Much new solid-state chemistry and many new structures are to be found in the little explored region of intermetallic phases involving, in part, the late transition and the early post-transition elements, i.e., groups 10-13. In addition, combinations that include an early active metal M exhibit additional bonding features because of their somewhat polar, salt-like nature. Such added coulomb forces give us a new and important component to structure and bonding in systems that are already metallic. For example, exploratory syntheses of such phases containing a ) an active metal M (Sr, Ba), b ) indium (In) or thallium (Tl), and, sometimes c ), a third earlier metal E, have revealed a rich variety of new compounds and numerous new or derivative structure types. More importantly, there are two virtually new features to these: first, the surprisingly persistent covalent network bonding among In and E atoms (when present) and the second, which requires that M be tightly encapsulated within this because of the inherent charge (electronegativity) differences between M and the somewhat negative or anionic network; basically, the importance of coulomb's law in stability.

We now also understand some things about the major structural changes that often ensue because of even small changes in atom sizes, as illustrated above. The 3D structure of BaIn4, as small portion of which is shown in 1 (upper left), is evidently the most efficient way to pack/bond these atoms and charges. Note the hexagonal prisms that, in part, surround each Ba (blue). Accordingly, substitution of the smaller Sr (by 0.16 ?, 10% in radius) requires as smaller cage in order to maintain reasonable Sr-In distances and to avoid "rattling" of Sr. This is best achieved in the lower symmetry structure of SrIn4 shown in 2 (upper right), Note that the Sr is now encaged in pentagonal, not hexagonal, prisms as shown in this projection nearly along the pseudo 5-fold axis. Our chemical studies of substitution reactions have now revealed two new ways to recover the tetragonal structure type in 1, both by shrinking the In network to better match Sr. One result in 3 , SrAuIn3 (lower left), is achieved by substitution of 50 % Au for In in one type of network site. (This is clearly a relativistic effect with Au since it occurs in the next heavier period.) Finally, 4 (SrMg1.5In2.5, lower right) shows how the network may also be shrunk by substitution of the naturally smaller Mg for most of the In in the other network site. Both substitutions involve electronic structure effects in the parent lattice as well. The last result in 4 has in fact suggested how we might also create new hydrogen storage substrates from lighter polar intermetallic phases.