Metal-carbonyl clusters skeletal bonding electrons

The relationship between boranes and metal-carbonyl clusters can be extended by considering the compound Fe5(CO)i5C, which has the square-based pyramidal structure shown in Fig. 13, with the carbide carbon atom just below the center of the Fe square, clearly contributing all its valence shell electrons to the cluster 24). The metal-carbonyl residue FeB(CO)i4 formally left by removal of this carbon as has the nido structure appropriate for a cluster with 5 skeletal atoms and seven skeletal bond pairs. 

It should be stressed that, in this treatment of metal-carbonyl clusters, the number of nonbonding electron pairs allocated to each metal atom [3 pairs for each ruthenium atom of H2Ru8(CO)i8 2 pairs for each rhodium atom of Rhe(CO)i6] is not arbitrary but is chosen with two objectives in mind (o) to reduce the number of electrons formally remaining for skeletal bonding to fewer than the number of orbitals remaining, because only then is it realistic to assume that all these electrons can be accommodated in bonding MO s and (6) to provide a suitable number of electron pairs on each metal atom for metal carbon... 

The foregoing examples show the relevance to metal-carbonyl cluster chemistry of the borane-oarborane structural and bonding pattern. Its relevance to other areas of chemistry may be explored readily using a systematic skeletal electron-counting procedure (161, 201). 

K. Wade, The Structural Significance of the Number of Skeletal Bonding Electron-pairs in Carboranes, the Higher Borane Anions, and Various Transition-metal Carbonyl Cluster Compounds, Chem. Comm. 1971, 792-793. 

Let us now ask how we could predict the correct total electron count, as just defined, for a stable cluster of known structure (i.e., closo, nido, or arachno). To do this for metal carbonyl clusters, it is postulated that in addition to the electrons necessary for skeletal bonding each metal atom will also have 12 nonskeletal electrons. The basis for this assumption is that in the pyramidal M(CO)3 unit each M—CO bond will comprise two formally carbon tr electrons that are donated to the metal atom and two formally metal it electrons that backbond, at least partially, to the CO ligand. Thus, in predicting the total electron count for a closo polyhedral cluster of n vertices, the result would be 12n + 2 n + 1). Similarly, for nido and arachno clusters that are derived from an n-vertex polyhedron (their parent polyhedron) by removal of one or two vertices, respectively, there will be 12 and 24 fewer total electrons, respectively. 

Many metal carbonyl clusters have interstitial atoms or groups located in the eenter of the polyhedron. Such interstitial atoms may be a light atom sueh as boron, carbon, or nitrogen a post-transition element such as germanium, tin, or antimony or a transition metal. Interstitial atoms most frequently provide all of their valence electrons as skeletal electrons since all of their valence orbitals are neeessarily internal orbitals because of the location of the interstitial atom in the center of the polyhedron. Exceptions to this rule may occur when some of the valence electrons of the interstitial atom occupy orbitals of symmetries which cannot mix with any of the molecular orbitals arising from the polyhedral skeletal bonding. 

Wade, K. (1971) The stmctural significance of the number of skeletal bonding electron-pairs in carboranes, the higher boranes and borane anions, and various transition-metal carbonyl cluster compounds. J. Chem. Soc., D Chem. Commun., (15), 792-793. 

If the square pyramidal metal carbonyl carbides Fe5(CO)i5C ° and Os5(CO)i5C are treated in a similar manner to I xyi ( ()) i T that is, as clusters in which all four of the core carbon atom s valence shell electrons are used for skeletal bonding, then they are seen to have the expected nido shapes of systems with five skeletal atoms (the metal atoms) held together by seven skeletal bond pairs. By contrast, if these carbide carbon atoms had occupied polyhedral vertex sites, with a lone pair of electrons in an exo-oriented sp hybrid orbital, then the number of skeletal bond pairs would have been reduced by one and the number of skeletal atoms would have increased by one. The five metal atoms and the carbide carbon atom would have had to be accommodated in some way on a trigonal bipyramidal skeleton. Clearly, the assumption that all four valence shell electrons from the carbide carbon atom are involved in the skeletal bonding is vindicated. 

In 1977 we reported a method based on graph theory for study of the skeletal bonding topology in polyhedral boranes, carboranes, and metal clusters Q). Subsequent work has shown this method to be very effective In relating electron count to cluster shape for diverse metal clusters using a minimum of computation. Discrete metal clusters treated effectively by this method Include post-transition metal clusters (, ) > osmium carbonyl clusters (O, gold clusters, platinum carbonyl clusters (J., 7 ) > and... 

The different bonding mode of ligands to a metal core, usually accompanied by a change in the electron contribution to the cluster electron count, seems to be one of the major causes of skeletal isomerism. These ligands are mainly organic, sulfur-donor, or carbonyl ligands. Some other causes of skeletal isomerism are the bonding versatility of AuPRs, the unsaturation of 16e Pt centers, or the mere presence of an heterometal. 

---