Machine-learning accelerated nonadiabatic dynamics for organic molecules

Project ID: 2228cd1355 (You will need this ID for your application)

Under Offer

UCL Lead department: Chemistry

Project Summary:

Relevant processes, such as photosynthesis and vision in biological systems, as well as various industrial processes, involve photochemical reactions. In the field of optoelectronics and photonics, understanding the behaviour of excited states is vital for developing and improving devices like lasers, LEDs (light-emitting diodes), and solar cells. In this context, semiclassical nonadiabatic dynamics (NAD) helps elucidate the underlying mechanisms and pathways for these reactions and design efficient materials with desired electronic properties, which is crucial for optimizing and controlling them. Due to the need for propagating multiple trajectories and long simulation times, these calculations are computationally expensive, even for small-size molecular systems. Machine learning (ML)-driven NAD dynamics enable the simulation of many trajectories, ranging from hundreds to thousands, over timescales extending to the nanosecond regime. Despite the recent developments in the field, ML-driven NAD simulations are still in their infancy. Thus, the development of new approaches has the potential to significantly impact our understanding of excited state processes. What Will You Be Doing? This PhD project will focus on the implementation and application of ML algorithms to accelerate NAD in molecular systems, with a particular focus on optimizing these simulations in the condensed phase. These codes will be tested for the dynamics of organic systems with applications in lasers and organic solar cells. Who Will You Be Working With? You will be directly supervised by Dr Rachel Crespo Otero, who is a computational chemist with experience in excited state dynamics and method development. The student will be actively involved in a collaborative project with Northeastern University (US). Who We Are Looking For: This project is for a highly motivated student with a great interest in method development, machine learning, and programming.