The aim of the project is to use advanced Machine Learning techniques to predict the anharmonic vibrational spectra of large  Polycyclic Aromatic Hydrocarbon (PAH) molecules. PAHs with size larger than 40 carbon atoms are widespread in astronomical environments and are responsible for a family of vibrational emission features dominating the spectra of entire galaxies. Anharmonic spectra for PAHs can be calculated using 2nd order Vibrational Perturbation Theory on Density Functional Theory derived Quartic Force Fields but are computationally very expensive when considering the size of PAHs relevant for astronomy and the large number of possible isomers. This is where machine learning techniques can help.

• You will collect and homogenize the available anharmonic DFT calculations of PAHs available in the team

• You will implement neural networks based on PAH molecular structure to predict their anharmonic spectrum

• You will test the accuracy of topological descriptors and neural network  for the task comparing with available experimental results

• You will conduct research, communicate results  via peer-review publication and at conferences

• You will be a proactive member of the research group and  have the opportunity to co-advise research projects of MSc and BSc students

• You will contribute to creating an open, safe and inclusive working environment

• A Ph.D. in chemistry, physics, astronomy, computer science or a related field completed upon your arrival

• A committed researcher, demonstrated by previous research experiences, publications and presentations at international conferences

• Experience with using neural networks and topological descriptors for molecules or materials 

• Excellent communication and writing skills in English

• The ability to work both independently and collaboratively in a multidisciplinary environment.