Tuesday Publication Post! 📖 Understanding how nanoparticles move in complex liquid environments is key to uncovering the molecular-level forces that drive assembly, reactivity, and transport. Using liquid-phase TEM with the #PoseidonAX holder, authors Zain Shabeeb, Naisargi Goyal, Pagnaa Attah Nantogmah and Vida Jamali directly visualized nanoparticle dynamics in their native environment!
To interpret these trajectories, the authors from #GeorgiaTech introduced LEONARDO, a deep generative model that combines a physics-informed loss function with an attention-based transformer architecture. This allows the model to learn the stochastic motion of nanoparticles even when no closed-form Langevin-style equation exists.
What can LEONARDO do?
💧 Reconstructs nanoparticle motion consistent with LPTEM observations
📈 Captures statistical signatures of heterogeneous, viscoelastic, and environmentally responsive liquid cell conditions
🔬 Provides a new route to connect observed motion to underlying interaction forces
Want to read the entire work? Find it here!
https://www.doi.org/10.1038/s41467-025-61632-1
