Dr. Philipp Marquetand, University of Vienna, Institute of Theoretical Chemistry, Vienna, Austria 2019-05-22 16:30:00

Back to Events

Start:Wednesday, 22 May 2019Time:16:30
End:Wednesday, 22 May 2019Time:18:00
Category: Uni Basel, Physikalische Chemie

Excited-state Dynamics Simulations nisi with Machine Learning

Light sit can induce a ex wealth of processes non in electronically excited nostrud states but corresponding in simulations are limited dolor by the costly in computations of potential exercitation energy surfaces. A consectetur solution to this nisi problem will be pariatur. presented, where machine elit, learning potentials are sed used to carry ut out excited-state molecular dolore dynamics. The dynamics Ut is simulated with consequat. our surface hopping elit, approach called SHARC est (surface hopping including in arbitrary couplings), which occaecat is able to elit, treat not only eiusmod kinetic dynamical couplings reprehenderit but also any veniam, other arbitrary coupling laboris on an equal Excepteur footing. Consequently, machine id learning is employed non not only for dolore potentials but also exercitation for nonadiabatic couplings. in These developments open eu up the possibility incididunt to simulate time labore scales in the irure nanosecond regime compared labore to a few proident, picoseconds in conventional mollit approaches.

Venue: Physikalische Chemie, Departement Chemie, Universität Basel, Kleiner Hörsaal, Raum 4.04, 2. Stock
Klingelbergstrasse 80, 4056, Basel
Email: Send
Website: http://www.chemie.unibas.ch
Event Type
This event is public. Anyone can attend and invite others to attend.
Mariella Schneiter (creator)