18-Electron rule limitations

This mode of calculation has been called the EAN rule (effective atomic number rule). It is valid for arbitrary metal clusters (closo and others) if the number of electrons is sufficient to assign one electron pair for every M-M connecting line between adjacent atoms, and if the octet rule or the 18-electron rule is fulfilled for main group elements or for transition group elements, respectively. The number of bonds b calculated in this way is a limiting value the number of polyhedron edges in the cluster can be greater than or equal to b, but never smaller. If it is equal, the cluster is electron precise. 

With metal clusters it is even harder than in other fields of inorganic chemistry to substantiate theoretical results by energy measurements. Only two such measurements have come to the attention of the author — the photoelectron spectrum of [CpFe(C0)]4 370) andbond energy determinations in 03(00)9CX-compounds 187). However, a considerable number of papers deal with metal-metal bonding in, and the symmetry properties of, clusters as related to their stoichiometry and their electron count. These studies have confirmed the wide apphcability of the simple 18-electron rule in predicting metal-metal bonds and structures, but they have also led to an understanding of the limits of this rule for clusters with more than four metal atoms. 

Structure C has 10 electrons surrounding nitrogen, but the octet rule limits nitrogen to 8 electrons. Structure C is incorrect. 

Although very reliable for Os3 clusters, the 18-electron rule is of limited use for larger ones. In the case of Os3 clusters with added transition metal atoms (Section VIII) one cannot simply apply the rule assuming that each short M—M contact involves a two-electron bond, but despite that the rule works well in most cases. 

The 18-electron rule can be a useful guide to stable organometallic compounds, especially when p-acceptor ligands are present, although it has the limitations referred to in Topic H9. Compounds 3, 5 and 6 obey this rule, but 1 without p bonding ligands has an electron count of only 12. Metallocenes [M(h5 C5H5)2 ] are known for the 3d series elements V-Ni, with 15-20 valence... 

Limitations of the 8- and 18-electron rule localized electron-deficient compounds... 

Limitations of the eight- and 18-electron rule delocalized bonding... 

In the chapters that follow you will find numerous exercises in counting electrons for clusters - elaborations of the eight- and 18-electron rules for these complex structures. The same factors that cause the eight- and 18-electron rules to fail will similarly limit cluster-counting rules based upon them. Like these fundamental rules, even when satisfied, the cluster-counting rules yield no detailed information on electronic structure. Hence, the bolder student occasionally asks, Why count by which he or she means Of what real value are these counting exercises if little is learned about where the electrons really are ... 

This complex has an electron count matching a nido structure, but it adapts the butterfly structure expected for arachno. This is one of the many examples in which the structure of metal clusters is not predicted accurately by Wade s rules. Limitations of Wade s rules are discussed in R. N. Grimes, Metal-lacarboranes and Metallaboranes, in G. Wilkinson, F. G. A. Stone, and W. Abel, eds.. Comprehensive Organometallic Chemistry, Vol. 1, Pergamon Press, Elmsford, NY, 1982, p. 473. 

Both of these reactions involve use of a strong reducing agent. In recent years, reduction of carbonyl complexes has been pushed to its limit with the synthesis of highly reduced anions such as [Mn(CO)4p, [Cr(CO)4] , [V(CO)5] , and [Ti(CO)5]. 2 t Since [MnlCOl ]", Cr(CO), and [V(CO)J were known to be relatively stable, there was some expectation, based on the 18-electron rule, that it would be possible to synthesize [Ti(CO)5] , even though Ti(CO)y is unknown. Often the expectation that a product should exist is a long way from synthetic success. The involved synthesis of [Ti(CO)(J illustrates the point.25 The overall reaction is... 

There is a special kind of site-selectivity which has been called periselectivity. When a conjugated system enters into a reaction, a cycloaddition for example, the whole of the conjugated array of electrons may be mobilized, or a large part of them, or only a small part of them. The Woodward-Hoffmann rules limit the total number of electrons (to 6, 10, 14 etc. in all-suprafacial reactions, for example), but they do not tell us which of 6 or 10 electrons would be preferred if both were feasible. Thus in the reaction of cyclopentadiene (355) and tropone (356), mentioned at the beginning of this book, there is a possibility of a Diels-Alder reaction, leading to 354, but, in fact, an equally allowed, ten-electron reaction is actually observed,121 namely the one leading to the adduct (357). The product is probably not thermodynamically much preferred to the... 

It has come to the point where some chemists, faced with a compound that does not exhibit a normal ll NMR spectrum, discard it and move on to a more promising project. In effect, our addiction to NMR spectroscopy has allowed the 18-electron rule to become a set of blinders limiting our view. 

---