In this article we discuss the Slater–Koster (SK) tight-binding (TB) method from the perspective of our own developments and applications to this method. We first present an account of our work in constructing TB Hamiltonians and applying them to a variety of calculations which require an accurate representation of the electronic energy bands and density of states. In the second part of the article we present the Naval Research Laboratory TB method, wherein we demonstrate that this elaborate scheme can accurately account for both the band structure and total energy of a given system. The SK parameters generated by this method are transferable to other structures and provide the means for performing computationally demanding calculations of fairly large systems. These calculations, including molecular dynamics, are of comparable accuracy to first-principles calculations and three orders of magnitude faster.

https://www.fis.cinvestav.mx/~rbaquero/SlaterKoster.pdf

The Slater-Koster tight-binding method: a computationally efficient and accurate approach

91Zr nuclear magnetic resonance in tetragonal ZrH2

47Ti and 49Ti nuclear magnetic resonance in TiH2

We report the measurements of nuclear magnetic resonance of 53Cr (for the first time) and magnetic susceptibility in hexagonal chromium hydride to search for electronic structure parameters characterizing this hydride. The Knight shift of 53Cr nuclei equal to 0.53% was constant in the temperature range 150–294 K. The magnetic susceptibility (χ) was measured in the 4.2–300 K temperature range. In contrast to its earlier interpretation, we suggest that the intrinsic paramagnetism is temperature-independent (χ = 226.7 × 10-6 emu/ mole) whereas the temperature-dependent component is due to a magnetic impurity. Both Knight shift and magnetic susceptibility have been partitioned into their spin and orbital contributions. The largest term for both values is the orbital (Van Vleck type) one.

53Cr knight shift and magnetic susceptibility of hexagonal CrH0.93

The electronic (or chemical) binding energy between one hydrogen atom and another hydrogen atom or a substitutional impurity in vanadium and niobium is estimated in a tight-binding formalism. The hydrogen-hydrogen binding energy is found to be considerably repulsive for nearest and next-nearest neighbouring positions. In the case of vanadium-based alloys we have estimated the binding energy between hydrogen and substitutional scandium, titanium, chromium and manganese. For niobium-based alloys we have estimated the binding energy between hydrogen and substitutional yttrium, zirconium and molybdenum. In both vanadium and niobium the binding energy between hydrogen and substitutional impurities located at nearest neighbouring positions is found to be attractive for impurities located to the left of the host metal in the periodic table and changes sign when the impurities are located to the right.

Electronic rearrangement in the dilute vanadium and niobium hydrides

Employing the tight binding coherent potential approximation method, we investigated the electronic structure of the substitutionally disordered transition metal dihydride series Ti1-yNbyHx, Ti1-yVyHx (0 ⩽ y ⩽ 1; 1.5 ⩽ x ⩽ 2.0). These alloys contain disorder on both the metal and hydrogen sublattices. We found that trends in the region of the Fermi level display rigid band-like characteristics for the substitution of niobium or vanadium for titanium, and for the hydrogen disorder as well. The calculations are compared with electronic specific heat measurements and predictions based on nuclear magnetic resonance experiments.

P.M. Laufer

Papers

Calculations of the electronic structure of disordered transition metal dihydrides