Reactions Changing the Metal Atom Composition

Such reactions, starting from a cluster and ending with a cluster of higher nuclearity, must be mechanistically complex. Normally several ligands have to be removed and several metal-metal bonds formed. Complete mechanistic information is not yet available for any case. Therefore, the material will be organized according to reagent types. 

The key step in a cluster expansion reaction is the attachment of the incoming metal unit. Once this has taken place, a sequence of metal-metal bond formations accompanied by ligand eliminations can occur which is the reversal of the cluster unfolding reactions described in Section III,C. In uncontrolled cluster expansions, the first step is the combination of coor-dinatively unsaturated cluster and monometallic units, and the reaction is unlikely to stop at this stage. Under mild conditions the attachment may result from a nucleophile/electrophile combination, the products of which have been isolable in a few cases (see below). More insight into possible 

Both HRe(CO)s and H2Os(CO)4 can be oxidatively added to Os3(CO),, (NCMe) (126 -128). This leads to external attachment of the new metal carbonyl unit as in 64 (127), and a second HRe(CO)s molecule can be incorporated the same way (126). In both cases just one metal-metal bond has been formed in the first step. CO elimination from 64 introduces one more metal - metal bond, one possible result of which is rhomboidal 65 (126), whereas further CO elimination under H2 leads to full aggregation to tetrahedral 66 (127). All three steps of a M3 + M capping sequence have thereby been performed. A similar two-step sequence leads from Os6(CO)17(NCMe) and H2Os(CO)4 via H2Os7(CO)21 to H2Os7(CO)20 (128). 

Anionic clusters are good nucleophiles (see Section III,A) and are often easy to make. On the other hand, the electrophilic nature of most monometallic complexes is obvious from ligand substitutions. The combination of these properties makes a strategy for cluster expansion. This strategy was used for the first time by Hieber (130) in making Fe4(CO)fc from Fe3(CO),7 and Fe(CO)s. It is probably active in many syntheses of large metal carbonyl clusters because the Re, Os, Rh, Ir, Ni, and Pt clusters involved are almost always anionic. However, simple stoichiometries can rarely be written for such reactions (122). This route makes mixed metal clusters accessible, e.g., 

Only two kinds of reagent have been extensively used for this type of reaction so far. These are the anionic carbido carbonyl clusters of iron and 

---