Optimal control of molecular motion for δ-target probability density