Advanced force fields in multiscale approaches to computational spectroscopy
Location: CECAM-FR-MOSER, Sorbonne Université Campus Pierre et Marie Curie, 4 place Jussieu, 75005, Paris
Organisers
Simulation approaches to spectroscopy (UV-visible, IR, Raman, sum frequency generation spectroscopy, 2D UV, 3D IR-UV,…) are extremely valuable to help interpret experimental spectra, assign bands and rationalize observed shifts. Typically, one tries to simulate the experimentally measured spectrum, and then analyses the simulation to bring additional molecular insight. However, computing spectra with sufficient accuracy has proved very challenging. In condensed phase, it is crucial to take into account the environment and its fluctuations, which implicates not only larger system sizes, but also longer sampling times, and has triggered the development of force field-based approaches, or mixed classical/quantum strategies.
For electronic spectroscopy, one needs at the same time to use quantum methods to obtain an accurate description of the ground and excited states energies, and to properly sample the geometries in both states. We thus need tools for the parameterization of intra-molecular and inter-molecular force fields from quantum mechanical data for flexible structures in both the ground state and the excited states, where more general force fields failed to reproduce the potential energy surface explored by the molecule. JOYCE [1-4] or other approaches to set up force fields [5,6] are used to sample the excited energy surface to take into account the solvation broadening and study phenomena involving excited states.
For vibrational spectroscopy, advanced polarizable force fields (e.g. AMOEBA) are used to obtain the dipole and polarizability fluctuations required for spectra calculations. Polarizable forces fields are used for biomolecular simulations [7], but also for the study of infrared spectra [8], solid states [9], charge transfer states, transitions states [10]. An alternative is offered by frequency maps, that are trained on quantum data, and allow to compute spectra from simulations using standard non polarizable force fields [11]. Very recently, several Machine Learning-enhanced approaches have been suggested [12-14].
Young researchers: Young researchers and students are particularly encouraged to participate and to present their results with posters. Thanks to our sponsor, we have some funding for rooms for PhD student and Postdoc student. To apply for this housing opportunity, please write a motivation and insert your CV when applying.
References
Carine CLAVAGUERA (CNRS - Paris Saclay University) - Organiser
Elise Duboué-Dijon (CNRS, Laboratoire de Biochimie Théorique) - Organiser
Simon Huppert (Sorbonne Université) - Organiser
Isabelle Navizet (Univ Gustave Eiffel) - Organiser