We make use of the Quantum Theory of Atoms in Molecules (QTAM) to partition the total energy of a many-electron system into intra- and interatomic terms, by explicitly computing both the one- and two-electron contributions. While the general scheme is formally equivalent to that by Bader et al., we focus on the separation and computation of the atomic self-energies and all the interaction terms. The partition is ultimately performed within the density matrices, in analogy with McWeeny's Theory of Electronic Separability, and then carried onto the energy. It is intimately linked with the atomistic picture of the chemical bond, not only allowing the separation of different two-body contributions (point-charge-like, multipolar, total Coulomb, exchange, correlation, ...) to the interaction between a pair of atoms but also including an effective many-body contribution to the binding (self-energy, formally one-body) due to the deformation of the atoms within the many-electron system as compared to the free atoms. Many qualitative ideas about the chemical bond can be quantified using this scheme.

No abstract

An exact energy partition method based on a physically sound decomposition of the nondiagonal first-order and diagonal second-order density matrices put forward by Li and Parr (J. Chem. Phys. 1986, 84, 1704) is presented. The method splits the total energy into intra- and interatomic components and is applicable on quite general wave functions. To explore it numerically, the energy components of three test molecules (H2, N2, and LiH) have been computed using four different partitions of the charge density ρ(r) into atomic densities. Several aspects on the chemical bond and the relative importance of different components of the binding energy are analyzed. The merits of different partitions of ρ(r) are also discussed.

Evidence that the bond paths of the quantum theory of atoms-in-molecules (QTAIM) signal preferred quantum-mechanical exchange channels is presented. We show how bond paths between an atom A and the atoms B in its environment appear to be determined by competition among the A-B exchange-correlation energies that always contribute to stabilize the A-B interactions. These pairwise additive stabilizations depend neither on the attractive or repulsive nature of the classical electrostatic interaction between the atoms' charge densities, nor on the change in the self energies of the atoms involved. These other terms may well cause an overall molecular-energy increase in spite of a possibly large A-B exchange-correlation stabilization. After our proposal, bond paths, both at and out of equilibrium geometries, are endowed with a specific energetic meaning that should contribute to reconcile the orthodox QTAIM interpretation with other widely accepted views, and to settle recent controversies questioning the meaning of hydrogen-hydrogen bonding and the nature of the so-called "steric interactions", the role of bond paths in endohedral complexes, and the generality of the results provided by the QTAIM. Implications for the nature of more general closed-shell interactions are also briefly discussed.

No abstract