Cluster Analysis
Identifying Nanoscale Structures from Elemental Segregation
Nova Science Corner
Cluster analysis is a key method in 3D atom probe data analysis for identifying local elemental enrichment and nanoscale clusters in materials. In complex alloys, microstructures do not always exhibit clear phase interfaces. During the early stages of precipitate formation, certain elements may first enrich locally, forming chemical clusters only a few nanometers or even smaller. Cluster analysis can identify these local chemical anomalies from large 3D atomic datasets, providing an important basis for studying the early microstructural evolution of materials.
Its basic workflow relies on the elemental species and 3D spatial coordinates obtained by APT. Depending on the research target and data characteristics, statistical methods such as the maximum separation method and density-based clustering can be applied. By setting parameters including maximum separation distance, minimum cluster size, elemental combination, and statistical threshold, atomic groups that are spatially close and chemically distinct can be identified. After parameter optimization and statistical validation, quantitative information such as cluster number, size, number density, composition, and spatial distribution can be obtained.
The key value of cluster analysis is that it goes beyond conventional "average composition" to reveal local chemical composition and spatial correlations. In superalloys and high-entropy alloys, for instance, nanoclusters formed by elemental enrichment may act as precursor structures for later precipitate formation. By comparing cluster number, size, and chemical composition under different heat treatment conditions, one can track the progression from elemental segregation → cluster formation → nanoscale precipitation → phase evolution, providing atomic-scale evidence for understanding phase transformation mechanisms and microstructural evolution.
APT's unique advantage is that it provides both near-atomic-scale elemental identification and 3D spatial coordinates. This allows cluster analysis to move beyond local bright spots in 2D images or regional average composition, enabling chemical clusters to be directly identified and statistically analyzed in 3D atomic data, while linking their spatial position, size, and chemical composition. For nanoscale early-stage segregation and precipitation that conventional characterization methods struggle to resolve, this 3D atomic-scale information is highly valuable.
Nova Scientific has conducted APT cluster analysis on complex alloys including nickel-based superalloys and high-entropy alloys, gaining hands-on experience from 3D atomic reconstruction and cluster identification to statistical quantitative analysis. Tailored to different material systems and research goals, we provide targeted analysis of elemental segregation, nanoclusters, and their evolution, delivering data support for studies on phase transformation mechanisms, thermal stability, and microstructure–property relationships.