Ligand Field Theory

What ligand field theory does

Take a bare transition-metal ion — say Fe³⁺ or Co²⁺. Its five 3d orbitals are all at the same energy (degenerate). Now surround it with six ligands in an octahedron. Ligand field theory (LFT) asks a single question: what happens to the energies of those five d orbitals when real M–L bonds form? The answer is that the degeneracy breaks, and the size of the split controls almost everything you can observe about the complex — its color, how many unpaired electrons it has, its stability, and its reactivity.

LFT is the molecular-orbital (MO) treatment of the metal center. The older crystal field theory (CFT) got most of the geometry right by pretending ligands were negative point charges that electrostatically repel the d electrons. That works surprisingly well, but it cannot explain covalency, and it gets the ordering of neutral ligands badly wrong. LFT keeps CFT's convenient labels (t₂g, eg, Δo) but derives them from orbital overlap instead of point charges — and in doing so explains the things CFT can't.

• σ-framework. Each ligand donates a lone pair into the metal. The two metal d orbitals that point directly at the ligands — d_z² and d_x²−y², the eg set — overlap strongly and become antibonding eg* molecular orbitals, pushed up in energy.
• The nonbonding trio. The other three d orbitals — d_xy, d_xz, d_yz, the t₂g set — point between the ligands, so they don't participate in σ-bonding and stay low, essentially nonbonding.
• The gap. The energy difference between t₂g and eg* is the ligand field splitting parameter Δo (the "o" is for octahedral). Δo is what you read off a UV–Vis spectrum, and it is the single most important number in the whole theory.

Building the d-orbital splitting, orbital by orbital

Here is the LFT construction for an octahedral ML₆ complex, step by step. Think of it as an electron-bookkeeping exercise where every orbital either goes up, goes down, or stays put.

1. Assemble the σ-donor orbitals. The six ligand lone pairs combine into six symmetry-adapted linear combinations (SALCs) that transform as a₁g + eg + t₁u under octahedral symmetry. These are the orbitals that will bond to the metal.
2. Match by symmetry. Only orbitals of the same symmetry mix. The metal's s orbital (a₁g), p orbitals (t₁u), and the d_z²/d_x²−y² pair (eg) all find ligand partners. Six bonding MOs form low in energy (filled by the twelve ligand electrons) and their antibonding partners rise.
3. The eg orbitals go up. Because d_z² and d_x²−y² overlap head-on with ligand lone pairs, their antibonding combination — labeled eg* — is destabilized. This is the top of the d-manifold.
4. The t₂g orbitals stay put. The d_xy, d_xz, d_yz trio has the wrong symmetry (t₂g) to interact with any σ-donor SALC, so with pure σ-donors it is strictly nonbonding — the floor of the d-manifold.
5. The metal d electrons fill from the bottom. Only the metal's own d electrons occupy the t₂g/eg* region. The split between them is Δo, and how those electrons distribute is decided next.
6. π-interactions fine-tune t₂g. If a ligand has orbitals of t₂g symmetry — filled p lone pairs (π-donor) or empty π* orbitals (π-acceptor) — they mix with the metal t₂g set and shift it, changing Δo. This π-step is the part CFT completely misses, and it is what makes the theory quantitative.

   free ion (5 × d, degenerate)      octahedral field

        — — — — —                          ___ ___   eg*  (dz², dx²−y²)   ← σ* antibonding
                                             ▲
                                             │  Δo   = ligand field splitting
                                             ▼
                                        ___ ___ ___   t₂g  (dxy, dxz, dyz) ← nonbonding (σ only)

   barycenter rule:  eg* rises +0.6 Δo,  t₂g falls −0.4 Δo   (weighted average unchanged)

The barycenter rule keeps the accounting honest: the average energy of the five d orbitals is unchanged by splitting, so the two eg* orbitals rise by +0.6 Δo (3/5) while the three t₂g orbitals drop by −0.4 Δo (2/5). Multiply occupancies by these values and you get the ligand field stabilization energy (LFSE) — the extra stability the complex gains from the split, which drives trends in hydration enthalpies and lattice energies across a transition series.

The π-story: why the spectrochemical series is what it is

The σ-only picture gives the right cast of orbitals but the wrong sizes. π-bonding is the plot twist that makes LFT predictive.

• π-donor ligands (F⁻, Cl⁻, Br⁻, I⁻, OH⁻, H₂O to some degree) carry filled p lone pairs of t₂g symmetry. These push the metal t₂g set up in energy (filled–filled repulsion). Raising the t₂g floor shrinks Δo → weak field.
• π-acceptor ligands (CO, CN⁻, phenanthroline, bipyridine, NO⁺, alkenes) carry empty π* orbitals of t₂g symmetry. The filled metal t₂g electrons donate into them (π-backbonding), which pulls the t₂g set down. Lowering the t₂g floor enlarges Δo → strong field.
• Pure σ-donors (NH₃, en, amines) have no accessible π orbitals, so t₂g stays nonbonding and they sit in the middle of the series.

This immediately resolves the paradox that broke crystal field theory: neutral CO produces a larger Δo than charged F⁻. Electrostatics says the anion should win; π-backbonding says the acceptor wins — and experiment agrees. The strength of a ligand is set by the direction and magnitude of its π-interaction, not by its formal charge.

The ingredients: metal, oxidation state, geometry, and Δo

Δo isn't one universal number; it scales with several physical factors, and knowing them lets you predict color and spin before running a spectrum.

• The ligand (biggest lever). Set by the spectrochemical series. For [Cr(III)] complexes, Δo climbs from ≈13,600 cm⁻¹ with 6 Cl⁻ to ≈17,400 cm⁻¹ with 6 H₂O to ≈26,600 cm⁻¹ with 6 CN⁻ — a two-fold swing from ligand choice alone.
• Oxidation state. Higher charge on the metal contracts the d orbitals and pulls the ligands closer, increasing overlap. Δo for a 3+ ion is roughly 30–80% larger than for the same metal as 2+: [Fe(H₂O)₆]²⁺ ≈ 10,400 cm⁻¹ vs [Fe(H₂O)₆]³⁺ ≈ 13,700 cm⁻¹.
• Position in the transition series. Δo grows going down a group: 3d < 4d < 5d, by roughly 30–50% per row. This is why second- and third-row complexes (Rh, Ir, Pt) are almost always low-spin — their Δo dwarfs the pairing energy.
• Geometry. Δt (tetrahedral) ≈ (4/9)Δo for the same metal and ligands, because only four ligands are present and none point straight at a d orbital. Square-planar splitting is largest of all and is the natural home of d⁸ ions.
• Pairing energy P. Not a field factor, but the number Δo must beat to force low-spin. P is an intrinsic property of the ion (electron–electron repulsion), typically 15,000–25,000 cm⁻¹ for first-row metals.

Crystal field theory vs ligand field theory

	Crystal field theory (CFT)	Ligand field theory (LFT)
Model of the M–L bond	Electrostatic point charges	Molecular orbitals from overlap
Nature of eg orbitals	Metal d, raised by repulsion	Antibonding σ* (eg*), metal + ligand
Nature of t₂g orbitals	Metal d, lowered	Nonbonding (σ) → π-tuned
Origin of Δo	Ligand charge / distance	σ-overlap ± π-interaction
Covalency	Ignored entirely	Built in (nephelauxetic effect)
Neutral π-acceptors (CO, CN⁻)	Predicts weak field (wrong)	Predicts strong field (correct)
Spectrochemical series	Cannot derive it	Reproduces the full ordering
Charge-transfer bands	No account	M→L / L→M transitions explained
Best for	Quick geometry & spin predictions	Quantitative bonding, spectra, series

The practical takeaway: reach for CFT when you just need to count unpaired electrons or sketch a splitting diagram; reach for LFT when you need to explain why a ligand sits where it does in the series, or interpret a charge-transfer band.

Worked example: is [Fe(CN)₆]³⁻ high-spin or low-spin, and what color?

Take the hexacyanoferrate(III) ion, the anion of the classic reagent potassium ferricyanide.

1. Count the d electrons. Fe is group 8. Fe³⁺ = [Ar]3d⁵. Five d electrons to place.
2. Find the ligand. CN⁻ is at the strong-field end of the spectrochemical series — a powerful π-acceptor. Expect a large Δo.
3. Compare Δo to P. For [Fe(CN)₆]³⁻, Δo ≈ 33,000 cm⁻¹, well above the pairing energy P ≈ 19,000 cm⁻¹. Since Δo > P, the complex is low-spin.
4. Fill the orbitals. All five electrons crowd into the three t₂g orbitals: (t₂g)⁵(eg*)⁰. That leaves exactly one unpaired electron — paramagnetic, but only weakly.
5. Predict the magnetic moment. Spin-only μ = √(n(n+2)) BM = √(1·3) = 1.73 μ_B, matching the measured ≈ 2.3 μ_B (orbital contribution raises it a little).
6. Predict color. The large Δo pushes the d–d absorption toward the blue/UV; ferricyanide solutions are pale yellow, with an intense low-energy ligand-to-metal charge-transfer band doing most of the visible absorption.

Contrast with [Fe(H₂O)₆]³⁺: water is weak-field, Δo ≈ 13,700 cm⁻¹ < P, so it is high-spin (t₂g)³(eg*)² with five unpaired electrons and μ ≈ 5.9 μ_B. Same Fe³⁺, same d⁵ count — swap the ligand and you flip the magnetism entirely. That is ligand field theory earning its name.

Real-world consequences

• The color of gemstones. Ruby is Cr³⁺ (d³) doped into corundum (Al₂O₃); the octahedral oxide field gives a Δo that absorbs green-yellow and violet, leaving the red glow. Emerald is the same Cr³⁺ ion in beryl — a slightly weaker field shifts the absorption and turns it green. One dopant, two gems, decided by Δo.
• Hemoglobin's spin switch. Deoxyhemoglobin holds Fe²⁺ (d⁶) high-spin; the large high-spin iron sits out of the porphyrin plane. When O₂ binds as a strong-field ligand it flips the iron low-spin, shrinking it enough to slip into the plane — the motion that triggers cooperative oxygen uptake across the whole protein.
• Pigments. Cobalt blue (CoAl₂O₄), chrome green, and the blue of cobalt glass all get their color from d–d transitions whose energy is set by the ligand field around the metal.
• Catalysis and CO poisoning. π-backbonding — the same effect that makes CO a strong-field ligand — is why CO binds heme iron ≈200× tighter than O₂, and why organometallic catalysts use CO and phosphines to stabilize low-valent metal centers.
• Spin-crossover materials. Some Fe(II) complexes sit right at Δo ≈ P, so a small change in temperature, pressure, or light flips them between high- and low-spin — the basis of molecular switches and pressure/temperature sensors.

Distortion and limitations

The clean t₂g/eg* picture assumes a perfect octahedron, but electron occupancy can distort it. When the eg* set is unevenly filled — most famously Cu²⁺ (d⁹, eg* holds 3 electrons) and high-spin Mn³⁺ (d⁴) — the Jahn–Teller effect elongates or compresses the octahedron to lift the degeneracy and lower the energy. That is why nearly every Cu²⁺ complex shows two long axial bonds and four short equatorial ones.

LFT also has honest limits. It is a one-electron, parametric model: Δo and the Racah parameters (B, C, describing electron repulsion) are fitted to experiment, not computed from first principles. It handles d–d transitions and magnetism well, but rigorous treatment of charge-transfer states, multi-electron term symbols, and heavy-metal spin–orbit coupling needs the full machinery of molecular orbital theory and, ultimately, computational methods like DFT. LFT is the sweet spot between CFT's cartoon and a full quantum calculation — powerful enough to be predictive, simple enough to do on a napkin.

Historical discovery

The lineage runs through three stages. In 1929 the physicist Hans Bethe worked out how an ion's energy levels split in crystals of different symmetry using group theory — the birth of crystal field theory. John Hasbrouck Van Vleck, who would share the 1977 Nobel Prize in Physics for his work on magnetism, showed in the 1930s that some covalent mixing was needed to fit real spectra and magnetic data, coining the more general "ligand field" viewpoint.

The theory reached its modern form in the 1950s through the Oxford school: J. S. Griffith and Leslie Orgel's influential 1957 review, and Orgel's diagrams correlating d–d transition energies with Δo across configurations, married group theory to experimental UV–Vis spectra. Their work turned a qualitative splitting picture into a quantitative tool that inorganic chemists still teach and use today, some seventy years later.