Important Research Achievement by the School of Chemistry and Chemical Engineering, SEU, Published in NANO LETTERS

Recently, the research team led by Prof. Sun Yueming and Prof. Dai Yunqian from the School of Chemistry and Chemical Engineering, SEU, has made significant progress in the field of the stability of nano-catalytic materials. The research, titled “Dynamic Stabilization of Ultrafine Pt Nanoparticles against Sintering: Insights from Machine Learning”, innovatively proposes a “dynamic confinement” strategy. By deeply integrating in-situ electron microscopy characterization and AI analysis, the team visually and quantitatively predicted the anti-sintering behavior of catalytic materials at extremely high temperatures, providing a brand-new theoretical framework and design paradigm for the development of practical nano-catalysts with both high activity and ultra-long service life. This achievement was published as a cover paper in the international journal Nano Letters.

In high-temperature heterogeneous catalysis, the agglomeration and deactivation of active components of catalysts due to sintering are a core bottleneck restricting their industrial application. Particularly for ultrafine platinum (Pt) nanoparticles smaller than 3 nm, their extremely high surface energy significantly reduces the Tammann temperature, allowing metal atoms to migrate at relatively low temperatures and causing rapid particle growth via the particle migration and coalescence (PMC) or Ostwald ripening (OR) mechanisms. Therefore, achieving a fundamental breakthrough that reconciles high activity with thermodynamic stability is a long-standing  major challenge in the field.

To address this challenge, the research team developed a dynamic confinement strategy. They anchored ultrafine Pt nanoparticles (<3 nm) onto a porous iron oxide (Fe₂O₃) support and enhanced the overall thermal stability using a Na₂Ti₃O₇ nanowire framework. Real-time observation via in-situ transmission electron microscopy revealed for the first time that at an extreme temperature of up to 850 °C, Pt nanoparticles remain mobile within the channels and edge regions of the support but are effectively confined within specific nanospaces. This achieves the unique state of “mobile but non-agglomerated” at the atomic scale, strongly suppressing sintering.

To deeply analyze and quantify this complex dynamic process, the team constructed an artificial neural network model. The model successfully established a quantitative structure-activity relationship between the particle microenvironment (e.g., distance from pores/edges), temperature, and particle size evolution, with predictions highly consistent with experimental data (R² > 0.92). The model clearly indicates that pore confinement and edge confinement are the key structural factors endowing nanoparticles with exceptional anti-sintering ability, advancing catalyst stability research from phenomenological observation to a predictable and quantitative new stage.

Verified under harsh conditions close to practical applications, the catalyst exhibits extraordinary performance. Using carbon monoxide oxidation as a model reaction, the catalyst aged at 500 °C achieves complete CO conversion at 150 °C. More notably, it shows no activity decay after continuous operation for 600 hours at 175 °C. Even in accelerated deactivation tests with increased reaction space velocity, its activity decay rate constant remains extremely low, demonstrating outstanding long-term operational stability. This provides an effective solution to the industrial problem of rapid catalyst sintering and deactivation caused by local overheating in strongly exothermic reactions.

Through a full-chain research paradigm of “rational material design – in-situ microscopic mechanism – machine learning prediction – application performance validation”, this work not only created a Pt-based catalyst with ultimate anti-sintering performance but also deeply revealed the microscopic mechanism of dynamic confinement in stabilizing nanoparticles, promoting the deep integration of artificial intelligence in catalytic material design and failure analysis.

Tang Mingyu, a PhD candidate from the School of Chemistry and Chemical Engineering, SEU, is the first author of the paper, and Prof. Dai Yunqian is the corresponding authorwithSEU as  the sole affiliation. This research was supported by the National Key R&D Program of China, the National Natural Science Foundation of China, and other projects.

Paper link: