Hypovalent bonding

Note that despite the superficial resemblance to the H-bridging geometry of diborane (B2H6, Section 4.4), the optimal NBO description of AI2CI6 reflects a quite different bonding pattern, with no appreciable three-center hypovalent bond character in the latter case. Thus, AI2CI6 provides a fairly direct intramolecular comparison between two-center cr bonds of distinct covalent versus coordinate character, with the latter bonds responsible for the unusual four-membered ring motif. 

The NBO donor-acceptor theory of hypovalency three-center bridge bonds... 

Soon after the quantum revolution of the mid 1920s, Linus Pauling and John C. Slater expanded Lewis s localized electronic-structural concepts with the introduction of directed covalency in which bond directionality was achieved by the hybridization of atomic orbitals.1 For normal and hypovalent molecules, Pauling and Slater proposed that sp" hybrid orbitals are involved in forming shared-electron-pair bonds. Time has proven this proposal to be remarkably robust, as has been demonstrated by many examples in Chapter 3. 

Although orbital hybridizations and molecular shapes for hypovalent metal hydrides of the early transition metals and the normal-valent later transition metals are similar, the M—H bonds of the early metals are distinctly more polar. For example, metal-atom natural charges for YH3 (+1.70), HfH4 (+1.75), and TaHs (+1.23) are all significantly more positive than those (ranging from +0.352 to —0.178) for the homoleptic hydrides from groups 6-10. Indeed, the empirical chemistry of early transition-metal hydrides commonly reveals greater hydricity than does that of the later transition-metal hydrides. 

Figure 4.12 Metal hydride bond (ctMh) and antibond (ctMh+) NBOs for hypovalent M = Hf, Ta.

Table 4.15 summarizes optimized bond lengths and NBO Lewis-like structures for 15 saturated normal-valent H MMH compounds (M = W-Pt) as well as corresponding hydrides of hypovalent Ta for comparison. The accuracy of the localized Lewis-like description (as measured by %pf) is found to be reasonably high both for normal-valent and for hypovalent species, typically >98% of the valence-shell density and >99.5% of the total density. 

As in main-group chemistry, hypovalent hydrides of the transition series have pronounced tendencies to form bridging tau bonds. In addition to H4TaTaH4, the multiply bonded species HTaTaH features two symmetrically bridging hydrides. Despite the complexity introduced by such tau bonds, the Ta-Ta interaction clearly has high bond order, as simple Lewis-like structures prescribe. 

As idealized computational models of metal hypovalency, let us therefore consider the early second-series transition-metal hydrides YH3, ZrH4, and NbfL (avoiding both the complications of lone-pair-bearing ligands and those associated with the lanthanide series). Figure 4.54 shows optimized structures of these species, and Table 4.33 summarizes the bonding (omh) and nonbonding (nM ) orbitals and occupancies at the metal center. 

The characteristic tendency of hypovalent transition metals to interact in a side-on (T-shaped) manner with nearby alkyl C—H bonds has been designated as the agostic effect.45 Agostic M- C—H interactions can also be identified with general 3c/2e donor-acceptor interactions of 2c— lc type. Availability of suitable lc acceptor orbitals is a signature of hypovalent early transition metals. 

Why are transition metals so well suited for catalysis A complete treatment of this critical question lies well beyond the scope of this book, but we can focus on selected aspects of bond activation and reactivity for dihydrogen and alkene bonds as important special cases. Before discussing specific examples that involve formal metal acidity or hypovalency, it is convenient to sketch a more general localized donor-acceptor overview of catalytic interactions in transition-metal complexes involving dihydrogen49 (this section) and alkenes (Section 4.7.4). 

---