Multi-AI agent framework reveals the "oxide gatekeeper" in aluminum nanoparticle oxidation.

Multi-AI agent framework uncovers the "oxide gatekeeper" mechanism regulating aluminum nanoparticle oxidation, revealing temperature-dependent dual modes of reactivity.

• Why it matters: Understanding atomic-level oxidation processes in aluminum nanoparticles is crucial for optimizing their use as energy-dense fuels, but existing computational methods face scale and fidelity limitations.
• What they did: Researchers developed a human-audited, multiagent workflow to create and validate a machine learning potential that combines quantum accuracy with scalability, enabling detailed simulations of ANP oxidation at various temperatures.
• The result: Simulations show a temperature-regulated dual-mode oxidation: a "breathing mode" with nanochannels at moderate temperatures and a "rupture mode" causing shell failure at higher temperatures, with aluminum cation diffusion dominating mass transfer by 2 to 3 orders of magnitude.