First-principles calculation of electric-field gradients

It has been shown recently for YD 3 (Zogal et al., Phys. Rev. B 64, 214110 (2001)) - a compound that has attracted much interest, because of the metal-semicoductor transition it undergoes at ambient conditions - that the comparison of accurate electric-field gradients (EFGs) calculated on the basis of the density-functional theory with experimental values makes it possible to distinguish between different structure models which show the same neutron-diffraction patterns. In the proposed project further hydrides/deuterides and some intermetallic compounds shall be investigated in a fashion similar to the one that has proved successful for YD 3 . Also spin-lattice relaxation rates and Knight shifts shall be calculated and compared to results from NMR measurements. While EFGs are directly related to the electron density, spin-lattice relaxation rates and Knight shifts depend mainly on the partial densities of states at the Fermi level which are thus also very sensible to the local environment of the respective atoms. The analysis of the theoretical results will provide useful information which cannot be obtained from the NMR experiments alone, namely on the relationship between the EFG components and chemical bonding and on the exact microscopic mechanism for the spin-lattice relaxation process. The required experimental results are in some cases already available or will be provided by Prof. Olgierd Zogal and Dr. Bogdan Nowak (Academy of Sciences in Wroclaw, Poland). Further collaborators will be Dr. Walter Wolf (Materials Design, Le Mans, France) and Dr. Peter Vajda (École polytechnique, Palaiseau, France). A fruitful collaboration with these persons has been established a few years ago which shall be continued and intensified during the period of this project.

Through investigations of solids by means of nuclear magnetic resonance, data (spin-lattice relaxation times, electric-field gradients) are obtained which provide valuable information on details of the electronic structure and the crystal structure, which otherwise cannot be acquired easily or which are completely inaccessible. By such a procedure the validity of structure models can be assessed, even if the structure determination is not uniquely possible from X-ray or neutron diffraction measurements. Moreover, in many cases the existence of theoretical results facilitates the evaluation of NMR spectra and yields information which can experimentally only be obtained for single crystals but not for powder samples. In this project a series of compounds (yttrium and lanthanum hydrides, lithium nickel nitride, borides of scandium, yttrium, zirconium and lutetium of different stoichiometries as well as yttrium aluminium borate) has been investigated using ab-initio methods. At the same time the corresponding experiments have been performed by various cooperation partners. The results of the investigations can be summarized as follows. In the borides the covalent bonds are determined by strong B-B interactions within the two- or three-dimensional B framework. For the majority of the investigated borides excellent agreement between the ab-initio calculations and the experiments was obtained. For scandium dodecaboride, however, differences have been observed which can either be caused by structural distortions or by the presence of iron in the sample. Therefore, new computational and experimental work will be required for this compound which is being performed at present. For LiNiN, which can be considered as a "one-dimensional" metal and which has extraordinary properties with respect to ionic and electronic conductivity, the combination of experimental and theoretical approaches provided information both concerning the bonding properties and the Li- ion diffusion which is important with regard to practical applications. As far as YAl 3 (BO3 ) 4 is concerned, which is a candidate for UV-visible lasers, the ab-initio calculations and the experimental work have led to consistent results after some difficulties in the preparation of single crystals had been overcome. For the hydrides good agreement between theory and experiment has been reached for YH 2 , whereas for lanthanum trihydride structure optimizations have been performed as a first step towards a successful interpretation of the NMR spectra.