Title: Accelerating Chemical Discovery: Computational Design of Molecules and Materials

Federico Zahariev

Ames National Laboratory Of US Dept of Energy

Iowa State University

Host: Dr. Mark Gordon

Abstract: Computational chemistry is transforming our ability to solve complex molecular problems. In this seminar, I will highlight my contributions to three key areas: First, I will discuss the development of a time-dependent density functional theory (TDDFT) method that enables accurate modeling of excited-state properties in solutions, with applications in photochemistry and spectroscopy. Next, I will showcase how machine learning models, trained on ab initio and experimental data, predict stability constants for rare-earth ligands 100× faster than traditional methods, directly guiding synthetic efforts. Furthermore, I will explore how quantum-classical computing, despite current limitations to small molecules, offers exceptional potential for high accuracy in describing systems with strong electron correlations—a significant challenge for classical computational approaches. These methodologies bridge theory and experiment, accelerating chemistry and materials science discoveries.