Abstract
The increasing demand for high-performance protective coatings in the aerospace and defense sectors—necessary to safeguard against corrosion and degradation—has led to a heightened reliance on essential minerals such as tungsten, titanium, chromium, and rare-earth elements (REEs). These minerals are critical for producing physical vapor deposition (PVD) cathode targets. PVD processes involve sophisticated techniques that utilize high-purity cathode targets. These targets are vaporized in a vacuum environment and deposited onto various substrates to create dense and durable coatings. Many metallic and ceramic components found in these targets, including tungsten, titanium, chromium, and rare-earth elements, are classified as critical minerals. Their importance arises from both their economic value and their vulnerability to supply chain disruptions, which pose risks to industries reliant on these materials. Additionally, the extraction of these minerals often incurs significant environmental costs and faces growing challenges related to supply shortages. As supply chains for REEs encounter increasing geopolitical and environmental constraints, there is an urgent need to identify viable alloy alternatives that can reduce or eliminate dependence on REEs without compromising coating performance. This research presents a machine-learning-driven framework for discovering REE-alternative alloys and validating their usage as cathode materials in PVD processes. The study employs data-driven alloy design tools to explore extensive compositional spaces and identify REE-lean or REE-free chemistries with desirable thermodynamic stability and mechanical properties. Machine learning models are utilized to predict composition-property relationships, optimize multi-element alloy systems, and screen candidates for manufacturability and compatibility with PVD processes. Selected machine-learning-designed alloys can be fabricated into cathode targets and tested for their ability to produce dense, adherent metalloceramic coatings with high hardness, controlled microstructure, and robust corrosion and erosion resistance. By integrating predictive modeling with experimental validation, this work establishes a pathway toward REE-independent cathode targets for next-generation PVD coatings, supporting long-term material resilience and sustainable innovation in high-performance engineering applications.