Journal of Alloys and Compounds 460 (2008) 138–141 Neutron powder diffraction and solid-state deuterium NMR studies of Ca 2 RuD 6 and the stability of transition metal hexahydride salts Ralph O. Moyer Jr. a , Sytle M. Antao b , Brian H. Toby b , Frederick G. Morin c , Denis F.R. Gilson c,∗ a Department of Chemistry, Trinity College, Hartford, CT 06106-3100, United States b Advanced Photon Source, Argonne National Laboratory, Argonne, IL 60439-4856, United States c Department of Chemistry, McGill University, Montreal, Que. H3A 2K6, Canada Received 12 April 2007; received in revised form 11 May 2007; accepted 22 May 2007 Available online 26 May 2007 Abstract The crystal structure of Ca 2 RuD 6 has been determined by neutron powder diffraction: space group Fm3m,K 2 PtCl 6 structure, as found for other hexahydride salts of group 8 metals with alkaline earth or lanthanide counter ions. No structural phase transition was observed between 340 K and 50 K. The deuterium nuclear quadrupole coupling constant, 54.7 kHz, leads to an ionic character of the Ru–D bond of 76%. The known trends in the behaviour of A 2 MH 6 salts are interpreted in terms of the ionization energies of the cation and the central metal atom. © 2007 Elsevier B.V. All rights reserved. Keywords: Ternary metal hydrides; Neutron diffraction; Deuterium NMR; Stability 1. Introduction The ternary transition metal hydrides/deuterides (TMH/Ds) of the alkaline earth and lanthanide salts of iron, ruthenium and osmium hydrides, A 2 MH 6 , exhibit interesting structural and spectroscopic trends. The unit cell dimensions depend upon the ionic radii of the alkali metal and the metal–deuteride bond distances in the Fe, Ru, Os group and increase in the order Fe < Ru < Os, as expected. They also depend markedly on the counter-ion, with Mg < Ca < Sr < Ba for all salts in the series [1]. The complete structure of Ca 2 RuD 6 has not been previ- ously reported and is the subject of the present report, together with a deuterium solid-state NMR study of Ca 2 RuD 6 . Kritikos and Nor´ eus [2] have shown that the infrared active metal–hydrogen stretching modes depend on the unit cell dimen- sion and attribute this to the size of the counter-ion, with the electropositive character of the group 2 ion of only secondary importance. However, Parker et al. [3] in a study of alkali metal salts of [PtH 6 ] 2- concluded that the alkali metal ion reduced the charge on the platinum and hydrogen, thus stabilizing the com- ∗ Corresponding author. E-mail address: denis.gilson@mcgill.ca (D.F.R. Gilson). plex. With this mechanism, the electropositive character of the counter-ion assumes much greater importance. Thus the inter- relationships between the bond lengths, unit cell dimensions and vibrational frequencies and the role of the electropositivity require attention and we suggest a more quantitative approach to these questions. 2. Experimental Ca 2 RuD 6 was prepared by the previously reported method [4] and the purity was confirmed by X-ray powder diffraction and infrared and Raman spectroscopy. Neutron powder diffraction data were collected using the 32 detector BT- 1 neutron powder diffractometer at the National Institute of Standards and Technology (NIST) Center for Neutron Research reactor, NBSR. The neutron diffraction data were modeled using the Rietveld method, as implemented in the GSAS and EXPGUI programs [5,6]. For the initial room temperature cubic structure, the starting atomic coordinates, cell parameters, isotropic displace- ment parameters, and space group, Fm3m, were taken from Moyer and Toby [7]. The Ca atom occurs on the 8c Wyckoff site, Ru at 4a, and D at 24e. The refined atomic coordinates were then used as input for the next structure at the different temperatures. The background was modeled using a 12-coefficient Chebyschev polynomial function. The reflection-peak profiles were fitted with three Gaussian coefficients (GU, GV, and GW). Individual isotropic displacement parameters were also refined. A full-matrix least-squares refinement with 4 structural param- eters, one lattice constant and 17 experimental parameters (12 background terms, 3 profile parameters, a scale factor, and a zero shift) were allowed to vary simulta- 0925-8388/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2007.05.078