classes remains a critical challenge in data-driven materials design. In this work, we evaluate the potential to extrapolate predictive models trained on one microstructure type (e.g., spinodal) to others (e.g., dendritic), using three distinct featurization strategies: two-point correlation functions, graph-based descriptors, and deep neural network embeddings. Our findings reveal that the Wasserstein distance is an excellent metric that correlates well with generalizability, serving as a model-agnostic yet data-aware signature of generalizability. Furthermore, we demonstrate that featurizations that conserve key microstructural features generalize better. Graphic

The properties of ceramic materials are strongly influenced by nanoscale defects such as oxygen vacancies and grain boundaries which drive phase transformations and affect functionality. Yttria-Stabilized Zirconia (YSZ), a widely used ceramic, is particularly sensitive to such defects at the atomic scale. In this work, to investigate these nanoscale features in YSZ we employ Atom Probe Tomography (APT) focusing on the Local Atomistic Environment (LAE) of yttrium atoms. We introduce a computational framework that constructs atomic neighborhoods using Voronoi tessellations, represents them as graphs, and quantifies their structure through customizable signature functions. These signature functions, when integrated with unsupervised learning techniques, enable the extraction of meaningful patterns and defect signatures from APT datasets. By representing LAEs as graph-based structures, our approach captures intricate local details of composition, coordination, and connectivity. This representation is particularly effective for indirectly identifying defects such as oxygen vacancies. Our analysis reveals significant differences in yttrium environments of YSZ sample at grain boundaries compared to those within grains, with grain boundary regions showing fewer and less connected oxygen ions. Additionally, we distinguish between two distinct types of grain boundaries of YSZ sample based on their local atomic environments. Finally, we demonstrate that a graph based order parameter derived from the LAE enables the detection of subtle variations in oxygen ion coordination and connectivity features indicative of potential vacancies that are otherwise difficult to resolve. While applied here to YSZ, the framework is generalizable to other material systems. Its flexibility in designing material specific signature functions makes it a powerful and versatile tool for atomic scale characterization of complex materials.

Understanding phase diagrams is essential for material selection and design, as they provide a comprehensive representation of the thermodynamics of mixtures. This work delivers a broad and systematic overview of possible ternary phase diagrams for amorphous systems representative of polymers, small organic molecules, and solvents. Thanks to computationally efficient methods, an unprecedented library of $>$80,000 ternary phase diagrams is generated based on a systematic screening of interaction parameters. Twenty-one phase diagram types, including unreported ones, are identified. They are classified according to simple rules related to the number of immiscible material pairs, of miscibility gaps, and of three-phase regions. They are mapped onto the three-dimensional interaction parameters space, providing a clear picture of their likelihood and existence conditions. Four well-known phase-diagram types with 0, 1, 2, or 3 immiscible pairs are found to be the most likely. The numerous uncommon phase diagrams are mostly observed within a small parameter window around the critical interaction parameter values. For the most common phase diagram types, we show that the size of the processability window becomes sensitive to interaction parameter variations close to critical values. The sensitivity decreases for materials with increasing molar size. Finally, successful comparisons of simulated and experimental phase diagrams showcase the real-world relevance of this theoretical analysis. The presented results lay a robust foundation for rational design of solution processing conditions and for blend morphology control. Immediate applications include organic thin films and the identification of green solvents for sustainable processing.

We develop an active workflow for calibrating microstructure–property relationships when a large dataset of microstructures is available, but the cost associated with evaluating the properties associated is high.

Gradient porous structures (GPS) are characterized by structural variations along a specific direction, leading to enhanced mechanical and functional properties compared to homogeneous structures. This study explores the potential of mycelium, the root part of a fungus, as a biomaterial for generating GPS. During the intentional growth of mycelium, the filamentous network undergoes structural changes as the hyphae grow away from the feed substrate. Through microstructural analysis of sections obtained from the mycelium tissue, systematic variations in fiber characteristics (such as fiber radii distribution, crosslink density, network density, segment length) and pore characteristics (including pore size, number, porosity) are observed. Furthermore, the mesoscale mechanical moduli of the mycelium networks exhibit a gradual variation in local elastic modulus, with a significant change of approximately 50% across a 30 mm thick mycelium tissue. The structure-property analysis reveals a direct correlation between the local mechanical moduli and the network crosslink density of the mycelium. This study presents the potential of controlling growth conditions to generate mycelium-based GPS with desired functional properties. This approach, which is both sustainable and economically viable, expands the applications of mycelium-based GPS to include filtration membranes, bio-scaffolds, tissue regeneration platforms, and more.