Construction of chromophore orbitals

A method is presented in which the delocalized orbitals of a large molecule containing one or more distinct chromophores are transformed to give orbitals which possess spatial properties similar to those of small isolated molecules which resemble each chromophore. The purpose of the method is to provide orbitals which, when employed in configuration interaction or other orbital‐based calculations, permits excited state wave functions to be characterized in terms of local chromophore and/or charge‐transfer excitations. The method is applicable to any set of orbitals given as a basis set expansion, obtained from either ab initio or zero differential overlap procedures. The effectiveness of the method is demonstrated with applications to a glycine dipeptide.