Crystal Field Theory

The puzzle CFT was built to solve

By the 1890s chemists could prepare hundreds of transition-metal compounds, and two facts cried out for explanation. First, they were spectacularly colored: aqueous Cu²⁺ is sky-blue, Cr³⁺ is violet-green, Ni²⁺ is grass-green, while the closed-shell ions next door — Zn²⁺, Ca²⁺, Al³⁺ — are all colorless. Second, the same metal in the same oxidation state could be magnetic or not depending only on what was bonded to it. [Fe(H₂O)₆]²⁺ has four unpaired electrons and is strongly paramagnetic; swap the six waters for six cyanides and [Fe(CN)₆]⁴⁻ has zero unpaired electrons and is diamagnetic. Same iron, same +2 charge, opposite magnetic behavior.

Crystal field theory, introduced by physicist Hans Bethe in 1929 and applied to magnetism by John Van Vleck through the 1930s, answers both with one idea. Forget covalent bonding for a moment and treat the metal–ligand interaction as purely electrostatic: the ligands are point negative charges (or the negative ends of dipoles) sitting around a positive metal ion. Those charges create an electric field that the metal's d-orbitals feel — and crucially, they feel it unequally, because the five d-orbitals point in different directions in space.

d-Orbital splitting: the heart of the model

A free gaseous metal ion has five degenerate d-orbitals — same energy. Their shapes matter: d_z² and d_x²−y² have lobes lying along the x, y, z axes, while d_xy, d_xz, and d_yz have lobes pointing between the axes.

Now bring six ligands in along the ±x, ±y, ±z axes — an octahedral arrangement. An electron in d_x²−y² or d_z² sits head-on toward the incoming negative charges, so it is electrostatically repelled and rises in energy. These two orbitals form the upper e_g set. The three orbitals that point between the ligands feel less repulsion and drop, forming the lower t_2g set. The gap between them is the crystal field splitting energy, written Δ_o (the "o" for octahedral) or sometimes 10 Dq.

Energy is conserved relative to the hypothetical spherical field (the barycenter). The center of mass of the five orbitals stays put, so the three t_2g orbitals each sink by 0.4 Δ_o (−4 Dq) and the two e_g orbitals each rise by 0.6 Δ_o (+6 Dq). Check: 3 × (−0.4) + 2 × (+0.6) = 0. That bookkeeping is the basis of every CFSE calculation below.

Why the complexes are colored

The magic is that Δ_o for most first-row complexes lands squarely in the visible range. A d-electron can absorb a photon whose energy exactly matches Δ_o and jump from t_2g up to e_g — a d–d transition. The wavelength obeys E = hc/λ, so Δ_o = 20,000 cm⁻¹ corresponds to λ ≈ 500 nm. That green light is absorbed, and the eye perceives the complementary color, in this case red-purple.

The textbook case is [Ti(H₂O)₆]³⁺, a d¹ ion with a single absorption band peaking near 20,300 cm⁻¹ (493 nm). It removes green-yellow light and looks purple. Tighten the field and Δ_o grows, shifting absorption toward the blue and the perceived color toward the orange-red. Loosen it and the reverse happens. This is why ligand identity changes color so dramatically: adding ammonia to pale-blue [Cu(H₂O)₄]²⁺ produces the deep royal-blue [Cu(NH₃)₄]²⁺ — NH₃ is a stronger-field ligand than water, so Δ increases and the absorption shifts.

Two corollaries fall straight out of the model. Ions with d⁰ (Ti⁴⁺, Sc³⁺) or d¹⁰ (Zn²⁺, Cu⁺) configurations have no possible d–d transition — empty or completely full — so their complexes are colorless. And d–d transitions are Laporte- and spin-forbidden, which is why metal-complex colors are pale (molar absorptivity ε ≈ 1–100 L mol⁻¹ cm⁻¹) compared with intense charge-transfer bands like the deep purple of permanganate (ε ≈ 2,000).

The spectrochemical series

How big Δ_o gets depends on the ligand. Ranking ligands by the splitting they produce gives the spectrochemical series, an empirical order discovered by Tsuchida (1938):

I⁻ < Br⁻ < S²⁻ < SCN⁻ < Cl⁻ < NO₃⁻ < F⁻ < OH⁻ < C₂O₄²⁻ < H₂O < NCS⁻ < NH₃ < en < bipy < phen < NO₂⁻ < CN⁻ ≈ CO

Weak-field ligands on the left (halides, water) give small splittings; strong-field ligands on the right (CN⁻, CO) give large ones. Purely electrostatic CFT cannot explain why neutral CO outclasses anionic I⁻ — that requires π back-bonding, the donation of metal electron density into empty ligand π* orbitals, captured by the covalent successor ligand field theory (a molecular-orbital treatment). CFT is the simple limit; ligand field theory is the honest one. Δ_o also rises with metal oxidation state (more positive metal pulls ligands closer) and down a group (4d and 5d orbitals are larger and more diffuse, so 2nd- and 3rd-row complexes are almost always low-spin).

High-spin vs low-spin and CFSE

For configurations d⁴ through d⁷ there is a genuine choice. After filling t_2g with three electrons, the fourth can either pay the pairing energy P to double up in t_2g, or pay Δ_o to climb into the empty e_g. Nature minimizes energy:

• Δ_o < P (weak field): cheaper to promote → electrons spread out → high-spin, maximum unpaired electrons, paramagnetic.
• Δ_o > P (strong field): cheaper to pair → electrons crowd t_2g → low-spin, fewer unpaired electrons.

This is the resolution of the iron puzzle from the opening. High-spin [Fe(H₂O)₆]²⁺ (d⁶, weak water field) keeps 4 unpaired electrons; low-spin [Fe(CN)₆]⁴⁻ (d⁶, strong cyanide field) pairs everything into t_2g for 0 unpaired electrons and is diamagnetic. The magnetic moment μ ≈ √[n(n+2)] Bohr magnetons makes this measurable: 4.9 μ_B versus 0.

The energetic payoff of filling the split orbitals is the crystal field stabilization energy (CFSE): CFSE = (−0.4 n_t2g + 0.6 n_eg) Δ_o + (extra pairing). For low-spin d⁶ [Co(NH₃)₆]³⁺ that is 6 × (−0.4) Δ_o = −2.4 Δ_o, a huge stabilization that makes Co(III) ammines famously inert. CFSE produces the characteristic double-humped curves seen when you plot hydration enthalpies, lattice energies, or ionic radii across Ca²⁺ → Zn²⁺: the d⁰, d⁵ (high-spin), and d¹⁰ ions have zero CFSE and sit on a smooth baseline, while d³ and d⁸ ions bulge with extra stability.

Geometry changes the picture

Crystal field splitting by coordination geometry

Geometry	Splitting	Lower set	Upper set	Typical spin
Octahedral (6 ligands)	Δo (= 10 Dq)	t2g (−0.4 Δo)	eg (+0.6 Δo)	high or low spin
Tetrahedral (4 ligands)	Δt ≈ 4/9 Δo	e (−0.6 Δt)	t2 (+0.4 Δt)	almost always high spin
Square planar (4 ligands)	large; dx²−y² far above	dxz, dyz, dz², dxy	dx²−y²	diamagnetic d⁸ (e.g. Pt²⁺)

The tetrahedral field inverts the octahedral pattern — the e set drops and t₂ rises — and the gap shrinks to Δ_t ≈ (4/9) Δ_o because only 4 ligands surround the metal and none points directly at a d-orbital. The small Δ_t almost never exceeds the pairing energy, so tetrahedral complexes such as [CoCl₄]²⁻ (deep blue) are reliably high-spin. Square-planar geometry, the limit of removing two trans ligands from an octahedron, leaves the d_x²−y² orbital pointing at four in-plane ligands and pushes it sky-high; d⁸ ions like Ni²⁺, Pd²⁺, and Pt²⁺ exploit this to make the popular diamagnetic square-planar complexes of organometallic catalysis (think cisplatin, cis-[Pt(NH₃)₂Cl₂]).

Jahn–Teller distortion and the blue of copper

When the splitting leaves an orbitally degenerate ground state — most famously high-spin d⁴ (Mn³⁺, Cr²⁺) and d⁹ (Cu²⁺) — the molecule lowers its energy by distorting, the Jahn–Teller theorem. Cu²⁺ (d⁹) has an unevenly filled e_g set, so the octahedron elongates along z: two axial bonds lengthen and four equatorial bonds shorten. That distortion is why "octahedral" Cu(II) is never quite regular and why copper complexes show broad, asymmetric absorption bands. It also underpins the brilliant tetragonal blue of [Cu(NH₃)₄(H₂O)₂]²⁺.

Where it matters

CFT is not a toy. The same d-orbital splitting governs the color of gemstones — ruby is Cr³⁺ doped into corundum (Al₂O₃), where the octahedral oxide field sets Δ_o to absorb violet and green, leaving red; emerald is the same Cr³⁺ in beryl, where a slightly weaker field shifts the absorption and turns the stone green. In biology, the change in iron's spin state on O₂ binding in hemoglobin — high-spin five-coordinate deoxy-Fe(II) flipping to low-spin six-coordinate oxy-Fe(II) — is the trigger that pulls the iron into the porphyrin plane and drives cooperative oxygen uptake. In materials and catalysis, CFSE and spin state control catalytic activity, the colors of pigments (cobalt blue, chrome green), and the stability of laser crystals (Cr³⁺ in ruby was the first laser medium, 1960). For a deeper covalent treatment that adds π-bonding and orbital overlap, see molecular orbital and ligand field theory.