Teaching lasers to control molecules in the presence of laboratory field uncertainty and measurement imprecision

An iterative optimization algorithm for designing laser fields to control molecular motion which utilizes laboratory input (test fields) and output (resulting product yields) information is proposed. Laboratory uncertainties such as laser field noise and limited precision in the product yield measurements are included in the simulations of the experiments. Two simulated examples of implementation of the algorithm are presented: selective electronic excitation in a model four‐state system and maximizing dissociation yield of the hydrogen fluoride molecule. Both examples demonstrate that, even with the inclusion of laboratory uncertainties, the experimental learning‐based algorithm is a potentially feasible method of controlling molecular motion and possibly manipulating chemical reactions.