I will show how to compute rigorously and within minutes at most some equilibrium solvation properties of a solute. We will focus on two key properties: (i) the solvation free energy of a solute, that is, its chemical potential, and (ii) the solvation profile, that is, where solvent molecules (dis)like to stay in the solvation shells of the solute.
This seminar will be about the molecular density functional theory (MDFT), a theory and a high performance software developed these 10 last years at École Normale Supérieure [Borgis, Belloni, Levesque]. It will not be a 40 minute talk around statistical mechanics.
The goal of this seminar will be to initiate a discussion about the constructive interference we may find between real world problems and a tool crafted from theory and, since very recently, is being shaped for real-world problems. I would also appreciate feedback on the quantities and precision you may expect from a useful tool.
I will first introduce the MDFT with its advantages (fast, rigorous, systematically improvable) and drawbacks (it’s a theory, and we do approximations).
Then, I will show benchmarks of MDFT versus experiments and versus the state of the art theoretical way of computing the properties discussed above: molecular dynamics with thermodynamic integration.
Finally, I will show how our most recent developments combining MDFT with artificial intelligence unlocked a prediction capability far better than any other method, on our test-set and to our knowledge.
[Borgis] Density functional theory of solvation and its relation to implicit solvent models, by Ramirez and Borgis, J. Phys. Chem. B 109, 6754 (2005)
[Belloni] Efficient molecular density functional theory using generalized spherical harmonics expansions, by Ding, Levesque, Borgis and Belloni, J. Chem. Phys. 147, 094107 (2017)
[Levesque] Bridge Functional for the Molecular Density Functional Theory with Consistent Pressure and Surface Tension and its Importance for Solvation in Water, by Gageat, Belloni, Borgis and Levesque, arXiv:1709:10139 (2018)