Fast spectroscopic signatures accelerated by deep learning

 

Often, simulations of spectroscopic measurements such as FTIR and Raman are used to assist experimentalists to characterize samples grown in their laboratories. Although this technique works (and is widely used), there can be a high computational cost associated with the spectral simulation. This project will utilize deep neural networks to develop an efficient and accurate methodology for simulating accurate spectroscopic signatures at significantly lowered computation cost than currently possible. This study will focus on 2D materials, beginning with graphene. Collaborator University of Montreal will provide expertise in computational materials, electronic structure and first principles simulations of nanoscale structure and spectra. The NRC will provide expertise on application of AI and deep learning to problems in nanoscience.

Biographies

Dr. Michel Côté

Michel, a professor at the University of Montreal, is an expert in computational materials, electronic structure, and first principles simulations of nanoscale structure and spectra.

Michel is a principal investigator on the fast spectroscopic signatures accelerated by deep learning project.

Find out more about Dr. Côté (French only).

Dr. Isaac Tamblyn

Isaac is a research officer at the National Research Council and is a Vector Institute faculty affiliate. He is also an adjunct professor of physics at the University of Ottawa. Isaac's current research interests are focussed on the application of AI and deep learning to problems in nanoscience, in particular materials and processes related to renewable energy.

Isaac is leading the AI for materials master project and is a principal investigator on several AI for Design collaborator projects.

Find out more about Dr. Tamblyn.

Highly qualified personnel (HQP) biographies

Kevin Ryczko

Kevin is a PhD student at the University of Ottawa and a Postgraduate Affiliate at the Vector Institute for Artificial Intelligence. He is working under the supervision of Dr. Isaac Tamblyn. He specializes in utilizing deep learning techniques to accelerate the discovery and computation of materials.

Find out more about his projects at github.com/kryczko or github.com/CLEANit.

Olivier Malenfant-Thuot

Olivier is a PhD student at Université de Montréal, under the supervision of Prof. Michel Côté. They worked together on a first-principles study of graphene and the incorporation mechanisms of nitrogen atoms in its structure. He is an expert in the use of ab initio codes such as ABINIT and BigDFT. He is now working on the use of deep neural networks to conduct such studies.

His coding projects are available at github.com/OMalenfantThuot.

Arnab Majumdar

Arnab is a postdoctoral fellow who will join the project efforts in the fall of 2020. His previous experience is in electronic structure methods and its applications to various problems such as high-pressure superconductors, geophysical properties, and topological materials. He obtained his PhD at the University of Saskatchewan under the supervision of Profs. John S. Tse and Yansun Yao, and he did his first postdoctoral research with Prof. Rajeev Ahuja at Uppsala University in Sweden.

Find out more about Arnab's publications at google scholar or research gate.