Bonding intermetallic solutions

In a study of the hydrolysis of platinum(II)-nitrile complexes to afford the corresponding amides, the first well-characterized examples of platinum(II) complexes that dimerize spontaneously to form weakly bonded Pt-Pt dimers with an unsupported intermetallic bond have been described.166 For example, treatment of cw-[PtCl2(NCBut)2] with excess KOH affords the amide species cw-[PtCl2 HN=C(OH)But 2], which exists as a dimer (49) both in solution and in the solid state, with a Pt-Pt bond length of 3.165(1) A, as determined by X-ray crystallography. Four... 

Bismuthides. Many intermetallic compounds of bismuth with alkali metals and alkaline earth metals have the expected formulas M3Bi and M3Bi2, respectively. These compounds are not salt ike but have high coordination numbers, interatomic distances similar to those found in metals, and metallic electrical conductivities. They dissolve to some extent in molten salts (eg, NaCl—Nal) to form solutions that have been interpreted from cryoscopic data as containing some Bi3 . Both the alkali and alkaline earth metals form another series of alloylike bismuth compounds that become superconducting at low temperatures (Table 1). The MBi compounds are particulady noteworthy as having extremely short bond distances between the alkali metal atoms. 

The circumstances under which intermetallics form were elucidated by the British metallurgist William Hume-Rothery (1899-1968) for compounds between the noble metals and the elements to their right in the periodic table (Hume-Rothery, 1934 Reynolds and Hume-Rothery, 1937). These are now applied to all intermetaUic compounds, in general. The converse to an intermetaUic, a solid solution, is only stable for certain valence-electron count per atom ratios, and with minimal differences in the atomic radii, electronegativities, and crystal structures (bonding preferences) of the pure components. For example, it is a mle-of-thumb that elements with atomic radii differing by more than 15 percent generally have very little solid phase miscibility. 

There is a close relationship between the solution-based syntheses of the naked clusters, as described above, and pure solid state chemistry. This relationship stems from the fact that homonuclear bonding and discrete clusters in many cases also exist within alloys of the type used in the synthesis of Zintl ions in solution. Indeed, such alloys are often to be regarded as well-defined intermetallic compounds. ... 

The y-phase is a solid solution with a face-centered crystal lattice and randomly distributed different species of atoms. By contrast, the y -phase has an ordered crystalline lattice of II2 type (Figure 10.2). In pure intermetallic compound NisAl the atoms of aluminum are placed at the vertices of the cubic cell and form the sublattice A. Atoms of nickel are located at the centers of the faces and form the sublattice B. The y -phase has remarkable properties, in particular, an anomalous dependence of strength on temperature. The y -phase first hardens, up to about 1073 K, and then softens. The interatomic bondings Ni-Al are covalent. 

Formation of solid solutions and replacement of the aluminum atoms decreases anisotropic factor because of occurrence of an additional quantity of collectivized electrons. A part of metallic bond increases. AQ the alloying elements are transition metals, that is, they have an unfilled d shell. These elements add electrons in the d band of the Ni3Al intermetallic compoimd. 

In the discussion of the bond valence method two important problems are skirted. The first, currently a topic of lively debate, is the problem of deciding whether two atoms that are in close proximity in a structure are in fact bonded together. The second is the problem of determining, a priori, the valence of an atom. These problems are circumvented by avoiding situations where their solution is not obvious . The treatment of the bond valence method owes much to the work of Brown [1,2] but also contains some new results. It complements an earlier account [3] of new methods of describing structures. Historically the concept of bond valence derives from the Pauling [4] bond number as applied to metals and intermetallic compounds and subsequently applied to oxides by Bystrom and Wilhelmi [5]. 

---