Q4(c)(i). Limitations of crystal field theory and the modifications

Limitations

1. Point-charge model: ligands are treated as charges or dipoles and the metal–ligand bond as purely ionic. Electrons are never shared.
2. Spectrochemical series unexplained: by charge, OH⁻ and F⁻ should split more than neutral H₂O, NH₃ or CO. In fact CO and CN⁻ give the largest Δₒ and halides the smallest. CO is a weak σ-donor but a strong π-acceptor; CN⁻ is a strong σ-donor and a weaker π-acceptor than CO. Neither ordering follows from charge.
3. Evidence of covalency: the interelectronic repulsion (Racah B) is smaller in complexes than in the gaseous ion (nephelauxetic effect); EPR of [IrCl₆]²- shows hyperfine splitting by the Cl nuclei, so the unpaired electron spends time on the ligands.
4. Other gaps: no account of intense charge-transfer bands, of metal–ligand π bonding in carbonyls and organometallics, or of the 18-electron rule; Δ cannot be calculated, only fitted.

Modifications

1. Ligand field theory (adjusted CFT): keeps the symmetry arguments but treats Δₒ, the Racah parameters B, C and the nephelauxetic ratio β=B_(complex)/B_(gaseous ion) as empirical parameters fitted to spectra.
2. Molecular orbital theory: ligand σ orbitals combine with metal a_(1g) (4s), t₁ᵤ (4p) and e_g (3d) orbitals; t_(2g) is non-bonding with σ-only ligands. Then π-donors raise t_(2g) and lower Δₒ, while π-acceptors lower t_(2g) and raise Δₒ (figure), which explains the spectrochemical series.
3. Angular overlap model: the energy of each d orbital is the sum of σ (e_σ) and π (e_π) contributions of each ligand, weighted by the angular overlap; it works for any geometry.