"For the success of the StrongLight project we worked mainly in two complementary research lines:
1) Electronic description of the Low Temperature and Hight Temperature phases of the (MeBr-dcnqi)2Cu molecular crystal:
Using a state-of-the-art level of theory such as the use of periodic-DFT at PBE including van der Waals interactions, for the proper description of the pi-system, and Hubbard-U method to include strong-correlations effects due to the on-site Coulomb repulsion, we determined the optimized geometry and the band structure for both phases. We have shown the need of a surprisingly high Hubbard-U value for the π-system in order to reproduce the experimentally observed charge separated phase at low temperature. Besides we determined the possible vibrational changes involved in the initial steps of the PIPT, being the most relevant the change in the copper coordination environment. A scientific paper is under preparation for its publication.
2) Scientific code and methodology developing:
We developed a new method based on a combination of the a local density analysis and a real-time propagation density functional theory (p-TDDFT) at the linear response regime that enable to determine the electronic excitation properties such as the transition dipole moments and transition densities. Besides, we proofed that this method is useful to compute the exciton coupling in complex systems. This method and its implementation in the OCTOPUS code have been published in peer-review journal. In addition, we have worked (and still going on) on a new way to determine molecular excitation spectra using machine learning techniques such as Kernel Ridge Regression and Deep Neural Networks.
Besides within the frame of another collaboration with Berkeley University, we have provide a mid-temperature synthesis of backbone-substituted polycylic aromatic hydrocarbons (PAH). Our DFT calculations demonstrated the plausible reaction mechanisms proving an energy barrier of less than 0.6 eV, in good agreement with the experimental values. The results of this grateful collaboration will soon published in a high impact peer-review journal.
During the reported period of the project, the Researcher has exploited its result by the following ways:
Papers:
Four peer-reviewed scientific publications: 2 published and 2 submitted. Besides, 3 more research papers are in preparation and we plan to submit them during the 2020 year.
Contributions in two conferences:
Oral contribution to the ""Theoretical Chemistry and Computational Modeling: 20 years promoting Excellence in Science (20TCCM)"" , May 30, 2019 to June 01, 2019, Donostia-San Sebastián (SPAIN). Talk title: ""Transition Density Formulation from TDDFT Enables Exciton Coupling Calculations in Large Systems”.
Poster contribution to ""HANDS-ON DFT AND BEYOND: HIGH-THROUGHPUT SCREENING AND BIG-DATA ANALYTICS, TOWARDS EXASCALE COMPUTATIONAL MATERIALS SCIENCE” August 26th to September 6th, 2019, Barcelona, Spain. Poster title: ""Predicting molecular conformers using Machine Learning""
Codes:
All codes developed by the Researcher are under the GNU General Public License, and are publicly available at:
- Octopus Code :
https://gitlab.com/octopus-code/octopus(s’ouvre dans une nouvelle fenêtre)- QuantumLight - The Game :https://gitlab.com/qjornet/quantumlight-game (still in preparation)
Besides, all calculations performed during the reported period have been or will be uploaded to the ab-initio calculation repository of NOMAD (NOvel MAterial Discovery Lab)."
