PREDICT teamPhotoinduced Reactive Dynamics in Computational Theory

Overview

The PREDICT team develops physics-based and data-driven multiscale models to predict laser-induced transformations in complex materials. Our research aims to bridge the gap between theoretical descriptions and predictive simulations across a range of fields, including laser–matter interaction, material nano- and micro-structuring, and out-of-equilibrium dynamics. We focus on linking ultrafast energy deposition to long-time structural, chemical, optical and magnetic evolution by coupling electronic excitation, thermal transport, atomistic dynamics and reactive processes within unified modeling frameworks.

This predictive approach enables quantitative understanding and control of phenomena such as laser-induced oxidation in semiconductors, nanoparticle formation in liquids, phase transitions, plasmon-mediated chemical reactions, and the evolution of optical and magnetic properties of nanostructured materials. The team also develops advanced numerical methods and optimization strategies for inverse process design.

Applications include photonic materials processing, laser-assisted catalysis, light-emitting nanostructures, and the synthesis of functional nanomaterials for biomedical and energy-related technologies.

Research topics include multi-scale and multi-physical simulations in following areas:

- Ultrafast laser–matter interaction
- Multiscale modeling of nanostructures
- Nanoparticles
- Plasmonic and magnetic structures: laser synthesis and properties
- Laser-induced oxidation and phase transitions
- AI-based optimization of material processing
-etc.

Expertise

The team has strong expertise in multi-physics and multi-scale modeling, in particular in electromagnetic fields and properties simulations, in classical and reactive molecular dynamics, two-temperature and nano-thermal nonequilibrium models (TTM / nTTM), hybrid continuum–atomistic coupling strategies, DFT and TD-DFT, optical and thermal transport modeling, and machine-learning-assisted parameter inference for inverse process design, etc.