Mainz 2026 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

A: Fachverband Atomphysik

A 41: Interaction with Strong or Short Laser Pulses II

A 41.3: Talk

Friday, March 6, 2026, 11:45–12:00, N 2

Deep learning for the retrieval of internuclear motion from photoelectron momentum distributions based on full quantum dynamics — •Nikolay Shvetsov-Shilovski and Manfred Lein — Leibniz Universität Hannover

We retrieve the time-dependent internuclear distance of a dissociating one-dimensional molecule using a neural network that takes photoelectron momentum distributions (PMDs) generated by strong-field ionization as input. The PMDs are calculated by solving the time-dependent Schrödinger equation, with nuclear motion treated fully quantum mechanically. We find that a neural network trained on distributions with fixed bond lengths can recover the time-varying internuclear distance with an absolute error of 0.3 a.u. This opens new perspectives for applying machine learning to real-time visualization of molecular dynamics.

Keywords: strong-field ionization; deep learning; photoelectron momentum distributions; molecular dynamics; real-time visualization