The 18 electron rule in chemistry is a useful tool in predicting the stability of metal complexes. For example in ferrocene the iron core provides 8 electrons and 2 cyclopentadienyl ligands add 2 x 5 making a total of 18: hence the compound is stable. Likewise in tetrakis(triphenylphosphine)palladium(0) Pd(P(Ph)3)4 palladium owns 10 electrons all by itself and as the four phosphine ligands each donate 2 electrons the grand total again is 18.
But what when said Pd(P(Ph)3)4 is the catalyst in one of those typical coupling reactions such as the Heck reaction?. All reaction schemes assume that the catalyst will willingly surrender two ligands to an electron-poor and reactive 14e compound before it can engage in actual business. On the other hand it is well known that the p-orbital in general is too high in energy for any electron sharing and the 18e rule falls apart.
In a recent publication Ahlquist & Norrby have some further doubts DOI and discuss a new computational model. Experimental evidence collected thus far suggests 4-coordinated Pd (P4) is in chemical equilibrium with tri-coordinated Pd (P3) and further dissociation to P2 is unfavourable. The popular theoretical modelling tool B3LYP is unable to replicate these dynamics (P3 formation too favourable) but Ahlquist & Norrby throw in London dispersion for a quick fix. In another find the d-orbital occupancy was found to decrease on going from P1 (9.72 electrons) to P4 (9.61 electrons). This means the driving force for the formation of P3 and P4 is back donation from the electron-rich Pd center to the ligands. This is a departure from the established driving force featuring an electron-deficient Pd center. Ahlquist & Norrby also offer an explanation for the formation of P2. They invoke ps orbital hybridisation which makes the molecule linear with long metal to ligand sigma bonds.