No less than 14 PhDs have begun their work and training in a European network on computational spectroscopy with strong DTU Chemistry participation.
The following article is from the DTU Chemistry Annual Report 2018. Read the full report here.
As synchrotron radiation sources, free electron lasers, and other advanced facilities become available, there is a growing need for computational chemistry.
“It is simply not possible to interpret results from advanced spectroscopy intuitively. You will need corresponding models and software tools,” explains Professor Sonia Coriani, DTU Chemistry. She heads a key work package in a new European training network (ETN) for computational spectroscopy. With 14 new PhD students hired, the network is set to shape the digital chemists of tomorrow.
“Education and training of students or early stage researchers in this field is not part of any standard curriculum, neither in chemistry nor physics. This is in clear contrast to the increasing importance of computational spectroscopy,” says Sonia Coriani.
Of the 14 young researchers, two – Torsha Moitra and Daniil Fedotov – are employed at DTU Chemistry with Sonia Coriani as their supervisor and Klaus B. Møller, Professor in Physical Chemistry at the department, as their co-supervisor. The two professors are also co-supervising Postdoc Shota Tsuru, employed at the department through DTU’s H.C. Ørsted COFUND programme.
In 2018, the group’s activities included a 3-month stay by Professor Henrik Kock, SNS & NTNU, as Visiting Professor at DTU Chemistry – funded by a grant from the Otto Mønsted foundation.
Model before you measure
The initiative is named COSINE (COmputational Spectroscopy In Natural science and Engineering). Sonia Coriani heads the work package on modeling of advanced spectroscopies while Klaus B. Møller contributes with his experience from computer simulation of chemical dynamics and interpretation of ultrafast X-ray scattering experiments.
“Scattering and spectroscopy are very different techniques, but they are alike in the sense that they both yield indirect information about the probed molecules. This implies that before you do your experiment, you need to have a model. There is just no way of working your way backwards from an experimental result,” says Klaus B. Møller.
Both professors have extensive experience from large-scale facilities abroad. And with the experience of the seven European academic partners and associated industrial partners included, the network covers a wide range of techniques. Examples are Near-Edge X-ray Absorption Fine Structure for ground and excited states, Resonance Raman Optical Activity, Resonance Inelastic X-ray Scattering, and Photo-Electron Spectroscopy.
Importantly, the network also offers specific training on programming and use of High-Performance Scientific Computing resources, both locally and through the PDC Center for High Performance Computing of Kungliga Tekniska Högskolan in Stockholm.
Predict the propertiesof new molecules
Both in her own research and in the COSINE project, Sonia Coriani focuses on excited electronic states.
“User-friendly software packages for ground-state chemistry already exists, but similar solutions for excited states are lagging behind,” she points out.
Over the last two decades it has become possible to model molecular ground-state properties on the computer with high accuracy. This enables chemists to predict the properties of new molecules virtually. Thereby a huge number of molecules can be pre-screened on computers prior to synthesis, avoiding costs. It is also easily tested whether a proposed change in the structure of a molecule, a particular substitution for example, can be expected to give the desired effect. This approach to design of molecules with specific properties – like molecules with low optical band gaps, high or low electron affinities or ionization potentials – is gaining momentum in both academia and industry.
“The ultimate goal of computational spectroscopy is, similarly, to be able to predict excited-state properties and spectra of real-life molecular species in gas and condensed phases, and to be able to study light-triggered reactions on the computer,” says Sonia Coriani.