The surface electronic structure of Ce in the \ensuremath{\alpha} and \ensuremath{\gamma} phase has been calculated using a film linearized-muffin-tin-orbitals method. The bonding in Ce is found to be mainly metallic in character. The width of the 4f band in \ensuremath{\alpha}-Ce is found to be about 0.7 eV both in the bulk and at the surface; in \ensuremath{\gamma}-Ce it is slightly narrower, about 0.6 eV. The calculated work function of \ensuremath{\alpha}-Ce is in good agreement with experimental data. Cross sections for the bremsstrahlung isocromat (BIS) processes have been calculated in a fully relativistic framework, and good agreement with experiment is obtained for the itinerant peak in the BIS data near ${\mathit{E}}_{\mathit{F}}$. We find our results consistent with an itinerant f-electron picture for \ensuremath{\alpha}-Ce, and thereby consistent with a picture for the \ensuremath{\gamma}\ensuremath{\rightarrow}\ensuremath{\alpha} transition that is from localized (and magnetic) to itinerant (bonding and nonmagnetic) f-electron behavior, i.e., a Mott transition, as opposed to a Kondo volume collapse picture involving a transition between localized states of the f electrons. We also predict that the surface of \ensuremath{\alpha}-cerium is \ensuremath{\gamma}-like.
The surface electronic structure of Ce in the alpha and gamma-phase has been calculated using' a film linearized-muffin-tin-orbitals method. The bonding in Ce is found to be mainly metallic in character. The width of the 4f band in alpha-Ce is found to be about 0.7 eV both in the bulk and at the surface; in gamma-Ce it is slightly narrower, about 0.6 eV. The calculated work function of alpha-Ce is in good agreement with experimental data. Cross sections for the bremsstrahlung isochromat (BIS) processes have been calculated in a fully relativistic framework, and good agreement with experiment is obtained for the itinerant peak in the BIS data near E(F). We find our results consistent with an itinerant f-electron picture for alpha-Ce, and thereby consistent with a picture for the gamma --> alpha-transition that is from localized (and magnetic) to itinerant (bonding and nonmagnetic) f-electron behavior, i.e., a Mott transition, as opposed to a Kondo volume collapse picture involving a transition between localized states of the f electrons. We also predict that the surface of alpha-cerium is gamma-like.
The electronic structure of NiUSn was obtained using a spin-polarized scalar relativistic linear muffin-tin orbital (LMTO) method. The self-consistent calculations were performed at the observed experimental lattice constant for the cubic MgAgAs crystal structure. Because of the open nature of this crystal structure the LMTO combined correction terms as well as additional empty spheres were used. Like transition-metal analogs (NiMnSb, for example) NiUSn is found to be a half-metallic magnet, with a semiconducting gap in the minority-spin bands and metallic behavior for the majority-spin bands. The calculated magnetic moment (1.99${\ensuremath{\mu}}_{B}$) was found to be about (2/3 of the experimental high-temperature value (3.08${\ensuremath{\mu}}_{B}$).
The first fully relativistic calculations that also include spin polarization are presented for the single-site t matrix for Ce and Pu. The pattern of resonant energies for Ce are compared to recent atomic eigenvalue results for the ${\mathrm{Ce}}^{3+}$ ion. The Ce band states are also determined along the 〈100〉 and 〈111〉 directions in the Brillouin zone by solving the Dirac equation with an internal magnetic field constructed from scalar relativistic linear muffin-tin orbital (LMTO) spin-polarized calculations. In the relativistic case the coupling between spin-orbit and spin-polarization interactions destroys rotational symmetry, unlike in the nonrelativistic spin-polarized formalism, and hence leads to completely nondegenerate f and d states in these systems (similar to the Zeeman effect). To see this we also give results for an applied magnetic field that only couples to the spins and hence resembles an effective internal spin-polarized field.