Crystal Field Theory

Crystal Field Theory (CFT) is a quantum-mechanical electrostatic model of coordination complexes that describes how the non-spherical electric field created by surrounding coordinating ligands (modeled as negative point charges or point electric dipoles) lifts the five-fold spatial degeneracy of the valence d orbitals (d_xy, d_xz, d_yz, d_x²−*y*², and d_z²)—or seven-fold degeneracy of f orbitals—of a central transition metal cation. By partitioning the five d orbitals into symmetry-adapted subsets separated by an energy gap Δ (or 10*Dq*), Crystal Field Theory quantitatively accounts for the vivid d–d electronic absorption colors, high-spin versus low-spin paramagnetism, Crystal Field Stabilization Energies (CFSE), double-humped lattice and hydration enthalpy curves, spinel mineral site preferences, and Jahn–Teller geometric distortions of transition-metal ions.

The mathematical foundation of Crystal Field Theory was developed in 1929 by the German-American theoretical physicist Hans Bethe, who applied point-group symmetry representation theory to calculate how crystalline electrostatic environments split atomic energy levels of ions embedded in ionic lattices. Between 1932 and 1935, the American physicist John Hasbrouck van Vleck (1977 Nobel Prize in Physics) extended Bethe's symmetry formalism to explain the magnetic susceptibilities of transition-metal and lanthanide salts, and later integrated CFT with molecular orbital theory to create Ligand Field Theory (LFT). In the early 1950s, Leslie Orgel, Carl J. Ballhausen, and Yukito Tanabe and Satoru Sugano adapted CFT to interpret the electronic absorption spectra and thermodynamic trends of discrete coordination complexes in solution.

In bioinorganic chemistry, Crystal Field Theory explains the electronic mechanisms of metalloproteins—most famously how oxygen binding to the iron(II) porphyrin center of hemoglobin switches the d⁶ Fe²⁺ ion from a paramagnetic high-spin (S = 2) state to a diamagnetic low-spin (S = 0) state, shrinking its ionic radius and triggering the Perutz cooperativity transition across the hemoglobin tetramer.

---

In an isolated, gas-phase spherical metal ion, all five valence *n*d orbitals are exactly degenerate in energy. However, their angular probability distributions fall into two distinct spatial families relative to the Cartesian (x, y, z) axes:

• The Axial e_g Set (d_z² and d_x²−*y*²): Their lobes of maximum electron density point directly along the Cartesian x, y, and z axes.
• The Inter-Axial t_2g Set (d_xy, d_xz, and d_yz): Their four lobes lie in the xy, xz, and yz planes and bisect the axes at 45°, pointing between the Cartesian axes.

1. Octahedral (O_h) Splitting (Δ_o or 10*Dq*)

Imagine bringing six negative point-charge ligands from infinity to distance R around a metal cation Mⁿ⁺. If the six charges were smeared uniformly over a spherical shell (the barycenter or center of gravity), all five d orbitals would be raised equally in energy by electrostatic repulsion between the d electrons and the negative ligand shell. When the six ligands are instead localized at the six vertices of a regular octahedron along the ±*x*, ±*y*, and ±*z* axes (point group O_h), conservation of total electrostatic energy requires the average energy of the five orbitals to remain at the spherical barycenter (the barycenter rule: Σ Δ*E*_i = 0).

Because the two e_g orbitals (d_x²−*y*² and d_z²) point directly at the six approaching negative ligands, they experience strong head-on electrostatic repulsion and are destabilized above the barycenter. Conversely, the three t_2g orbitals (d_xy, d_xz, d_yz) point into the regions between the ligands and are stabilized below the barycenter. Denoting the total energy separation between the t_2g and e_g sets as the octahedral crystal field splitting parameter Δ_o (or 10*Dq*), the barycenter balance 3(E_*t*2g) + 2(E_*e*g) = 0 with E_*e*g − E_*t*2g = Δ_o yields:

E(t_2g) = −0.4 Δ_o = −(2/5) Δ_o = −4 Dq

E(e_g) = +0.6 Δ_o = +(3/5) Δ_o = +6 Dq

In point-charge perturbation theory, Δ_o scales inversely with the fifth power of the metal–ligand bond distance (Δ_o ∝ q_L ⟨*r*⁴⟩ / R⁵), making crystal field splitting highly sensitive to metal oxidation state and ionic radius.

2. Tetrahedral (T_d) Splitting (Δ_t)

In a four-coordinate tetrahedral complex (point group T_d, which lacks an inversion center i, so the g gerade parity subscript is omitted), the four ligands occupy alternate vertices of a cube surrounding the metal atom at the center, while the Cartesian x, y, and z axes pass through the centers of the six cube faces. Now, the three t₂ orbitals (d_xy, d_xz, d_yz) point toward the cube edges (35.26° from the ligands), whereas the two e orbitals (d_x²−*y*², d_z²) point toward the cube faces (54.74° from the ligands). Consequently, the tetrahedral splitting diagram is inverted relative to the octahedral case:

E(e) = −0.6 Δ_t = −6 Dq_t (doubly degenerate lower set)

E(t₂) = +0.4 Δ_t = +4 Dq_t (triply degenerate upper set)

Because a tetrahedron has only 4 ligands instead of 6 (a factor of 4/6 = 2/3) and none of the d orbitals point directly at the tetrahedral vertices (reducing angular overlap by another factor of 2/3), the tetrahedral splitting for the same metal, ligands, and bond length is fixed at:

Δ_t = (4/9) Δ_o ≈ 0.44 Δ_o

Because Δ_t is less than half of Δ_o, it is almost universally smaller than the spin-pairing energy P (Δ_t < P). As a result, virtually all tetrahedral first-row transition-metal complexes (e.g., [CoCl₄]²⁻, [FeCl₄]⁻) are high-spin.

3. Tetragonal Elongation and Square-Planar (D_4h) Splitting

If the two axial z-axis ligands of an octahedral complex are pulled outward (a tetragonal Jahn–Teller elongation reducing symmetry from O_h to D_4h) and ultimately removed entirely to leave four ligands on the ±*x* and ±*y* axes, a square-planar complex is formed. Removing the ±*z* ligands sharply stabilizes every orbital possessing a z-component (d_z², d_xz, d_yz), while the four in-plane x, y ligands move closer and strongly destabilize the d_x²−*y*² orbital:

E(d_xz, d_yz) < E(d_z²) < E(d_xy) ≪ E(d_x²−*y*²)

Because the energy gap below the top d_x²−*y*² orbital is extremely large (≈1.3 Δ_o), d⁸ metal ions with large intrinsic splitting—specifically 4*d*⁸ and 5*d*⁸ metals (Pd²⁺, Pt²⁺, Au³⁺, Rh⁺, Ir⁺) such as the anticancer drug cisplatin (cis-[PtCl₂(NH₃)₂]) and [PtCl₄]²⁻, as well as strong-field 3*d*⁸ [Ni(CN)₄]²⁻—universally adopt diamagnetic (S = 0) square-planar geometries with eight paired electrons filling the lowest four orbitals.

---

For an octahedral complex with n_*t*2g electrons in the lower t_2g set and n_*e*g electrons in the upper e_g set, the Crystal Field Stabilization Energy (CFSE) relative to the spherical barycenter is:

CFSE (O_h) = ( −0.4 n_*t*2g + 0.6 n_*e*g ) Δ_o + m P

where m is the number of additional electron pairs forced to form by the ligand field beyond those already present in the free-ion Hund's rule high-spin state, and P is the mean spin-pairing energy. For configurations d¹, d², d³, and d⁸, d⁹, d¹⁰, only one ground-state electronic arrangement is possible. However, for octahedral d⁴, d⁵, d⁶, and d⁷ ions, two competing configurations exist:

• Weak-Field / High-Spin (Δ_o < P): The fourth (to seventh) electron ascends into the upper e_g orbitals rather than paying the pairing penalty P, maximizing the number of unpaired spins in accordance with Hund's first rule.
• Strong-Field / Low-Spin (Δ_o > P): The orbital gap Δ_o exceeds P, so electrons pair completely within the lower t_2g orbitals before occupying e_g.

Thermodynamic Consequences: Double-Humped Curves and Spinels

When standard hydration enthalpies (Δ*H*_hyd^° of [M(H₂O)₆]²⁺) or lattice energies (U₀ of MCl₂ and MO) are plotted across the first transition row from Ca²⁺ (d⁰) to Zn²⁺ (d¹⁰), the experimental values do not follow a smooth monotonic line: instead, they trace a characteristic double-humped curve. The three ions with CFSE = 0—Ca²⁺ (d⁰), high-spin Mn²⁺ (d⁵, t_2g³e_g²), and Zn²⁺ (d¹⁰)—define a straight baseline (generated solely by progressive ionic radius contraction across the period), while all other dⁿ ions sit below the line (more exothermic) by an amount exactly equal to their octahedral CFSE, peaking at V²⁺ (d³, −1.2 Δ_o) and Ni²⁺ (d⁸, −1.2 Δ_o). Similarly, in spinel oxides (AB₂O₄) containing both tetrahedral and octahedral oxide cavities, comparing octahedral vs. tetrahedral CFSE values explains why MgAl₂O₄ is a normal spinel whereas magnetite (Fe²⁺Fe³⁺₂O₄) is an inverse spinel: high-spin d⁵ Fe³⁺ has zero CFSE in both sites, while d⁶ Fe²⁺ gains −0.4 Δ_o in an octahedral site vs. −0.6 Δ_t ≈ −0.27 Δ_o in a tetrahedral site, driving Fe²⁺ into the octahedral holes.

---

Measuring the wavenumber (ν̃ = 1/λ = Δ_o / hc) of the t_2g → e_g optical absorption band across diverse complexes led the Japanese chemist Ryutaro Tsuchida (1938) to order ligands empirically into the spectrochemical series of increasing Δ_o:

I⁻ < Br⁻ < S²⁻ < SCN⁻ < Cl⁻ < NO₃⁻ < F⁻ < OH⁻ < C₂O₄²⁻ < H₂O < NCS⁻ < edta⁴⁻ < NH₃ < en < bpy ≈ phen < NO₂⁻ < PPh₃ < CN⁻ < CO

Similarly, for a fixed ligand set, Δ_o increases with metal oxidation state (M³⁺ > M²⁺ by ~50%) and increases by 30–50% down a periodic group (3*d* < 4*d* < 5*d*, explaining why 4*d* and 5*d* octahedral complexes are nearly always low-spin).

Pure electrostatic Crystal Field Theory cannot explain why neutral carbon monoxide (CO, dipole moment μ = 0.11 D) produces a far larger Δ_o than negatively charged halide anions (F⁻, Cl⁻). Ligand Field Theory (LFT) resolves this paradox by incorporating covalent molecular orbital π interactions between metal t_2g orbitals and ligand frontier orbitals:

1. π-Donor Ligands (Halides I⁻, Br⁻, Cl⁻, F⁻, OH⁻, O²⁻): Possess filled *p*π lone pairs lower in energy than metal t_2g; π-overlap pushes the metal t_2g (now π* antibonding) upward toward e_g, shrinking Δ_o (weak-field).
2. σ-Only Ligands (NH₃, en): Lack π-symmetry lone pairs or low-lying π* orbitals, leaving t_2g non-bonding (intermediate Δ_o).
3. π-Acceptor Ligands (CN⁻, CO, PR₃, bpy, O₂): Possess empty, low-lying antibonding π* orbitals higher in energy than metal t_2g; dπ → pπ* backbonding stabilizes and pushes the metal t_2g set downward, widening Δ_o (strong-field). In hemoglobin, binding of the strong-field π-acceptor ligand O₂ (or CO) to five-coordinate high-spin deoxy-Fe²⁺ (d⁶, t_2g⁴e_g², r = 78 pm) widens Δ_o above P, collapsing all six d electrons into the t_2g⁶ low-spin configuration (r = 61 pm); depopulating the antibonding d_x²−*y*² orbital allows the contracted Fe²⁺ ion to snap into the plane of the porphyrin ring, pulling the proximal histidine (His F8) to trigger the T → R cooperative transition.

---

Why do the e_g orbitals lie higher in energy than the t_2g orbitals in an octahedral complex?

In an octahedral complex, six negatively charged or dipolar ligands approach the central metal cation directly along the Cartesian ±*x*, ±*y*, and ±*z* axes. Because the lobes of the two e_g orbitals (d_x²−*y*² and d_z²) point directly along these axes at the ligands, electrons in e_g experience strong electrostatic and σ-antibonding repulsion (+0.6 Δ_o), whereas the three t_2g orbitals (d_xy, d_xz, d_yz) point between the axes and are stabilized (−0.4 Δ_o).

Why are almost all tetrahedral transition-metal complexes high-spin?

In a tetrahedral geometry, only four ligands coordinate the metal (4/6 as many as in an octahedron) and none of the five d orbitals point directly at the ligand vertices. Consequently, the tetrahedral crystal field splitting is only four-ninths of the octahedral splitting (Δ_t = [4/9]*Δ*_o ≈ 0.44 Δ_o). Because this small orbital gap is nearly always smaller than the electron spin-pairing energy (Δ_t < P), electrons populate the upper t₂ orbitals singly before pairing.

Why is neutral carbon monoxide (CO) a much stronger-field ligand than a negative chloride ion (Cl⁻)?

A purely electrostatic point-charge model predicts that anionic Cl⁻ should split d orbitals more strongly than neutral CO, yet the spectrochemical series shows the opposite. Ligand Field Theory explains that Cl⁻ has filled *p*π lone pairs that donate electron density into the metal t_2g orbitals, raising t_2g energy and shrinking Δ_o. Conversely, CO possesses empty antibonding π* orbitals that accept electron density from the metal t_2g orbitals (*d*π–*p*π* backbonding), lowering t_2g energy and widening Δ_o.

---

1. Bethe, H. (1929). "Termaufspaltung in Kristallen." Annalen der Physik, 395(2), 133–208. https://doi.org/10.1002/andp.19293950202
2. Van Vleck, J. H. (1932). "Theory of the variations in paramagnetic anisotropy among different salts of the iron group." Physical Review, 41(2), 208–215. https://doi.org/10.1103/PhysRev.41.208
3. Orgel, L. E. (1952). "The effects of crystal fields on the properties of transition-metal ions." Journal of the Chemical Society, 1952, 4756–4761. https://doi.org/10.1039/JR9520004756
4. Perutz, M. F. (1970). "Stereochemistry of cooperative effects in haemoglobin." Nature, 228(5273), 726–734. https://doi.org/10.1038/228726a0