Pentalanthanum zinc diplumbide, La5Zn1−xPb2+x (x ≃ 0.6)

The title non-stoichiometric pentalanthanum zinc diplumbide, La5Zn1−xPb2+x (x ≃ 0.6), was prepared from the elements in an evacuated silica ampoule. It adopts the Nb5Sn2Si-type structure (space group I4/mcm, Pearson symbol tI32), a ternary ordered superstructure of the W5Si3 type. Among the four independent crystallographic positions, three are fully occupied by La (Wyckoff 16k), La (4b), and Pb (8h) and one is occupied by a statistical mixture [occupancy ratio 0.394 (12):0.606 (12)] of Zn and Pb (4a). The structure is constructed by face-sharing 10-vertex polyhedra around the unmixed Pb sites. These fragments enclose channels of trans-face-sharing tetragonal antiprisms occupied by the disordered Zn and Pb sites.

The coordination polyhedron of the La1 atoms is a 15-vertices polyhedron (CN=15). The nearest neighbours of La2 atoms form 14-vertices polyhedra with CN=14 (see Fig. 1a). The 10-vertices polyhedra around Pb3 can be seen as the major building block of the structure. The face-sharing 10-vertices polyhedra form a three-dimensional framework encasing channels of face-sharing tetragonal antiprisms which in turn accommodate Zn4 and Pb4 with C·N. 10 in a statistical way (see Fig. 1 et al., 1986;Nowak et al., 1991) with the additional implementation of two algorithms, the electron localization function (ELF) (Becke & Edgecombe, 1990) and -for quantifying nearest-neighbour bonding interactions -crystal orbital Hamilton population (COHP) (Dronskowski & Blöchl, 1993) analyses. Results arising from these analyses were put in relation to selected crystallographic data of La 5 Zn 2 Sn (Oshchapovsky et al., 2011a), namely a differential electronic density map obtained from diffraction data.
In all above mentioned compounds bonding interactions evolve similarly. Atoms of the rare-earth elements usually form metallic bonds with a certain ionic component as they donate part of their electrons to other atoms. The atoms of d-and p-metals, namely zinc and tin, accept these electrons used for covalent bonds among each other. This outcome can be easily explained by the higher electronegativity of zinc and tin compared with rare-earth and alkaline metals (Lange, 1999). Judged on the basis of integrated COHP, (iCOHP) values, R-{Zn,Sn} bonds are significantly weaker than {Zn,Sn}-{Zn,Sn} bonds. Latter mostly show significant bond length contraction compared to the next contacts in the elemental metals. The differential electronic density map of the structure of La 5 Zn 2 Sn, obtained from experimental diffraction data, indicate donation of electrons to zinc and tin atoms in the similar way as in other compounds, for which electronic structure calculations were performed.
The results of electronic structure calculations for (I) are given in Fig. 2 and 3. As an algorithm for calculating electronic structures for substances with mixed-occupied sites is not implemented in TB-LMTO-ASA package, a fictitious, crystallographically ordered phase was introduced by assuming the statistically occupied Zn/Pb site solely occupied by Zn without change of other crystallographic data.
Similar to R-Zn binary and R-Zn-Sn ternary compounds, large values of the electron localization function (ELF) around Pb3 and Zn4 (originally mixed-occupied) are found. In accordance with similar results for previously cited, structurally related compounds, these findings are interpreted as resulting from donation of electrons from lanthanum supplementary materials sup-2 Acta Cryst. (2014). E70, i2-i3 atoms to zinc and lead atoms (see Fig. 2a,b).
The values of ELF around Pb3 atoms are significantly larger than around Zn4 atoms, which indicate that bonds between Pb3 and their neighbours should be significantly stronger than bonds between Zn4 atoms and their neighbours. The calculated iCOHP values enable qualitative estimation of energies of two-center bonds and confirm previous assumption.
Negative values of iCOHP together with bond lengths and bond contractions are given in Table. The DOS plot indicates metallic conductivity for the ordered model compound (see Fig. 3a) with small dip at -1 eV relative to Fermi level. Based on a rigid band approximation it can be suggested that these features don't change qualitatively for the title compound. A small dip in the DOS plot near Fermi level could be indirect proof of donation of electrons from lanthanum atoms towards zinc and lead atoms.
The graph negative iCOHP values versus. distance contraction shown in Fig. 3 b substantiates the assumption that the polyhedra around Pb3 are indeed the building blocks of the structure if seen as condensed into the given threedimensional framework by sharing faces. Apparently, this interpretation is not only useful from a crystallographic point of view but also supported by the hierarchy of bonding interactions in the solid: La-Pb bonds are stronger and show larger distance contractions than most La-La bonds. An exception is the particularly strong La2 IX -La2 X bond enhancing the stability of the framework by connecting adjacent building blocks.

Experimental
A small irregular grey single-crystal of the title compound of suitable quality was isolated from an alloy with composition La 7 ZnPb 2 prepared in the course of a systematic investigation of the ternary La-Zn-Pb system. The preparation process was according to that described for La 5 Zn 2 Sn (Oshchapovsky et al., 2011a).
The sample was prepared by melting pieces of pure metals (99.9% La, 99.999% Zn, 99.99% Pb) in an evacuated silica ampoule in a resistance furnace with subsequent annealing at 600°C for 30 days. Final phase analysis revealed the presence of the title compound in samples with composition La 7 ZnPb 2 together with other lanthanum-rich ternary alloys.
The latter samples, however, had not reached completely the liquid state, since the use of silica ampoules constraints the maximum temperature to 900°C. This limitation may hamper complete equilibration of the sample.

Refinement
The fractional site occupancies of Pb4 and Zn4 were constrained to sum to unity.      Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.