Uncovering Mechanisms in Closed Loop Molecular Learning

This research project focuses on the understanding of how intense, shaped ultrafast laser pulses can be used as photonic reagents for controlling basic chemical reactions such as unimolecular dissociation. The experiments make use of a learning algorithm to discover optimal shaped laser pulses for controlling molecular fragmentation in situations where it is impossible to calculate optimal pulse shapes a priori. By combining the optimal pulse shapes with ab initio molecular structure calculations and carefully chosen studies of the molecular fragmentation as a function of pulse shape, it is possible to uncover the basic physical mechanisms underlying control. The ultimate goals of the efforts are to use the understanding of control mechanisms in these experiments to develop learning control into a useful spectroscopic tool for hard to access regions of molecular phase space, and to improve the range and degree of control such that useful applications of control (such as discriminating between different chemical species and controlling bimolecular reaction chemistry) are viable. The broader impact of the program involves student training at both the graduate and undergraduate levels as well as outreach to the local high school. This project is jointly funded by the Division of Physics and the Division of Chemistry in the Mathematical and Physical Sciences Directorate.