The caesium chloride structure

A unit cell of the caesium chloride structure is shown in Figure 1.30. It shows a caesium ion, Cs, at the centre of the cubic unit cell, surrounded by eight chloride... 

The second common structure is the caesium chloride structure. In this structure, ions of one type are located at the corners of a cubic lattice... 

All the alkali metal halides except the cliloride, bromide and iodide of caesium form cubic crystals with the rock salt lattice and show a co-ordination number of 6. The exceptions are also cubic, but have the caesium chloride structure (Fig. 133) characterised by a co-ordination number of 8. The radius ratio for CsCl, Cs /Cl" = 0.93, allows 8 co-ordination, but is so near the ratio for 6 co-ordination that caesium chloride is dimorphous, changing, at 445°, from the caesium chloride to the rock salt structure. The crystalline halides are generally markedly ionic, though, as expected, lithium iodide is somewhat covalent, for iodide is the largest and most easily polarised simple anion and lithium, the smallest alkali metal cation, possesses the strongest polarising power. 

The structure of caesium chloride is included here because, although it is not close packed, it is often confused with, and written as, body centred when it is not. The structure of caesium chloride is shown in Figure 1.17. The chloride ions are on the cube comers and the ion at the centre is a caesium. In Section 1.4 we saw that a body-centred cubic lattice refers to an identical set of points with identical atoms at the comers and at the centre of the cube. This means that the stmcture of caesium chloride is not body-centred cubic. Many alloys, such as brass (copper and zinc) possess the caesium chloride structure. 

Fig. 3.07. Section through a unit cell of the caesium chloride structure on a vertical diagonal plane. The solid circles represent the cations. In (a) the ions are shown in their correct relative sizes for CsCl (b) corresponds to the critical radius ratio for anion-anion contact.

The radius ratio r+jr for each of the alkali halides is shown in table 3.03. Consideration of these values reveals that CsCl, CsBr and Csl would, indeed, be expected to have the caesium chloride structure, and that the majority of the remaining halides would be expected to show the sodium chloride arrangement. There are, to be sure, a number of halides with r+jr > 0 7 which, nevertheless, have the sodium chloride rather than the caesium chloride structure. Fig. 3.08, however, emphasizes that energetically there is little difference between these two structures when the radius ratio is large, and there are in any case other factors contributing to the lattice energy which we have so far ignored in our discussion. 

In the same way, the caesium chloride and sodium chloride structures are related by a simple displacive mechanism. If the caesium chloride structure be suitably distended in a direction parallel to one of the cube diagonals each ion loses two of its neighbours (those along the direction of distortion) while the remaining six assume positions at the comers of a regular octahedron. 

The structures of a number of ammonium salts have already been described. In many of them the NH4+ ion is in free rotation and can be treated as a spherical ion of radius 1 48 A so that the structures observed are determined by the usual geometrical considerations. Thus most of the ammonium halides have either the sodium chloride or the caesium chloride structure. In ammonium fluoride, however, the cation is not in free rotation and the wurtzite structure results. 

Even with the / and y phases the position is not always as simple as in the case just discussed, and in fact it appears that the pattern of sites occupied, and not the actual distribution of the two kinds of atom, is the only significant feature of the structure, for a given phase may appear in different systems at widely different compositions. Thus in the copper-tin system the / phase appears at a composition of about 17 atomic per cent of tin, corresponding approximately to the formula Cu5Sn. Such a composition is clearly not consistent with the caesium chloride structure, but it is found that the phase is actually a body-centred cubic arrangement with the atoms distributed at random in these proportions. 

The caesium-chloride structure can be considered to be derived from the ccp structure by having Cl-ions occupy all the primitive lattice points and octahedral sites, with all tetrahedral sites occupied by Cs+ ions. This is exceedingly difficult to visualize and describe without carefully constructed figures or models. Refer to S.-M. Ho and B. E. Douglas, J. Client. Educ. 46, 208, 1969, for the appropriate diagrams. 

---