Chemical bonding in phosphane and amine complexes of main group elements and transition metals

The geometries and bond dissociation energies of the main group complexes X₃B-NX₃, X₃B-PX₃, X ₃Al-NX₃, and X₃Al-PX₃ (X = H, Me, Cl) and the transition metal complexes (CO)₅M-NX₃ and (CO)₅M-PX₃ (M = Cr, Mo, W) have been calculated using gradient-corrected density functional theory at the BP86/TZ2P level. The nature of the donor-acceptor bonds was investigated with an energy decomposition analysis. It is found that the bond dissociation energy is not a good measure for the intrinsic strength of Lewis acidity and basicity because the preparation energies of the fragments may significantly change the trend of the bond strength. The interaction energies between the frozen fragments of the borane complexes are in most cases larger than the interaction energies of the alane complexes. The bond dissociation energy of the alane complexes is sometimes higher than that of the borane analogues because the energy for distorting the planar equilibrium geometry of BX₃ to the pyramidal from in the complexes is higher than for AIX₃. Inspection of the three energy terms, ‡E_Pauli, ‡E_orb, and ‡E _elstat, shows that all three of them must be considered to understand the trends of the Lewis acid and base strength. The orbital term of the donor-acceptor bonds with the Lewis bases NCl₃ and PCl₃ have a higher π character than the bonds of EH₃ and EMe ₃, but NCl₃ and PCl₃ are weaker Lewis bases because the lone-pair orbital at the donor atoms N and P has a high percent s character. The calculated ‡E_int values suggest that the trends of the intrinsic Lewis bases' strengths in the main-group complexes with BX₃ and AIX₃ are NMe₃ > NH₃ > NCl₃ and PMe₃ > PH₃ > PCl₃. The transition metal complexes exhibit a somewhat different order with NH ₃ > NMe₃ > NCl₃ and PMe₃ > PH₃ > PCl₃. The slightly weaker bonding of NMe ₃ than that of NH₃ comes from stronger Pauli repulsion. The bond length does not always correlate with the bond dissociation energy, nor does it always correlate with the intrinsic interaction energy.