Acidities of Hydride Complexes

The pJCj values are now available for many hydride complexes. Extensive tables have been compiled recently by Bullock and by Tilset. The rate of proton transfer to and from transition metals is rather slow (see below), so it is often possible to detect separate NMR signals for M-H and M , and tiius to determine the position of proton transfer equilibria between hydride complexes (M-H) and bases (B), or metal bases (M ) and organic acids (HA). The pX values in Table 3.1 have been obtained in acetonitrile, an excellent solvent for acid-base chemistry because it solvates cations well enough to minimize ion pair formation it is both a weak acid and a weak base, with a very low autoprotolysis constant (ion product).  

In less polar solvents it is impractical to determine pK values, either because the solvent is not readily protonated (CH Cy or because the solvent permits extensive ion pair formation and its protonated form is unstable (THF). However, the positions of proton transfer equilibria in such solvents (e.g.. Equation 3.125) have been used to estimate relative aqueous pfC/s for hydride complexes that are insoluble in (or unstable in) water. CyjPH, whose aqueous pfC is 9.7, has often been used as an anchor—that is, the acidities of other RjPH and hydride complexes have been measured relative to CyjPH, leading to the pseudo-aqueous pJ f values in Table 3.2. Morris has published acidity scales in CHjClj and in THF. However, complications due to ion pairing make such pK values less reliable than direct measurements in CHjCN (or water). Morris has linked his THF acidity scale with acidities in DMSO, a solvent that minimizes ion pair formation, but is incompatible with many organometallic complexes.  

The pfC of HCo(CO) has been reported as 8.4 in CHjCN, but should be considered uncertain, as the bond strength (67 kcal mol ) that results from the use of that pK in a thermodynamic cycle is too high to be consistent with the reactivity of HCo(CO) in H  

The most acidic hydrides in Table 3.2 have pK values comparable to that of HCl (which has a pK between 8 and 9 in CH3CN)/ while the least acidic have pK values exceeding that of phenol. Electron-donating substituents (i.e., the replacement of a carbonyl by a phosphiiTe or phosphite) predictably increase the pK of hydride complexes, as does descending a column of the periodic table (Cr Mo W, Fe Ru Os, and Mn Re). 

The rates at which protons can be removed from transition metal hydrides (their kinetic acidities ) generally parallel their thermodynamic acidities pK values). However, the removal of a proton from a metal is much slower than the removal of a proton from an electronegative atom like nitrogen or oxygen. The reverse is also true the protonation of a metal (to form a hydride) is slower than the protonation of a nitrogen or an oxygen examples are shown in Equations 3.113 and 3.114. The low kinetic acidity of transition metal hydrides is much like that of carbon acids in organic chemistry. - -  

---

In view of the acidity of hydride complexes, one might expect them to form hydrogen-bonded intennediates during proton-transfer reactions. However, it is relatively uncommon for neutral hydride complexes to form hydrogen bonds M-H A because most M-H bonds are polarized as M(8+)-H(8-), not as M(8-)-H(8+). (The need to reverse the observed polarization during deprotonation is a major cause of the low kinetic acidity of transition metal hydrides, mentioned previously.) The first M-H Ahydrogen bond from a neutral hydride has just been reported CpM(CO)3H (M = Mo and W) serves as a hydrogen bond donor to (octyl)3P=0 and even to pyridine, apparently because there is M(8-)-H(8+) polarization in its M-H bond. °... 

---