Complex Shape - Not Just Any Which Way

J0rgensen s chain theory for some cobalt(III) complexes (top), compared with the Werner representations (bottom). Werner s representation for [CoC13(NH3)3] fits the nonelectrolyte behaviour seen experimentally, unlike the Jprgensen model. 

We now know conclusively that Werner was correct, or at least nearly so. Interestingly, although his octahedral shape is supremely dominant in six-coordination, we now know that there are trigonal prismatic six-coordinate shapes also - albeit with ligand systems to which Werner had no access. What sophisticated modern physical methods tell us conclusively is that coordination complexes don t just happen upon a shape. Each different coordination number known, and these vary from 2 to 14, supports a limited number of basic shapes. The shape, or stereochemistry, depends both on the metal and its oxidation state and on the form of the ligand - each contributes in different ways. 

Distributions predicted for from two to six point charges on a spherical surface (left), and shapes of complexes evolving from this point-charge model, when a central atom is placed at the core of the sphere and considered to bond to each donor atom located where a point charge occurs. 

---