No abstract

No abstract

It is shown that generalized gradient approximations (GGAs) for exchange only, due to their very limited form, quite generally can not simultaneously reproduce both the asymptotic forms of the exchange energy density and the exchange potential of finite systems. Furthermore, mechanisms making GGAs formally approach at least one of these asymptotic forms do not improve the corresponding quantity in the relevant part of the asymptotic regime of atoms. By constructing a GGA which leads to superior atomic exchange energies compared to all GGAs heretofore but does not reproduce the asymptotic form of the exact exchange energy density it is demonstrated that this property is not important for obtaining extremely accurate atomic exchange energies. We conclude that GGAs by their very concept are not suited to reproduce these asymptotic properties of finite systems. As a byproduct of our discussion we present a particularly simple and direct proof of the well known asymptotic structure of the exchange potential of finite spherical systems.

No abstract

Density functional theory calculations on Ybusing the Hartree Fock system and the analogous Dirac-Hartree-Fock methods predict that Ybis stable in the Py,z state. The electron affinity obtained after the correlation energy functionals used are calibrated against Caand Lu is 2f 1 mHartree (54+ 27 meV).The calculations indicate that the Pilz state is unlikely to be bound due to spin-orbit coupling reducing its binding energy.

Recently, most of the lanthaoide anions have been reported to be stable (Ganvan erno including Tm'. We present skong theoretical evidence, based on a Dinr-Hamee-Fock density functional lheory (DHFDFT) study of the 70 electmn system, that Tmis stable in the electron configuration lXe]4f136s26p1 and that it is not stable in the canventionally assumed lXeI4fl46s2 confrguntion, the ground slate of Bo. The EA of Tm-is estimated to be in the range 1 to 5 mliartrees. When this result is combined with previous predictions of a 6p electron being bound in Yb-(Vosko er al) and Lu-(Vosko and Chewy), it is natural to conjecture that lhe other lanthanides bind 6p electrons to form their stable anions.

No abstract

AbstrocL DiraoHartree-Fock theory (DHF). and density functional lheory Hartree-Fock system (DFI-HF) calculations predid that Lu-(2 = 71) is stable in the D(e]4f'4Sd16sz6p' odd-parify electmn configuration and is not b u n d in the commonly assumed D(e]4ft45dz6sz even-parity mnfiguration of H P (2 = 72). Combining the mulls of thex methods gives an electron amnily e 7 f 4 mHartree The Hartree-Fock energy dominates the crossover from even lo odd parity ground state as Z is lowered f " 72 to 71 in this 72 electron system. The relativistic eiiecu, which are inherent to the DHF melhod, are important in unbinding the even-parity state and have k e n included penurbativeiy in the DFPHF alculations. llle very large experimental sd interconfiguration energy of Luo is wed to support the conclusion that the even-parity configuration is unstable. A mmparison ot the electron mnfiguration of singly charged (*) and neutral Lu with those of the group 1115 elemenls indicates that Lu is most similar to Y. * Supported in pan by the Natural Sciences and Engineering Research Council of Canada.