A game starts with the empty graph on $n$ vertices, and two player alternate adding edges to the graph. Only moves which do not create a triangle are valid. The game ends when a maximal triangle-free graph is reached. The goal of one player is to end
the game as soon as possible, while the other player is trying to prolong the game. With optimal play, the length of the game (number of edges played) is called the $K_3$ game saturation number. In this paper we prove an upper bound for this number.
We present a new method to compute the electronic structure of correlated materials combining the hybrid functional method with the dynamical mean-field theory. As a test example of the method we study cerium sesquioxide, a strongly correlated Mott-b
and insulator. The hybrid functional part improves the magnitude of the pd-band gap which is underestimated in the standard approximations to density functional theory while the dynamical mean-field theory part splits the 4f-electron spectra into a lower and an upper Hubbard band.
We present ab initio calculations of the evolution of anisotropic magnetoresistance (AMR) in Ni nanocontacts from the ballistic to the tunnel regime. We find an extraordinary enhancement of AMR, compared to bulk, in two scenarios. In systems without
localized states, like chemically pure break junctions, large AMR only occurs if the orbital polarization of the current is large, regardless of the anisotropy of the density of states. In systems that display localized states close to the Fermi energy, like a single electron transistor with ferromagnetic electrodes, large AMR is related to the variation of the Fermi energy as a function of the magnetization direction.
It is shown how unbound electron wave functions can be expanded in a suitably chosen localized basis sets for any desired range of energies. In particular, we focus on the use of gaussian basis sets, commonly used in first-principles codes. The possi
ble usefulness of these basis sets in a first-principles description of field emission or scanning tunneling microscopy at large bias is illustrated by studying a simpler related phenomenon: The lifetime of an electron in a H atom subjected to a strong electric field.
