Origin of the size-dependence of the polarizability per atom in heterogeneous clusters: The case of AlP clusters

Abstract

An analysis of the atomic polarizabilities α in stoichiometric aluminum phosphide clusters, computed at the MP2 and density functional theory (DFT) levels, the latter using the B3LYP functional, and partitioned using the classic and iterative versions of the Hirshfeld method, is presented. Two sets of clusters are examined: the ground-state Al(n)P(n) clusters (n=2-9) and the prolate clusters (Al(2)P(2))(N) and (Al(3)P(3))(N) (N≤6). In the ground-state clusters, the mean polarizability per atom, i.e., α/2n, decreases with the cluster size but shows peaks at n=5 and at n=7. We demonstrate that these peaks can be explained by a large polarizability of the Al atoms and by a low polarizability of the P atoms in Al(5)P(5) and Al(7)P(7) due to the presence of homopolar bonds in these clusters. We show indeed that the polarizability of an atom within an Al(n)P(n) cluster depends on the cluster size and the heteropolarity of the bonds it forms within the cluster, i.e., on the charges of the atoms. The polarizabilities of the fragments Al(2)P(2) and Al(3)P(3) in the prolate clusters were found to depend mainly on their location within the cluster. Finally, we show that the iterative Hirshfeld method is more suitable than the classic Hirshfeld method for describing the atomic polarizabilities and the atomic charges in clusters with heteropolar bonds, although both versions of the Hirshfeld method lead to similar conclusions.

Quantum chemical mass spectrometry:ab initioprediction of electron ionization mass spectra and identification of new fragmentation pathways

“…Harmonic frequency calculations were performed to verify that the resulting structures are minima on the potential energy surface. Bond orders were calculated using the Fractional Occupation Hirshfeld Iterative (FOHI) formalism, which is based on the Hirshfeld partitioning of the electron density . Transition‐state energies were calculated using the STQN method (QST3) implemented in Gaussian 09.…”

Hirshfeld‐based intrinsic polarizability density representations as a tool to analyze molecular polarizability

“…[25] Using the latter family of partitionings, some of the present authors have widely explored the monomer interaction of various kinds of clusters formed by water [26] together with monovalent anions and sodium cation, [27] methanol, [28] hydrated methane sulfonic acid, [29] and aluminum-based clusters. [30,31] Moreover, they have extended the study to higher polarizability contributions such as the linear dipole-dipole, dipole-quadrupole, and quadrupole-quadrupole polarizabilities. Thus, they have proposed a partitioning of these multipole polarizabilities into atomic contributions to study the accuracy of the interaction energies and the transferability of functional groups for linear alkanes, water clusters, and small organic molecules.…”

Importance of anisotropy in the evaluation of dispersion interactions

“…We have validated the Hirshfeld method for partitioning dipole polarizabilities on a broad range of systems [32][33][34][35], including the evaluation of the isotropic dispersion coefficients and the anisotropic C 6 coefficient [9,12]. In order to evaluate the anisotropic C 7 and C 8 coefficients, we now introduce the definitions of the intrinsic quadrupole-dipole and quadrupole polarizability tensors:…”