Jahn-Teller Distortion

Jahn-Teller distortion is not a museum-piece symmetry curiosity; it shapes real chemistry and technology across scales. In bioinorganic chemistry, blue copper proteins (plastocyanin, azurin) exploit a deliberately frustrated Jahn-Teller geometry — the protein scaffold imposes a distorted tetrahedral-ish coordination on Cu(II)/Cu(I) that is intermediate between the ions' individually preferred geometries, lowering the reorganization energy for electron transfer (the "entatic state" concept) and enabling extremely fast redox cycling in photosynthetic electron transport chains. In materials science, cooperative Jahn-Teller ordering in perovskite manganites (La$_{1-x}$Sr$_x$MnO$_3$, La$_{1-x}$Ca$_x$MnO$_3$) — where high-spin Mn$^{3+}$ ($d^4$) sites distort cooperatively through the lattice — underlies colossal magnetoresistance, a property exploited in magnetic sensors and read-heads. The same Mn$^{3+}$ Jahn-Teller physics drives structural phase transitions in LaMnO$_3$ that are directly relevant to solid oxide fuel cell cathode materials and battery cathode chemistry (Jahn-Teller-active Mn$^{3+}$ in LiMn$_2$O$_4$ spinel cathodes causes structural degradation and capacity fade — a genuine engineering headache in lithium-ion battery design). In coordination chemistry and catalysis, the labilization of axial ligands caused by Jahn-Teller elongation (weaker axial bonds, as in the copper-aqua example) explains the characteristically fast ligand-exchange kinetics and distinctive reactivity patterns of Cu(II) and high-spin Mn(III)/Cr(II) centers, relevant to their roles as Lewis-acid catalysts and in bioinorganic oxidation chemistry. In spectroscopy and instrumentation, EPR and optical spectroscopists routinely use the axial or rhombic splitting patterns diagnostic of Jahn-Teller distortion to determine coordination geometry and electronic structure of paramagnetic transition-metal centers in catalysts, enzymes, and doped semiconductors, feeding directly into the more complete molecular-orbital treatments developed in Ligand Field Theory: MO Approach.

Ask the tutor: How does the Jahn-Teller stabilization energy change if the metal ion is high-spin $d^7$ in a tetrahedral field instead of octahedral $d^9$?

Start from an octahedral $d^9$ ion, copper(II), in a purely ionic crystal-field picture. The five $d$ orbitals split into $t_{2g}$ (lower, three-fold degenerate, orbitals $d_{xy}, d_{xz}, d_{yz}$) and $e_g$ (upper, two-fold degenerate, orbitals $d_{z^2}, d_{x^2-y^2}$). With nine $d$ electrons, the configuration is $t_{2g}^6 e_g^3$. The three lower orbitals are completely full and spherically symmetric in their combined occupation — no problem there. But the $e_g$ set holds three electrons in two orbitals: one orbital gets two electrons, the other gets one. There are two ways to do this — put the odd electron in $d_{z^2}$, or put it in $d_{x^2-y^2}$ — and in a perfect octahedron these two arrangements are exactly degenerate in energy. This is precisely the situation the Jahn-Teller theorem forbids for a stable equilibrium. The electronic wavefunction is degenerate at the symmetric nuclear geometry, and — crucially — the $e_g$ orbitals point directly along the metal–ligand bond axes ($d_{z^2}$ along $z$, $d_{x^2-y^2}$ along $x,y$). This means the electron density difference between the two possible occupations directly and strongly affects the electrostatic repulsion felt by the ligands along each axis. If the molecule elongates along $z$ (pushing the axial ligands out) while compressing in the $xy$-plane, the axial ligands feel less electron density if the odd electron sits in $d_{z^2}$ — but the true effect is subtler and must be tracked through both the electronic energy and the elastic cost of distortion. The resolution is that distortion along the tetragonal (elongation) coordinate splits the $e_g$ degeneracy: $d_{z^2}$ drops in energy (its lobes point at the receding axial ligands, so electron-electron repulsion with ligand charge decreases) while $d_{x^2-y^2}$ rises (its lobes point at the ligands that have moved closer in-plane). Simultaneously $t_{2g}$ splits into a lower $e_g$-derived pair ($d_{xz}, d_{yz}$) and a higher $b_{2g}$-derived level ($d_{xy}$), though this splitting is much smaller because $t_{2g}$ orbitals do not point at the ligands. The nine $d$-electrons then reoccupy these new levels, and because the odd electron in the upper (formerly degenerate) set is put into whichever new orbital is lower, the electronic energy drops. The energy cost of the actual bond distortion (a "spring" term, quadratic in the distortion amplitude) is overwhelmed, for small distortions, by the electronic stabilization (linear in the distortion amplitude), guaranteeing a net lowering of energy for some nonzero distortion — that is the essence of the vibronic-coupling argument given below.

The vertical direction in the figure is energy; the key feature is that the $e_g$ level splits by a large amount $\delta_1$ (because $e_g$ orbitals point at ligands) while $t_{2g}$ splits by a small amount $\delta_2$ (because $t_{2g}$ orbitals point between ligands). For $d^9$, the odd electron occupies the lower orbital of the split $e_g$ pair ($d_{z^2}$, if we elongate along $z$), giving a net stabilization of $\tfrac{1}{2}\delta_1$ relative to the (unstable) degenerate configuration — precisely the electronic driving force for elongation.

Group theory guarantees a degeneracy-lifting distortion exists but does not, by itself, say whether elongation or compression is favored — that requires including second-order (vibronic coupling to $t_{2g}$, and anharmonicity) effects. Empirically and computationally, elongation is overwhelmingly preferred for $d^9$ and high-spin $d^4$ octahedral complexes because it better relieves antibonding $\sigma^*$ overlap while keeping in-plane bonding strong; compressed geometries are thermodynamically less stable and rare.

The Jahn-Teller theorem is a statement about symmetry and degeneracy, provable by direct inspection of the totally symmetric part of the vibronic Hamiltonian. Its content can be summarized in two governing relations.

Symmetry selection rule for Jahn-Teller activity. A distortion along a normal coordinate $Q$ of symmetry $\Gamma_Q$ can couple linearly to (and split) an electronic state of symmetry $\Gamma_e$ only if

where: $\Gamma_e \times \Gamma_e$ is the symmetric direct product of the degenerate electronic representation with itself, and $\Gamma_Q$ is the symmetry of the candidate nuclear distortion (a non-totally-symmetric normal mode). If some non-totally-symmetric $\Gamma_Q$ appears in this product, that distortion linearly splits the degenerate electronic term, and the symmetric geometry is unstable.

Reading: for an octahedral $E_g$ electronic term (the $d^9$/high-spin-$d^4$ case), $E_g \times E_g = A_{1g} + A_{2g} + E_g$ in $O_h$. Since $E_g$ appears (and $E_g$ is the symmetry of the tetragonal distortion, i.e. differential elongation/compression along $z$ versus $x,y$), a Jahn-Teller-active vibrational mode of $E_g$ symmetry exists and couples linearly to the electronic degeneracy: the octahedron distorts tetragonally.

Vibronic energy as a function of distortion. Treating the electronic energy to first order and the lattice/bond energy to second order in a single distortion coordinate $Q$ gives

where: $E_0$ is the energy at the symmetric geometry, $k_{\text{vib}}$ is the linear vibronic coupling constant (how strongly the electronic energy responds to the distortion), $K$ is the force constant (elastic stiffness) of the mode, and the $\pm$ denotes the two branches of the split degenerate state.

Reading: because the destabilizing term is linear in $Q$ while the restoring term is quadratic, for any $k_{\text{vib}} \neq 0$ there is always a nonzero $Q^\ast$ that lowers $E(Q)$ below $E_0$ — the symmetric point $Q=0$ is a local energy maximum (or saddle point) on the lower branch, never a minimum. This is the algebraic heart of the theorem.

Example 1 — Foundational: identifying Jahn-Teller-active configurations by symmetry product. Determine whether the following high-spin octahedral configurations are (strongly) Jahn-Teller active: $d^1$, $d^4$ (high spin), $d^6$ (high spin), $d^8$. For each, write the ground electronic term's orbital degeneracy and check whether the degenerate set is $e_g$ (strong) or $t_{2g}$ (weak/negligible), or absent (none).

• $d^1$: configuration $t_{2g}^1$. The single electron can occupy any of three degenerate $t_{2g}$ orbitals — orbitally degenerate, but since $t_{2g}$ orbitals point between ligand axes, the coupling constant $k$ (Step 1 of the first derivation) is small. Weak J-T effect — real but structurally almost undetectable; distortions of a few thousandths of an Å.
• $d^4$ high spin: configuration $t_{2g}^3 e_g^1$. The single $e_g$ electron can occupy $d_{z^2}$ or $d_{x^2-y^2}$ — the same strong degeneracy as $d^9$, because $e_g$ orbitals point directly at ligands. Strong J-T effect. (This is the Mn$^{3+}$, Cr$^{2+}$ case.)
• $d^6$ high spin: configuration $t_{2g}^4 e_g^2$. The $e_g$ set is evenly occupied (one electron each in $d_{z^2}$ and $d_{x^2-y^2}$) — by the same cancellation logic as Step 2 above, an evenly half-filled $e_g^2$ set has no orbital degeneracy to begin with in the ground term (it is an orbital singlet, $^5E_g$ is NOT the ground term — high-spin $d^6$ octahedral is $^5T_{2g}$, arising from the $t_{2g}^4$ hole). The unevenly filled $t_{2g}^4$ set (3 orbitals, 4 electrons, hole degeneracy) gives a weak J-T effect (analogous to $d^1$ by hole-particle symmetry).
• $d^8$: configuration $t_{2g}^6 e_g^2$. With two electrons in two $e_g$ orbitals, high-spin coupling places one electron in each orbital (Hund's rule) — again an orbital singlet, ground term $^3A_{2g}$, not Jahn-Teller active by the $e_g$ mechanism. (This is why Ni$^{2+}$ octahedral complexes are regular, undistorted octahedra.)

This example demonstrates the central diagnostic: count electrons in the $e_g$ set specifically; only uneven occupation of $e_g$ (three or one electron in the two-orbital set, i.e. $e_g^1$ or $e_g^3$) gives strong, structurally obvious Jahn-Teller distortion.

Example 2 — Applied: predicting and rationalizing bond lengths in $[\text{Cu}(\text{H}_2\text{O})_6]^{2+}$. Copper(II) is $d^9$ ($t_{2g}^6 e_g^3$), strongly Jahn-Teller active. Given that the "expected" (undistorted, mean) Cu–O bond length estimated from ionic radii sums is about 2.15 Å, and that measured structures show four short bonds at 2.08 Å and two long bonds at 2.38 Å, rationalize the direction and rough magnitude of the distortion. Direction: the odd electron ($e_g^3$ = 2 in one orbital, 1 in the other) is lower in energy when placed in $d_{z^2}$ after axial elongation (Step 2 of the second derivation), because $d_{z^2}$'s lobes point directly at the receding axial ligands and lose antibonding $\sigma^*$ overlap. Hence elongation (not compression) along $z$ is predicted, matching observation (two long, four short bonds — the classic "4+2" copper(II) coordination footprint). Magnitude: real complexes show $\Delta r \approx 0.30$ Å (2.38 − 2.08), a large distortion by structural chemistry standards (compare typical bond-length precision of 0.01–0.02 Å in a good crystal structure). This is large precisely because the $e_g$ orbitals point at the ligands, so $\delta_1$ (Step 1 of the second derivation) is large — of order several thousand cm$^{-1}$ for Cu(II)-oxygen bonds, comparable in scale to $\Delta_o$ itself (roughly 12,000–13,000 cm$^{-1}$ for $[\text{Cu}(\text{H}_2\text{O})_6]^{2+}$). Consequence check: because the distortion is so large, some crystallographers describe copper(II) octahedra as "so distorted they are barely octahedral," and in several Cu(II) salts (e.g. $\text{CuCl}_2\cdot2\text{H}_2\text{O}$) the axial ligands are so weakly bound that the geometry is better described as square-planar plus two very long, semi-ionic axial contacts — the extreme limit of the same elongation trend.

Example 3 — Exam/competition level: quantitative estimate of the Jahn-Teller stabilization energy and its effect on measured $\Delta_o$. A high-spin $d^4$ octahedral complex $[\text{Mn}(\text{H}_2\text{O})_6]^{3+}$ shows two $d$-$d$ absorption bands at 18,900 cm$^{-1}$ and 24,500 cm$^{-1}$ (rather than a single band expected for an undistorted octahedron), attributed to splitting of the single spin-allowed $^5E_g \to {}^5T_{2g}$ transition by the Jahn-Teller distortion. Estimate the $e_g$ splitting parameter $\delta_1$ and the electronic stabilization energy $\Delta E_{JT} = \tfrac{3}{5}\delta_1$, then compare to typical $\Delta_o$ values for Mn(III)-aqua complexes (~21,000 cm$^{-1}$).

Step 1. In a tetragonally distorted $d^4$ ($t_{2g}^3 e_g^1$) system, the ground state $^5E_g$ (from the $e_g^1$ hole/electron degeneracy) splits by $\delta_1$, and the excited $^5T_{2g}$ state (derived from a $t_{2g}$-hole configuration) splits by a smaller amount $\delta_2$. To leading order, treat the two observed bands as origin transitions from the lower ground component to the two components of the (mildly split) excited state; a simplified but standard estimate attributes the majority of the observed splitting to the ground-state $e_g$ splitting itself, since it dominates:

Step 2. Apply the stabilization-energy relation derived above (adapted for $d^4$: one electron in the lower $e_g$-derived orbital rather than the $d^9$ case of one hole; the algebra is identical by electron-hole symmetry, giving the same coefficient):

Step 3 — sanity check against $\Delta_o$. Compare $\Delta E_{JT}$ to the parent octahedral splitting, $\Delta_o \approx 21{,}000\ \text{cm}^{-1}$:

A ratio of order 0.1–0.2 is physically reasonable: the Jahn-Teller stabilization is a genuine but secondary perturbation on top of the primary octahedral crystal field, consistent with the general rule that Jahn-Teller splittings ($10^3$ cm$^{-1}$ scale) are typically 5–20% of $\Delta_o$ itself. This also explains why the octahedral CFSE framework (from Crystal Field Stabilization Energy (CFSE)) remains the right zeroth-order picture, with Jahn-Teller effects entering as a perturbative correction — exactly the same logic used when Square-Planar Crystal Field Splitting is derived as the extreme limit of tetragonal distortion carried to completion (axial ligands removed entirely).

Limiting-case note. If the two observed bands coincided ($E_2 - E_1 \to 0$), the analysis would give $\delta_1 \to 0$ and $\Delta E_{JT}\to 0$, correctly recovering the un-split, symmetric-octahedron spectral signature — i.e., no detectable Jahn-Teller splitting, which would (correctly) flag either a non-Jahn-Teller-active ion or a case where the distortion is dynamic/averaged rather than static (see the table below).