Revealing spectral field features and mechanistic insights by control pulse cleaning

Control pulse cleaning (CPC) is the process of experimentally removing extraneous control field features in a closed-loop quantum dynamics optimization experiment. We demonstrate CPC in ionization processes in the model systems K{sub 2} and NaK by applying evolution strategies during the closed loop shaping of fs pulses. The CPC technique operates by applying genetic pressure on the spectral components with appropriate cost functions, and CPC is investigated in the case of weak (K{sub 2}) and of strong cleaning (NaK). A Pareto-optimal curve is constructed for the latter case, which reveals the correlation of the two generally conflicting goals of meeting the physical objective versus cleaning the control pulse. At weak genetic pressure unnecessary pulse components could be removed with marginal influence on the ionization process, whereas at strong genetic pressure the optimal degree of ionization is affected and, interestingly, electronic transitions to particular vibrational states are exposed. Thus, employing CPC while seeking optimal controls allows for extracting information about the chosen dynamical pathways including the significant intermediate states.