Concurrent Pathways in the Phase Transitions of Alloys and Oxides: Towards an Unified Vision of Inorganic Solids - Inorganic 3D Structures - page 157

6.1 Description of BaCl 2 and Its Relation to the CoSn-Type
Structure

The description of the hexagonal structure of BaCl 2 (Fe 2 P type) is not easy. Hyde
and Andersson [ 57 ] described it on the basis of the so-called TCTPs, like those
existing in both the cotunnite- and the Ni 2 In-type structures. These prisms are
formed by six Cl (Fe) atoms and are centred by Ba and P atoms, respectively.
Three additional Cl atoms cap the lateral faces of the prisms, giving rise to a
CN 9 for the central Ba (P) atoms (Fig. 14b ).

The structure can also be described by considering only the Ba(1) and Cl(2)
atoms. In this case, the novelty is that we obtain a partial skeleton of composition
Ba 2 Cl 3 , formed by interconnected BaCl 6 trigonal prisms. This subarray is repre-
sented in Fig. 14c and shows that the prisms are sharing the edges perpendicular to
the projection plane, forming distorted hexagonal tunnels that run parallel to the c
axis. These tunnels are centred by the Ba(2) atoms.
Contrarily, when only the remaining Ba(2) and Cl(1) atoms are taken into
account, columns of trigonal prisms of composition BaCl 3 are formed (Fig. 14d ) .
They are located at the origin of the cell, filling the hexagonal tunnels of Fig. 14c .
Thus, BaCl 2 can be regarded as the sum of two substructures, Ba 2 Cl 3 and BaCl 3 ,
which are complementary, forming jointly the structure of Fig. 14a . The stoichiom-
etry can then be written as Ba 2 Cl 3 þ

¼

BaCl 3 , equivalent to BaCl 2 .
If the Ba(2) atoms, at (0, 0, 0), centering the hexagonal tunnels, are included into
the Ba 2 Cl 3 subarray of Fig. 14c , the partial formula becomes Ba 3 Cl 3 . This new
skeleton, drawn in Fig. 14e , is in fact a distortion of the hexagonal RT phase of NiIn
(CoSn type), represented in Fig. 14f . By comparing both figures, one sees that the
distortion consists of a slight deviation of the Cl atoms from the line connecting the
Ba atoms. Recall that Ba and Cl atoms are not coplanar.
When the Cl atoms are connected by red lines (Fig. 14e ), one obtains planar nets,
which are a distortion of the regular 3 · 6 · 3 · 6 nets (Kagom´ nets) forming the NiIn
(CoSn)-type structure (Fig. 14f ). It should be reminded that such Kagom´ nets are
also forming part of the cubic Laves phases (Cu subarray in MgCu 2 ) and, hence, are
also present in the cation array of spinels. In BaCl 2 ,Fe 2 P and CoSn, the Kagom ´
nets are eclipsed ( ... AAA ... sequence), whereas in MgCu 2 they are stacked in a ...

ABCABC sequence, perpendicular to [1 1 1].

The interest of this analysis of the BaCl 2 structure resides in that it allows for a
simple and rational explanation of the two skeletons in which we have partitioned the
BaCl 2 structure. Once more, the explanation conforms to the

...

[ 14 , 33 ] . Thus,
starting with the substructure Ba 3 Cl 3 , composed of Ba(1), Ba(2) and Cl(1) (Fig. 14e )
and assuming that each Ba atom transfers one electron to the three remaining Cl(2)
atoms (the green atoms drawn in Fig. 14d ) , we would obtain the pseudo-formula
(

EZKC

-CsCl) that
adopts the structure of the CoSn (NiIn) type filled with three C -Ar atoms (Fig. 14f ).
This interpretation could be regarded as artificial because this structure has
not been found for any alkali halide. For this reason, an alternative explanation

C -Ar) 3 ( C -CsCl) 3 , which is a pseudo-compound of stoichiometry 1:1 ( C