Several concepts of information theory (IT) are extended to cover the complex probability amplitudes (wave functions) of molecular quantum mechanics. The classical and non-classical aspects of the electronic structure are revealed by the electronic probability and phase distributions, respectively. The information terms due to the probability and current distributions are accounted for in the complementary Shannon and Fisher measures of the resultant information content of quantum states. Similar generalization of the information-distance descriptors is also established. The superposition principle (SP) of quantum mechanics, which introduces the conditional probabilities between quantum states, is used to generate a network of quantum communications in molecules, and to identify the non-additive contributions to physical and information quantities. The phase-relations in two-orbital model are explored. The orbital communication theory of the chemical bond introduces the entropic bond multiplicities and their partition into IT covalent/ionic components. The conditional probabilities between atomic orbitals, propagated via the network of the occupied molecular orbitals, which define the bond system and orbital communications in molecules, are generated from the bond-projected SP. In the one-determinantal representation of the molecular ground state the communication amplitudes are then related to elements of the charge and bond-order matrix. Molecular equilibria are reexamined and parallelism between the vertical (density-constrained) energy or entropy/information principles of IT and the corresponding thermodynamic criteria is emphasized.