Annu. Rev. Mater. Sci. 1985.15:1-42 About three quarters of all elements are metals or semimetals, which means more than half of all binary mixtures exist in the metallic state. As the binary and ternary systems of metals usually do not contain fewer compounds than those of nonmetals, one would guess that the chemistry of inter metallics would have at least the same significance in research and chemistry textbooks as that of compounds containing nonmetals. However, intermetallic chemistry remained beyond the scope of most chemists, even when intermetallic compounds had gained tremendous technical and economic importance. Their characterization was mainly performed by physicists and metallurgists, who concentrated their efforts on describing those properties essential to the technological use of these materials. The chemical characterization of these compounds is still incomplete. Our knowledge of the number of compounds possible in the systems, their chemical properties, and most important, the bonding mechanism in these phases is far less developed than in other chemical fields. This huge part of chemistry is-with few exceptions (2}-mostly summarized in a very general way, based on the classical concept of the metallic bond (1, 1a). There is no differentiation in classes of compounds, as has long been accepted in other fields of chemistry. This situation is 1 0084--6600/85/0801-0001$02.00 Annu. Rev. Mater. Sci. 1985.15:1-42 2 SCHAFER demonstrated best by the fact that, according to several authors, the word "compound" should not be used for intermetallic phases (3). These scientists feel that stoichiometric sharp compounds do not exist in these systems at all, and that some range of composition is always absolutely necessary to define a metallic phase. Chemically based valence rules, used in other classes of chemical compounds, were thought to be impossible (3). In contrast with this view, which mainly originated among the physical and metallurgical scientists, and which is surely valid for some families of i ntermeta lH cs, t h e ChemI st Ed uard Z intl set f orth a model based on chemical valences for at least some classes of intermetallics. Ionic and covalent bonds in intermetallic phases were a matter of course for Zintl, who believed in the cognition, especially valid for chemistry, that natura non facit saltus.Eduard Zintl was born" on January 21, 1898, in Weiden-in-der-Oberpfalz, Bavaria. He took his PhD at the University of Munich in 1923, as a co worker of Professor O. Honigschmid, who is famous for the accurate determination of the atomic weights of several elements. Two years later Zintl received the venia legendi for chemistry, as assistant of the chemical laboratories of the Bavarian Academy of Science. At that time he published the determination of several atomic weights, and analytical developments concerning potentiometric titrations. He left Munich in 1928 and, as a professor in Freiburg, began his work on intermetallic compounds. He was very successful in this field. With many publications to ...