Events Calendar

 
Seminar

"Computational materials design: From multiscale modeling to machine learning" by Dr. Pascal Friederich, KIT and University of Toronto, Canada.

Tuesday, 19 March 2019, 16:00-17:30
KIT, Campus Nord
Institute of Nanotechnology
Bldg. 640 Seminar Room 0-167
Hermann-von-Helmholtz-Platz 1
76344 Eggenstein-Leopoldshafen

Talk given by

Dr. Pascal Friederich
Marie Skłodowska-Curie Fellow
Karlsruhe Institute of Technology & University of Toronto

 

Abstract:

The development of novel technologies is tightly coupled to the discovery and development of new materials with increasingly demanding specifications on function and performance. The chemical space of potential material candidates, both inorganic and organic, is large, widely unexplored and in its entirety inaccessible by standard methods. The established Edisonian process of materials discovery and development is costly and typically requires up to 10 to 20 years from the discovery of a new material to its commercialization. The development of reliable computational methods offers the potential to significantly speed up this process and helps to discover yet unknown parts of the chemical space.

My recent work has focused on the computational design of organic semiconductors for
organic electronic applications. I will present a multi-scale modeling approach which was
used to design novel semiconductor materials, including molecules for charge transport and light emission in OLEDs and charge extraction in perovskite solar cells.​[1] The model
accurately predicts experimental properties of amorphous thin films, including properties of the morphology and molecular orientation as well as electronic properties such as the
charge carrier mobility.​[2,3] We used the model to demonstrate the de novo design of organic semiconductors with improved charge carrier mobility.​[4]

In the second part of the talk I will present my work on machine learning methods, a field that has opened entirely new prospects in materials science and computational materials design.​[5] I propose to combine methods based on machine learning with physics based computational approaches to further improve and accelerate computational materials modeling and design.


[1] Friederich ​et al.​, Advanced Functional Materials 26 (31), 5757-5763 (​2016​)
[2] Friederich ​et al.​, ACS Applied Materials & Interfaces 10 (2), 1881-1887, (​2018​)
[3] Friederich ​et al.​, Chemistry of Materials 29 (21), 9528-9535, (​2017​)
[4] Friederich ​et al.​, Advanced Materials 29 (43), 1703505, (​2017​)
[5] Friederich ​et al.​, Scientific Reports 8, 2559 (​2018​)

 

This event is part of the eventgroup INT Talks
Speaker
Dr. Pascal Friederich

KIT and University of Toronto, Canada
The Matter Lab
Organizer
Prof. Dr. Wolfgang Wenzel
INT
KIT
Karlsruhe
Mail: wolfgang wenzel does-not-exist.kit edu
https://www.int.kit.edu/wenzel.php
Targetgroup
Interested / Everyone
Service-Menu