Crystal field stabilization energy and the spectrochemical series

Intuition [Beginner]

In an octahedral complex the five $d$-orbitals split into two energy levels: a lower trio $t_{2g}$ (three orbitals pointing between the ligands) and an upper pair $e_g$ (two orbitals pointing at the ligands). The size of the gap between them, written ${\Delta }_o$, is the central parameter of crystal-field theory. The question this unit answers is what an electron's choice of orbital costs or saves, and what determines the size of ${\Delta }_o$ for a particular metal-ligand combination.

Three concrete payoffs come out of the answer. First, a number called the crystal-field stabilisation energy (CFSE) measures how much energy the splitting saves the electrons relative to the unsplit baseline. Each electron in the lower trio saves $0.4{\Delta }_o$ worth of energy; each electron in the upper pair costs $0.6{\Delta }_o$. Add them up, and you get the total CFSE for a given d-electron configuration.

Second, when the metal has between four and seven d-electrons, the electrons face a real choice. They can pile into the lower trio and pair up there, paying a pairing penalty $P$ but saving the ${\Delta }_o$ gap. Or they can spread out into the upper pair, paying the gap but avoiding the pairing penalty. Whichever choice has the lower total energy wins. Large ${\Delta }_o$ favours pairing (low-spin); small ${\Delta }_o$ favours spreading (high-spin). Iron(II) bound to six water molecules has four unpaired electrons and is high-spin; iron(II) bound to six cyanide ions has zero unpaired electrons and is low-spin. Same metal, same oxidation state, same d-count — only the ligand has changed.

Third, the ligands themselves can be ranked. From the smallest ${\Delta }_o$ producer to the largest:

iodide $<$ bromide $<$ chloride $<$ fluoride $<$ hydroxide $<$ water $<$ ammonia $<$ ethylenediamine $<$ cyanide $<$ carbon monoxide.

This ranking is the spectrochemical series. It is one of the most useful empirical orderings in chemistry: predict where a ligand sits in the series, predict its ${\Delta }_o$, predict the resulting colour and magnetism. The ranking ignores oxidation state and metal identity, capturing only the ligand contribution.

Why is the ranking what it is? The simple electrostatic picture (ligand point charges repelling d-electrons) gets the direction right but not the magnitude, and it cannot explain why a neutral CO molecule produces a larger gap than a doubly charged oxide ion. The fuller story uses molecular-orbital theory: ligands donate electron density through a sigma channel (raising the upper pair) and exchange electron density through pi channels (raising or lowering the lower trio, depending on whether the pi orbitals on the ligand are filled or empty). The result is a sensible three-way classification of ligands as sigma-donors, pi-donors, and pi-acceptors.

A fourth idea ties everything together: when the splitting pattern leaves the highest-energy occupied orbital partly filled in a way that has an orbital degeneracy, the geometry itself becomes unstable. The molecule distorts — bonds stretch unequally — until the degeneracy is lifted and the energy is lowered. This is the Jahn-Teller distortion (1937), and it is the reason essentially every copper(II) complex is not a perfect octahedron but an axially elongated one. The intermediate and master tiers below develop each of these ideas with the supporting computation.

Visual [Beginner]

The Beginner picture has two halves. On the left, the canonical two-up-three-down splitting pattern in an octahedral complex with the gap ${\Delta }_o$ marked. On the right, a sketch of the spectrochemical series running from weak-field ligands at the bottom (small ${\Delta }_o$) to strong-field ligands at the top (large ${\Delta }_o$).

The picture conveys two facts at once: the universal geometric splitting pattern and the ligand-dependent magnitude that governs colour and magnetism.

Worked example [Beginner]

Consider iron(II) in two different complexes: aqueous $[\mathrm{Fe}({\mathrm{H}}_2\mathrm{O}{)}_6{]}^{2+}$ and the cyanide complex $[\mathrm{Fe}(\mathrm{CN}{)}_6{]}^{4-}$. Both have iron in the $+2$ oxidation state, both have $d^6$ configuration, both have octahedral geometry. The difference is the ligand.

Step 1. Count $d$-electrons. Iron(II) is $[\mathrm{Ar}]3d^6$ once the two $4s$ electrons are removed. So six $d$-electrons in both complexes.

Step 2. Place each ligand on the spectrochemical series. Water sits roughly in the middle. Cyanide sits at the top — it is one of the strongest-field ligands known. So ${\Delta }_o$ for the cyanide complex is far larger than for the aqua complex.

Step 3. Decide high-spin vs low-spin. For the aqua complex, ${\Delta }_o<P$, so high-spin: electrons fill the lower trio with three (one each, parallel spins by Hund), then the upper pair with two (one each, parallel), then the sixth electron pairs in the lower trio. Result: $t_{2g}^4{e}_g^2$, four unpaired electrons. For the cyanide complex, ${\Delta }_o>P$, so low-spin: all six electrons pile into the lower trio with three pairs. Result: $t_{2g}^6{e}_g^0$, zero unpaired electrons.

Step 4. Compute the CFSE. High-spin aqua: CFSE $=4\cdot (0.4{\Delta }_o)-2\cdot (0.6{\Delta }_o)=1.6{\Delta }_o-1.2{\Delta }_o=0.4{\Delta }_o$. Low-spin cyanide: CFSE $=6\cdot (0.4{\Delta }_o)=2.4{\Delta }_o$, minus the cost of two extra electron pairings, i.e. CFSE $=2.4{\Delta }_o-2P$.

Step 5. Predict magnetism. The aqua complex has four unpaired electrons; predicted spin-only moment $\sqrt{4\cdot 6}{\mu }_B=\sqrt{24}{\mu }_B\approx 4.9{\mu }_B$. The cyanide complex has zero unpaired electrons; it is diamagnetic. Observation matches both predictions.

What this tells us: with only the d-count, the ligand position on the spectrochemical series, and the CFSE formula, one can predict whether a given complex is high-spin or low-spin, count its unpaired electrons, and compute its energetic stabilisation from the ligand field. Two complexes with the same metal and the same d-count can have entirely different colours and magnetism — driven by ligand identity alone.

Check your understanding [Beginner]

Formal definition [Intermediate+]

Let $M$ be a transition-metal ion at the origin of ${\mathbb{R}}^3$ surrounded by six identical ligands at $(\pm a,0,0)$, $(0,\pm a,0)$, $(0,0,\pm a)$. From the analysis in 16.03.02 pending, the five real d-orbitals decompose under $O_h$ as $V_d=E_g\oplus T_{2g}$, with the two-dimensional $E_g$ block carrying $\{d_{z^2},d_{x^2-y^2}\}$ and the three-dimensional $T_{2g}$ block carrying $\{d_{xy},d_{xz},d_{yz}\}$. The first-order energy splitting raises $E_g$ by $+\frac{3}{5}{\Delta }_o$ and lowers $T_{2g}$ by $-\frac{2}{5}{\Delta }_o$ relative to the spherically symmetric baseline, satisfying the barycentre rule $2(+\frac{3}{5}{\Delta }_o)+3(-\frac{2}{5}{\Delta }_o)=0$.

Definition (Crystal-field stabilisation energy, CFSE). Let a $d^n$ configuration in an octahedral field occupy the lower-trio set $T_{2g}$ with $n_t$ electrons and the upper-pair set $E_g$ with $n_e$ electrons, where $n_t+n_e=n$. The crystal-field stabilisation energy is

$$\mathrm{CFSE}=-n_t\cdot \frac{2}{5}{\Delta }_o+n_e\cdot \frac{3}{5}{\Delta }_o,$$

measured as the lowering (negative number means stabilised, but for tabulation the sign convention often follows magnitudes: $\text{CFSE}=0.4n_t{\Delta }_o-0.6n_e{\Delta }_o$ with the understanding that positive values are stabilisation). When the configuration involves more pairings than the free-ion ground state would require, a pairing-energy correction $+p\cdot P$ is added, with $p$ counting the number of extra spin pairs created relative to the spherically symmetric reference and $P$ the pairing energy per pair (consisting of Coulomb-repulsion and lost-exchange contributions).

Definition (Pairing energy). Following Griffith and Ballhausen, the pairing energy $P$ for a $d^n$ configuration is decomposed as $P=P_C+P_E$, where $P_C$ is the Coulomb-repulsion contribution from forcing two electrons into the same spatial orbital and $P_E$ is the exchange-energy loss from no longer having two parallel-spin electrons in different orbitals. Both terms scale with the Racah parameter $B$ (a measure of inter-electron repulsion in the partly filled d-shell); typical first-row values are $P\approx 17,000$ to $28,000$ cm${}^{-1}$.

Definition (High-spin vs low-spin boundary). For $d^n$ with $4\le n\le 7$ in an octahedral field, the high-spin configuration places electrons in $e_g$ before pairing in $t_{2g}$, paying ${\Delta }_o$ per electron promoted; the low-spin configuration pairs in $t_{2g}$ before populating $e_g$, paying $P$ per extra pair. The crossover satisfies ${\Delta }_o=P$:

• ${\Delta }_o>P$ $\to$ low-spin energetically preferred;
• ${\Delta }_o<P$ $\to$ high-spin energetically preferred.

The crossover is sharpest for $d^5$ and $d^6$, where the number of unpaired electrons drops most dramatically (from 5 to 1 for $d^5$; from 4 to 0 for $d^6$).

Definition (Spectrochemical series). The spectrochemical series is the empirical ordering of ligands by the magnitude of ${\Delta }_o$ they produce at a fixed metal-ion centre. In the standard form,

$${\mathrm{I}}^{-}<{\mathrm{Br}}^{-}<{\mathrm{S}}^{2-}<{\mathrm{SCN}}^{-}<{\mathrm{Cl}}^{-}<{\mathrm{N}{\mathrm{O}}_3}^{-}<{\mathrm{F}}^{-}<{\mathrm{OH}}^{-}<{\mathrm{H}}_2\mathrm{O}<{\mathrm{NCS}}^{-}<\mathrm{N}{\mathrm{H}}_3<\mathrm{en}<{\mathrm{N}{\mathrm{O}}_2}^{-}<\mathrm{phen}<{\mathrm{CN}}^{-}<\mathrm{CO}.$$

A complementary series for metal-ion centres orders ions by ${\Delta }_o$ at fixed ligand: increasing oxidation state and going down a group both increase ${\Delta }_o$. Roughly, ${\Delta }_o(\text{4d})\approx 1.5{\Delta }_o(\text{3d})$ and ${\Delta }_o(\text{5d})\approx 2{\Delta }_o(\text{3d})$ at the same oxidation state and ligand.

Definition (Jahn-Teller theorem). Any non-linear molecule whose electronic ground state belongs to a degenerate irreducible representation of its point group is geometrically unstable: there exists a nuclear-coordinate distortion whose displacement breaks the symmetry, lifts the orbital degeneracy, and lowers the total electronic energy. The distortion mode must transform as an irrep contained in the symmetric square of the degenerate electronic irrep. Kramers degeneracy (odd-electron-number spin degeneracy) is exempt, because spin-only degeneracy is robust to geometric distortion.

Counterexamples to common slips

• CFSE is not the only energetic factor. CFSE measures the splitting contribution to the d-electron energy; the full thermodynamics of complex formation also includes hydration energies, lattice energies, and the spherical part of the metal-ligand attraction. CFSE plotted across $d^0$ through $d^{10}$ explains the deviations from a smooth trend, not the trend itself. The Irving-Williams series of ${\mathrm{M}}^{2+}$ stability ordering (${\mathrm{Mn}}^{2+}<{\mathrm{Fe}}^{2+}<{\mathrm{Co}}^{2+}<{\mathrm{Ni}}^{2+}<{\mathrm{Cu}}^{2+}>{\mathrm{Zn}}^{2+}$) is the cleanest empirical manifestation of CFSE effects.

• The spectrochemical series does not always rank by ligand charge. The strongest-field ligand on the standard list is CO, a neutral molecule, while one of the weakest is the doubly charged sulfide ion ${\mathrm{S}}^{2-}$. The series ranks by pi-bonding character, not Coulomb attraction. A purely electrostatic CFT would predict the opposite ordering; the LFT extension with pi-acceptors at the top is required to reproduce the observed series.

• Tetrahedral fields invert the splitting and forbid low-spin in practice. In ${\mathrm{T}}_d$ geometry, the e set sits below the $t_2$ set, and ${\Delta }_t\approx \frac{4}{9}{\Delta }_o$. Because ${\Delta }_t$ is small (~4/9 of an already smaller ${\Delta }_o$, plus only four ligands contribute), ${\Delta }_t<P$ in essentially every tetrahedral complex, so tetrahedral complexes are essentially always high-spin. The single counterexample family — low-spin tetrahedral — requires highly specialised pi-acceptor ligands and remains rare.

• Jahn-Teller distortions in $T_{2g}$ ground terms are weak. The Jahn-Teller theorem predicts distortion for any orbitally degenerate ground term, but the magnitude depends on the irrep. $E_g$-degenerate terms ($d^9$, $d^4$ HS, $d^7$ LS) distort strongly because $E_g$ orbitals point at ligands and feel the distortion directly. $T_{2g}$-degenerate terms ($d^1$, $d^2$, $d^5$ LS, $d^6$ LS at the boundary) distort weakly because $T_{2g}$ orbitals point between ligands. Observed distortions in $T_{2g}$ ground terms are often dynamically averaged out and look as if the geometry is undistorted on a thermal-averaging timescale.

Key theorem with proof [Intermediate+]

Theorem (CFSE and the high-spin / low-spin boundary). Let a $d^n$ ion sit in an octahedral field with splitting parameter ${\Delta }_o$ and pairing energy $P$. Define the high-spin configuration as the one maximising the total spin $S$ subject to the constraint that lower-energy orbitals are filled first only insofar as Hund's rule allows; define the low-spin configuration as the one minimising the total energy of the d-electrons by preferentially pairing in $t_{2g}$. Then for $4\le n\le 7$:

(i) The high-spin configuration has total d-electron energy $$ E_{\mathrm{HS}}(n) = -,n_t^{\mathrm{HS}} \cdot \tfrac{2}{5}\Delta_o + n_e^{\mathrm{HS}} \cdot \tfrac{3}{5}\Delta_o + p^{\mathrm{HS}} P, $$ where $(n_t^{\mathrm{HS}},n_e^{\mathrm{HS}},p^{\mathrm{HS}})$ takes values $(3,1,0)$, $(3,2,0)$, $(4,2,1)$, $(5,2,2)$ for $n=4,5,6,7$.

(ii) The low-spin configuration has total d-electron energy $$ E_{\mathrm{LS}}(n) = -,n_t^{\mathrm{LS}} \cdot \tfrac{2}{5}\Delta_o + n_e^{\mathrm{LS}} \cdot \tfrac{3}{5}\Delta_o + p^{\mathrm{LS}} P, $$ where $(n_t^{\mathrm{LS}},n_e^{\mathrm{LS}},p^{\mathrm{LS}})$ takes values $(4,0,1)$, $(5,0,2)$, $(6,0,3)$, $(6,1,3)$ for $n=4,5,6,7$.

(iii) The low-spin configuration is energetically preferred precisely when ${\Delta }_o>P$.

Proof. The first-order energy of an electron in $t_{2g}$ is $-\frac{2}{5}{\Delta }_o$ and in $e_g$ is $+\frac{3}{5}{\Delta }_o$ relative to the spherical baseline (the barycentre, see 16.03.02 pending). The total orbital-population contribution to the energy is therefore ${\sum }_i{ϵ}_i=-n_t\cdot \frac{2}{5}{\Delta }_o+n_e\cdot \frac{3}{5}{\Delta }_o$. Pairing-energy additivity contributes $+pP$ where $p$ is the number of doubly occupied spatial orbitals counted relative to the spherical baseline (the baseline has the maximum-multiplicity Hund filling of the free-ion ground term).

The high-spin filling rule for $4\le n\le 7$: place electrons one per orbital across all five d-orbitals respecting Hund's rule, then pair up additional electrons starting in the lowest-energy ($t_{2g}$) orbital. Counting:

• $n=4$: $t_{2g}^3{e}_g^1$, four unpaired, no extra pairs $\to (n_t,n_e,p)=(3,1,0)$.
• $n=5$: $t_{2g}^3{e}_g^2$, five unpaired, no extra pairs $\to (3,2,0)$.
• $n=6$: $t_{2g}^4{e}_g^2$, four unpaired, one extra pair (the sixth electron pairs in $t_{2g}$) $\to (4,2,1)$.
• $n=7$: $t_{2g}^5{e}_g^2$, three unpaired, two extra pairs $\to (5,2,2)$.

The low-spin filling rule: fill $t_{2g}$ completely (three pairs) before populating $e_g$. Counting:

• $n=4$: $t_{2g}^4{e}_g^0$, two unpaired, one extra pair $\to (4,0,1)$.
• $n=5$: $t_{2g}^5{e}_g^0$, one unpaired, two extra pairs $\to (5,0,2)$.
• $n=6$: $t_{2g}^6{e}_g^0$, zero unpaired, three extra pairs $\to (6,0,3)$.
• $n=7$: $t_{2g}^6{e}_g^1$, one unpaired, three extra pairs $\to (6,1,3)$.

This establishes (i) and (ii).

For (iii), compute the energy difference $E_{\mathrm{HS}}(n)-E_{\mathrm{LS}}(n)$ for each $n$:

$$\begin{array}{rl}n=4:& E_{\mathrm{HS}}-E_{\mathrm{LS}}=[(-3\cdot \frac{2}{5}+1\cdot \frac{3}{5}){\Delta }_o+0]-[(-4\cdot \frac{2}{5}){\Delta }_o+P]\\ & =(-\frac{6}{5}+\frac{3}{5}+\frac{8}{5}){\Delta }_o-P={\Delta }_o-P.\end{array}$$ $$\begin{array}{rl}n=5:& E_{\mathrm{HS}}-E_{\mathrm{LS}}=[(-\frac{6}{5}+\frac{6}{5}){\Delta }_o]-[-2\cdot {\Delta }_o+2P]\\ & =0+2{\Delta }_o-2P=2({\Delta }_o-P).\end{array}$$ $$\begin{array}{rl}n=6:& E_{\mathrm{HS}}-E_{\mathrm{LS}}=[(-\frac{8}{5}+\frac{6}{5}){\Delta }_o+P]-[-\frac{12}{5}{\Delta }_o+3P]\\ & =(-\frac{2}{5}+\frac{12}{5}){\Delta }_o-2P=2{\Delta }_o-2P=2({\Delta }_o-P).\end{array}$$ $$\begin{array}{rl}n=7:& E_{\mathrm{HS}}-E_{\mathrm{LS}}=[(-\frac{10}{5}+\frac{6}{5}){\Delta }_o+2P]-[(-\frac{12}{5}+\frac{3}{5}){\Delta }_o+3P]\\ & =(-\frac{4}{5}+\frac{9}{5}){\Delta }_o-P={\Delta }_o-P.\end{array}$$

In every case, $E_{\mathrm{HS}}-E_{\mathrm{LS}}$ is a positive multiple of ${\Delta }_o-P$. So $E_{\mathrm{HS}}>E_{\mathrm{LS}}$ — that is, low-spin is preferred — precisely when ${\Delta }_o>P$. The crossover boundary is ${\Delta }_o=P$ in all four cases, with the magnitude of the energy difference largest at $n=5$ and $n=6$ (factor 2) and smaller at $n=4$ and $n=7$ (factor 1). This explains why spin-crossover materials (sharp transitions in the unpaired-electron count) are typically iron(II) ($d^6$) or iron(III) ($d^5$) systems near the crossover.

$\square$

Bridge. The CFSE / pairing-energy boundary builds toward 16.04.01 coordination chemistry, where the ligand-substitution thermodynamics depend on the CFSE difference between starting and product complexes, and appears again in 16.06.01 bioinorganic chemistry, where heme proteins exploit the $d^6$ Fe(II) spin-crossover boundary to couple oxygen-binding to a large geometric change. The foundational reason CFSE explains the Irving-Williams series of M(II) hydrate-formation constants is exactly that the $d^3,d^8,d^9$ configurations (Cr, Ni, Cu) have the largest CFSE values for their row, and putting these together with the additivity of the pairing-energy correction identifies the spin-state energy with the linear functional $E(n_t,n_e,p)=-0.4n_t{\Delta }_o+0.6n_e{\Delta }_o+pP$. The bridge is from this single-ion local energy to the lattice-extended Hubbard model of correlated electrons in transition-metal-oxide solids, where the same CFSE pattern controls the on-site occupation.

Exercises [Intermediate+]

CFSE quantification: high-spin vs low-spin and the energetics of d-electron configurations [Master]

The Key Theorem of this unit establishes the CFSE formula and the crossover boundary ${\Delta }_o=P$. The master tier develops the consequences of that boundary across the periodic table and unifies it with the broader thermodynamic and kinetic picture of d-electron systems.

The CFSE landscape. Tabulating CFSE for all octahedral $d^n$ high-spin and low-spin configurations gives the double-humped pattern of Exercise 9. The maxima sit at $d^3$ ($t_{2g}^3{e}_g^0$, CFSE = $1.2{\Delta }_o$) and $d^8$ ($t_{2g}^6{e}_g^2$, CFSE = $1.2{\Delta }_o$); the zeroes sit at $d^0$, $d^5$ (HS), and $d^{10}$. The $d^5$ HS zero is the cleanest organic-chemistry-style "the trend evidently breaks" signature: high-spin Mn(II) and Fe(III) complexes are notably weaker complex-formers than would be expected from ionic radius alone, because the CFSE that stabilises every other M(II) hydrate is absent at $d^5$. The Irving-Williams series ordering ${\mathrm{Mn}}^{2+}<{\mathrm{Fe}}^{2+}<{\mathrm{Co}}^{2+}<{\mathrm{Ni}}^{2+}<{\mathrm{Cu}}^{2+}>{\mathrm{Zn}}^{2+}$ for divalent first-row hydrate stabilities is exactly the CFSE pattern modulated by ionic-radius effects; the maximum at Cu${}^{2+}$ ($d^9$, CFSE = $0.6{\Delta }_o$) is reinforced by the Jahn-Teller stabilisation that comes from the tetragonal distortion (an additional ~$1000$ cm${}^{-1}$ of stabilisation typical for Cu(II) hexaaqua).

Pairing energy across the series. The pairing energy $P$ has two contributions, both scaling with the Racah parameter $B$ that measures inter-electron repulsion within the d-shell. The Coulomb contribution $P_C$ is the energetic cost of forcing two electrons into the same spatial orbital; the exchange contribution $P_E$ is the loss of exchange stabilisation that comes from no longer having the maximally parallel-spin configuration. Both are positive (destabilising) and add. Typical first-row TM values are $P\approx 17,000$ to $28,000$ cm${}^{-1}$, comparable to the ${\Delta }_o$ range $\approx 10,000$ to $30,000$ cm${}^{-1}$.

The crucial fact is that $P$ decreases substantially going from 3d to 4d to 5d. The 4d and 5d orbitals are spatially larger, so inter-electron repulsion within them is smaller, and $B$ drops by roughly a factor of two between 3d and 5d. Meanwhile ${\Delta }_o$ increases by about 50 percent (3d to 4d) and 100 percent (3d to 5d), because the larger 4d/5d orbitals overlap more effectively with the ligand orbitals. The combined effect — smaller $P$, larger ${\Delta }_o$ — makes second-row and third-row TM complexes essentially always low-spin. Examples: $[\mathrm{Rh}(\mathrm{N}{\mathrm{H}}_3{)}_6{]}^{3+}$ ($4d^6$) and $[\mathrm{Ir}(\mathrm{N}{\mathrm{H}}_3{)}_6{]}^{3+}$ ($5d^6$) are both diamagnetic low-spin in the same conditions where Co(III) is borderline; $[\mathrm{Pd}(\mathrm{Cl}{)}_4{]}^{2-}$ and $[\mathrm{Pt}(\mathrm{Cl}{)}_4{]}^{2-}$ are diamagnetic square-planar where Ni(II) is paramagnetic tetrahedral. The 4d/5d uniformly-low-spin pattern is a primary feature of the heavier-d-block organometallic and catalytic chemistry.

Spin crossover. Some first-row complexes sit so close to the ${\Delta }_o=P$ boundary that thermal energy can tip them between high-spin and low-spin states. The canonical examples are Fe(II) complexes with intermediate-field ligands: $[\mathrm{Fe}(\mathrm{phen}{)}_2(\mathrm{NCS}{)}_2]$ is high-spin above 176 K and low-spin below it, with the transition driven by entropy (the high-spin state has more accessible vibrational microstates and dominates at high $T$). Spin-crossover materials are technologically important for switchable molecular magnets and pressure-sensitive sensors; the parameters that tune the transition are the ligand-field strength (${\Delta }_o$ modulated by ligand identity and bond length) and the pairing energy ($P$ modulated by configuration). The full thermodynamics of the transition is governed by an Ising-style two-state model with cooperative interactions between neighbouring metal sites in the solid state.

The Tanabe-Sugano diagram for $d^6$ shows the term-energy crossing at ${\Delta }_o/B\approx 20$ that marks the high-spin-to-low-spin transition; physical complexes near this crossing are spin-crossover candidates. The full development of these diagrams uses the configuration-interaction methodology of Racah and Tanabe-Sugano, with the inter-electron-repulsion parameters $B$ and $C$ (related by $C\approx 4B$ approximately) supplementing ${\Delta }_o$ as independent inputs.

The spectrochemical series and ligand-field perturbation theory [Master]

The empirical spectrochemical-series ordering is one of the most-replicated experimental regularities in inorganic chemistry. The Crystal-Field-Theory level of treatment cannot account for it: a pure electrostatic point-charge model predicts that ligand charge and bond length should be the only relevant variables, which would put ${\mathrm{O}}^{2-}>{\mathrm{F}}^{-}>{\mathrm{OH}}^{-}>{\mathrm{H}}_2\mathrm{O}$ at the top. Observation is the reverse for the top of the series — neutral CO produces the largest ${\Delta }_o$ — so the electrostatic CFT picture must be extended.

The sigma / pi classification of Schäffer and Jørgensen. The 1965 paper of Schäffer and Jørgensen ^{[Schäffer-Jørgensen 1965]} reframed the spectrochemical series in MO-theoretic terms by classifying the M-L interaction into a sigma channel (the M-L bond axis) and a pi channel (perpendicular to the bond). For an octahedral complex:

• The six ligand sigma-donor lone pairs combine into SALCs spanning $A_{1g}\oplus E_g\oplus T_{1u}$ in $O_h$. The $E_g$ SALC pairs with metal $E_g$ ($d_{z^2},d_{x^2-y^2}$) to form a $\sigma$-bonding pair (mostly ligand) and a $\sigma$-antibonding pair (mostly metal). The metal $T_{2g}$ ($d_{xy},d_{xz},d_{yz}$) has no sigma partner among the ligand SALCs — sigma-only ligands leave $T_{2g}$ formally non-bonding.

• The twelve ligand pi orbitals (two per ligand, perpendicular to each M-L axis) combine into SALCs spanning $T_{1g}\oplus T_{2g}\oplus T_{1u}\oplus T_{2u}$ in $O_h$. The $T_{2g}$ portion pairs with metal $T_{2g}$ to form a pi-bonding pair (mostly the lower-energy partner) and a pi-antibonding pair (mostly the higher-energy partner).

The three-way classification follows:

• Sigma-only ligands (saturated amines like NH${}_3$ and en; alkyl): only the sigma channel operates. Metal $T_{2g}$ stays non-bonding; metal $E_g$ rises by the sigma-antibonding splitting. ${\Delta }_o$ is moderate, set by the sigma overlap alone.

• Sigma-donor + pi-donor ligands (halides, oxide, sulfide, hydroxide): both channels operate, with the pi channel using filled ligand pi orbitals that lie below the metal $T_{2g}$. The pi interaction pushes metal $T_{2g}$ upward, reducing ${\Delta }_o$. The strength of the pi-donor effect increases with the filled-pi-orbital energy of the ligand: heavier halides (more diffuse, higher-energy lone pairs) are stronger pi-donors than fluoride, explaining why ${\mathrm{I}}^{-}$ sits below ${\mathrm{F}}^{-}$ in the series despite being a weaker electrostatic donor.

• Sigma-donor + pi-acceptor ligands (CO, CN${}^{-}$, NO, bipyridine, phenanthroline, NO${}_2^{-}$): the pi channel uses empty ligand ${\pi }^{\ast }$ orbitals that lie above the metal $T_{2g}$. Back-donation from filled metal $T_{2g}$ into ligand ${\pi }^{\ast }$ stabilises the metal $T_{2g}$ (pushes it downward), increasing ${\Delta }_o$. The signature is bond-order reduction in the ligand: CO coordinated as a terminal carbonyl shows a C-O stretch shifted from $2143$ cm${}^{-1}$ (free CO) to $\approx 2000$ cm${}^{-1}$ (Cr(CO)${}_6$ to ~$1900$ cm${}^{-1}$ in lower-oxidation-state carbonyls), reflecting partial ${\pi }^{\ast }$ occupation and consequent loss of C-O bond order.

The empirical series direction — halides at the bottom, sigma-only amines in the middle, pi-acceptors at the top — is the direct translation of this MO classification into the ${\Delta }_o$ ordering.

The angular overlap model (AOM). A quantitative parameterisation due to Schäffer in the late 1960s ^{[Schäffer 1968]} assigns each metal-ligand pair two parameters $e_{\sigma }$ (sigma overlap, always positive) and $e_{\pi }$ (pi overlap; positive for donors, negative for acceptors). The energy of each d-orbital in the complex is then a sum of contributions from each ligand position, weighted by geometric factors that follow from the projection of the M-L overlap onto the d-orbital basis. For the octahedral case, the result is:

• $E(e_g)=3e_{\sigma }$ (each $e_g$ orbital has cumulative sigma weight 3 over the six ligand positions).
• $E(t_{2g})=4e_{\pi }$ (each $t_{2g}$ orbital has cumulative pi weight 4).
• ${\Delta }_o=E(e_g)-E(t_{2g})=3e_{\sigma }-4e_{\pi }$.

For sigma-only ligands ($e_{\pi }=0$): ${\Delta }_o=3e_{\sigma }$. For pi-donors ($e_{\pi }>0$): ${\Delta }_o$ is reduced. For pi-acceptors ($e_{\pi }<0$): ${\Delta }_o$ is increased.

AOM extends cleanly to non-octahedral geometries. For square-planar four-coordination, the four equatorial ligands contribute different geometric factors than they would in $O_h$: the result is a four-level pattern $a_{1g}<e_g<b_{2g}<b_{1g}$ with the gap between the highest-occupied and lowest-unoccupied (the LUMO-HOMO gap for $d^8$) considerably larger than ${\Delta }_o$ would be in the corresponding octahedral six-coordinate complex. This is the foundation of the structural preference of $d^8$ low-spin for square-planar geometry. For trigonal-bipyramidal five-coordination, AOM predicts a $a_1^{\prime }<e^{\prime }<e^{\prime \prime }$ pattern with axial vs equatorial ligand contributions weighted differently. The transferability of AOM parameters from one geometry to another (the same metal-ligand pair retains its $e_{\sigma },e_{\pi }$ values to good approximation) makes the model widely useful for predicting splittings in irregular coordination environments.

Modern computational refinement. Density-functional theory (DFT) calculations on transition-metal complexes can reproduce ${\Delta }_o$ values to within $\sim 0.1$ eV accuracy for many systems. Multiconfigurational methods (CASSCF, NEVPT2, density-matrix renormalisation group adaptations) are needed for systems near term crossings or with strong static correlation (spin crossover, intermediate-coupling regimes, heavy elements with strong spin-orbit). The state-of-the-art combines an active space of metal d-orbitals plus relevant ligand $\sigma$ and $\pi$ orbitals with second-order perturbation theory; the resulting energies match experimental absorption maxima for first-row TM complexes to within $1000$ cm${}^{-1}$ when carefully tuned.

The LFT / AOM framework remains the interpretive vocabulary even when the quantitative values come from DFT or CASSCF. A DFT-computed ${\Delta }_o$ is interpreted as a particular combination of sigma and pi metal-ligand interactions; the spectrochemical-series intuition that ranks ligands by combined $\sigma +\pi$ contribution is the conceptual scaffold around which the computational result is reported. Without LFT / AOM, the DFT energies would lack interpretive structure.

Jahn-Teller distortion and the lifting of orbital degeneracy [Master]

The Jahn-Teller theorem ^{[Jahn-Teller 1937]} addresses the geometric consequences of orbital degeneracy in non-linear molecules. Its statement and proof were among the earliest applications of full group-theoretic reasoning in molecular structure theory, and the theorem's implications continue to organise transition-metal coordination chemistry.

Statement. Let $G$ be the point group of a non-linear molecule, and let the electronic ground state belong to a degenerate irreducible representation $\Gamma$ of $G$ with $\dim \Gamma \ge 2$. Then there exists a nuclear-coordinate distortion mode that lowers the molecular symmetry, lifts the orbital degeneracy, and lowers the total electronic energy. The distortion mode must transform as an irrep $\gamma$ contained in the symmetric square ${\mathrm{Sym}}^2(\Gamma )$, and $\gamma$ must be different from the totally symmetric representation $A_{1g}$ (otherwise the mode is a breathing distortion that preserves symmetry).

Why the symmetric square? The first-order vibronic coupling between an electronic state of symmetry $\Gamma$ and a nuclear-coordinate distortion of symmetry $\gamma$ is non-zero precisely when $\Gamma \otimes \gamma \otimes \Gamma$ contains the totally symmetric representation. For a degenerate electronic state ($\dim \Gamma \ge 2$), this reduces (using the standard inner product on irrep representations and the property that $\Gamma \otimes \Gamma ={\mathrm{Sym}}^2(\Gamma )\oplus {\Lambda }^2(\Gamma )$) to the condition that $\gamma$ is contained in ${\mathrm{Sym}}^2(\Gamma )$ minus the totally symmetric part. The Jahn-Teller theorem then states that for non-linear molecules this set of allowed $\gamma$ is always non-empty for $\dim \Gamma \ge 2$ — a representation-theoretic existence result that does not depend on the specific molecule, only on its symmetry.

Application to octahedral complexes. The two octahedral irreps that produce degenerate ground states are $E_g$ (dimension 2) and $T_{2g}$ (dimension 3). Compute the symmetric squares:

• ${\mathrm{Sym}}^2(E_g)=A_{1g}\oplus E_g$: the $A_{1g}$ part is breathing; the $E_g$ part contains the symmetry-breaking tetragonal modes $Q_2$ and $Q_3$. So $E_g$-degenerate ground states distort along a tetragonal mode and lower the symmetry to $D_{4h}$.

• ${\mathrm{Sym}}^2(T_{2g})=A_{1g}\oplus E_g\oplus T_{2g}$: the $A_{1g}$ part is breathing; the $E_g$ part is tetragonal; the $T_{2g}$ part is a trigonal distortion. $T_{2g}$-degenerate ground states can in principle distort along either the tetragonal or the trigonal mode, but the energy gain is generally much smaller than for $E_g$ distortions because $T_{2g}$ orbitals point between the ligands and are less sensitive to bond-length changes.

Configurations and predictions.

• $E_g$ orbital degeneracy: $d^4$ HS ($t_{2g}^3{e}_g^1$), $d^7$ LS ($t_{2g}^6{e}_g^1$), $d^9$ ($t_{2g}^6{e}_g^3$). Strong J-T; tetragonal distortion of order $0.1$ to $0.3$ Å in bond-length differential.

• $T_{2g}$ orbital degeneracy: $d^1$ ($t_{2g}^1$), $d^2$ ($t_{2g}^2$), $d^4$ LS ($t_{2g}^4$), $d^5$ LS ($t_{2g}^5$). Weak J-T; distortions of order $0.01$ to $0.05$ Å, often dynamically averaged at room temperature.

• Closed-shell or orbitally non-degenerate: $d^3,d^5$ HS, $d^6$ LS, $d^8,d^{10}$ in $O_h$. Not J-T active.

The canonical example is Cu(II) $d^9=t_{2g}^6{e}_g^3$, with ground term ${}^2{E}_g$. The tetragonal distortion to $D_{4h}$ splits $e_g$ into $a_{1g}(d_{z^2})+b_{1g}(d_{x^2-y^2})$. The unpaired electron sits in $b_{1g}$, and the geometry is a tetragonally elongated octahedron with four short equatorial Cu-O bonds and two long axial ones. Cu(II) is so consistently distorted that "perfect octahedral Cu(II)" is a rare laboratory curiosity, found only when crystal packing forces an averaging over the three equivalent tetragonal axes.

Static vs dynamic Jahn-Teller. The Jahn-Teller theorem predicts that the molecule cannot rest at the perfectly symmetric octahedral geometry, but it does not specify whether the system locks into a single distorted minimum (static J-T) or tunnels coherently between equivalent distorted minima (dynamic J-T). For $E_g$ ground terms, the potential surface has three equivalent minima corresponding to elongation along $x$, $y$, or $z$; tunnelling among them at low temperature can produce a dynamic-J-T spectrum that looks octahedrally symmetric on slow timescales but tetragonally distorted on fast (vibrational) timescales. EPR spectroscopy at variable temperature distinguishes the two regimes: static-J-T spectra are axial and temperature-independent; dynamic-J-T spectra collapse to isotropic at high temperature and resolve to axial at low temperature.

Pseudo Jahn-Teller (second-order JT). Even when the ground state is orbitally non-degenerate, a low-lying excited state of compatible symmetry can produce a second-order vibronic coupling that drives a distortion. This pseudo Jahn-Teller effect was systematised by Öpik and Pryce (1957) and Bersuker (1980s onward) ^{[Bersuker 2010]}. It explains many subtle structural features — the bent geometry of $\mathrm{X}{\mathrm{Y}}_2$ molecules with ${\mathrm{n}}_{\text{val}}=16$ to 20 valence electrons (e.g., $\mathrm{N}{\mathrm{H}}_2$, $\mathrm{N}{\mathrm{O}}_2$); the puckered shape of cyclic conjugated systems near a HOMO-LUMO crossing; the off-centre displacement of heavy ions in ferroelectric perovskites (BaTiO${}_3$, PbTiO${}_3$). The pseudo Jahn-Teller effect is the structural analogue of the same vibronic coupling that drives Peierls distortions in one-dimensional metals.

Bridge to organic stereochemistry. The Jahn-Teller theorem and its pseudo extension generalise to organic systems: an organic molecule in a degenerate $\pi$-electron configuration distorts to break the degeneracy. The classic example is cyclobutadiene (${\mathrm{C}}_4{\mathrm{H}}_4$), whose square geometry would give a degenerate $e_g$ HOMO with two electrons; the molecule distorts to a rectangle (alternating bond lengths), breaking the degeneracy to a non-degenerate $b_{2g}$ HOMO. The full classification of degeneracy-driven distortions across organic, inorganic, and solid-state systems is the modern subject called vibronic coupling theory, with Bersuker's monograph providing the canonical reference ^{[Bersuker 2010]}.

Connection to UV-vis spectroscopy: d-d vs charge-transfer transitions [Master]

The colour of a transition-metal complex is the result of selective absorption of visible light by the d-electron system. The visible spectrum spans roughly $14,000$ to $25,000$ cm${}^{-1}$ ($400$ to $700$ nm); ${\Delta }_o$ for first-row TM complexes falls in roughly the same range. The pattern of absorption bands and their intensities reveals the metal-ligand interaction in detail.

Selection rules. Three rules govern electric-dipole transitions in centrosymmetric octahedral complexes:

1. Laporte rule. Transitions between states of the same parity (gerade-to-gerade or ungerade-to-ungerade) are electric-dipole-forbidden. d-orbitals are all gerade, so d-d transitions are Laporte-forbidden. They become weakly allowed by vibronic coupling to odd-parity vibrations that temporarily destroy the inversion centre. Observed molar absorption coefficients $\epsilon \sim 1$ to $100$ M${}^{-1}$ cm${}^{-1}$, typically two to three orders of magnitude weaker than fully allowed transitions.

2. Spin rule. Transitions cannot change the total spin: $\Delta S=0$. Spin-forbidden transitions become weakly allowed by spin-orbit-coupling-induced mixing of states of different multiplicity, with intensity scaling as $\sim {\zeta }^2/\Delta E^2$ where $\zeta$ is the spin-orbit constant and $\Delta E$ is the energy gap to a state of mixed multiplicity. Observed $\epsilon \sim 0.01$ to $1$ M${}^{-1}$ cm${}^{-1}$ for spin-forbidden d-d.

3. Symmetry rule. Even within the Laporte-allowed channel, the product of the irreps of the initial state, the transition operator (transforming as the position vector $T_{1u}$ in $O_h$), and the final state must contain the totally symmetric representation $A_{1g}$. This is the group-theoretic version of the dipole matrix element non-vanishing requirement.

For $d^5$ high-spin Mn(II), the ground term is ${}^6{A}_{1g}$ ($S=5/2$); every excited d-d term has lower multiplicity (quartets, doublets, etc.). Every d-d transition is therefore both Laporte- and spin-forbidden, giving the characteristic faint pink colour of solid Mn(II) salts and a $\epsilon \sim 0.01$ M${}^{-1}$ cm${}^{-1}$ in aqueous solution. By contrast, the spin-allowed but Laporte-forbidden ${}^4{A}_{2g}\to {}^4{T}_{2g}$ transition in $d^3$ Cr(III) has $\epsilon \approx 5$ to $20$ M${}^{-1}$ cm${}^{-1}$, producing the well-known deep colours of Cr(III) complexes.

Charge-transfer bands. Beyond d-d transitions, transition-metal complexes show charge-transfer (CT) bands that involve electron transfer between metal and ligand:

• Ligand-to-metal charge transfer (LMCT) transitions involve excitation from a filled ligand orbital to a partly filled metal d-orbital. The transition is Laporte-allowed (different-parity initial and final states), with $\epsilon \sim 10^3$ to $10^4$ M${}^{-1}$ cm${}^{-1}$. LMCT bands dominate the spectrum of high-oxidation-state metals with reducing ligands: ${\mathrm{Mn}{\mathrm{O}}_4}^{-}$ (deep purple, LMCT ${\mathrm{O}}^{2-}\to \mathrm{Mn}(\mathrm{VII})$), ${\mathrm{Cr}{\mathrm{O}}_4}^{2-}$ (yellow, LMCT ${\mathrm{O}}^{2-}\to \mathrm{Cr}(\mathrm{VI})$), $\mathrm{Fe}(\mathrm{III})$ thiocyanate (blood-red, LMCT ${\mathrm{SCN}}^{-}\to \mathrm{Fe}(\mathrm{III})$). The colour intensity of LMCT bands distinguishes these complexes from simple d-d-coloured species at a glance.

• Metal-to-ligand charge transfer (MLCT) transitions involve excitation from a filled metal d-orbital to an empty ligand orbital. The transition is also Laporte-allowed with $\epsilon \sim 10^3$ to $10^4$ M${}^{-1}$ cm${}^{-1}$. MLCT bands dominate the spectrum of low-oxidation-state metals with pi-acceptor ligands: $[\mathrm{Ru}(\mathrm{bpy}{)}_3{]}^{2+}$ (orange, MLCT $\mathrm{Ru}(\mathrm{II})\to \mathrm{bpy}{\pi }^{\ast }$, with photophysics that underpins modern dye-sensitised solar cells), $[\mathrm{Fe}(\mathrm{phen}{)}_3{]}^{2+}$ (red, MLCT $\mathrm{Fe}(\mathrm{II})\to \mathrm{phen}{\pi }^{\ast }$, the basis of the ferroin redox indicator), $\mathrm{Fe}(\mathrm{CO}{)}_5$ (LMCT and MLCT both present in the UV).

The distinction between LMCT and MLCT is diagnostic of the metal's oxidation state and the ligand's pi character: high-oxidation metals with electron-rich ligands give LMCT; low-oxidation metals with electron-poor pi-acceptor ligands give MLCT.

Tanabe-Sugano diagrams. The complete classification of d-d transitions for a given $d^n$ configuration in an octahedral field uses the Tanabe-Sugano diagrams ^{[Tanabe-Sugano 1954]}. For a $d^n$ ion, the free-ion Russell-Saunders terms ${}^{2S+1}L$ split under $O_h$ into Mulliken-labelled sub-terms, whose energies depend on the dimensionless ratio ${\Delta }_o/B$. The diagrams plot the energy of each sub-term relative to the ground term as a function of ${\Delta }_o/B$, with the inter-electron-repulsion Racah parameters $B$ and $C$ ($C\approx 4B$) supplying the off-diagonal mixing.

For $d^2$ (Ti(II), V(III)), the free-ion ground term is ${}^{3}F$ ($L=3,S=1$), splitting under $O_h$ into ${}^3{T}_{1g}(F){\oplus }^3{T}_{2g}{\oplus }^3{A}_{2g}$. The diagram plots all three triplet sub-terms (plus the higher singlet ${}^{1}G{,}^{1}D{,}^{1}S$ sub-terms) as ${\Delta }_o/B$ varies from 0 (free ion) to 50 (strong field). The ground term ${}^3{T}_{1g}(F)$ traces a particular curve; the first two spin-allowed excited terms ${}^3{T}_{2g}$ and ${}^3{T}_{1g}(P)$ produce two observed bands in the visible-near-UV. Assigning the bands to specific ${\Delta }_o/B$ values gives both ${\Delta }_o$ and $B$ for the complex.

For $d^5$ (Mn(II), Fe(III)), the ground term is the spin-sextet ${}^6{A}_{1g}$, and every excited term has lower multiplicity. The Tanabe-Sugano diagram for $d^5$ shows the high-spin-to-low-spin transition at ${\Delta }_o/B\approx 28$, marked by a sharp change in slope of the lowest term (the ground term switches from ${}^6{A}_{1g}$ at low field to ${}^2{T}_{2g}$ at high field). Spin-crossover materials sit near this crossing.

The seven Tanabe-Sugano diagrams for $d^2$ through $d^8$ remain the standard tool for assigning the visible spectra of first-row TM complexes. They are reproduced in every graduate inorganic-chemistry textbook and are taught alongside the spectrochemical series as the primary translation between absorption-spectrum data and structural conclusions.

Synthesis. The crystal-field stabilisation framework builds toward 16.06.01 bioinorganic chemistry, where the same ${\Delta }_o$, CFSE, and high-spin/low-spin machinery decides the function of heme iron, the copper site of cytochromes, and the zinc-finger binding of regulatory proteins, and appears again in 12.17.01 pending in the lattice extension where a periodic array of TM ions with on-site CFSE pattern becomes the multi-orbital Hubbard model whose Mott-insulator behaviour is the foundation of transition-metal-oxide magnetism. The foundational reason CFSE explains both the Irving-Williams aqua-stability ordering and the structural preferences of $d^8$ low-spin for square-planar geometry is exactly that it identifies the d-electron energy with the linear functional $-0.4n_{t_{2g}}{\Delta }_o+0.6n_{e_g}{\Delta }_o+pP$ over (configuration, geometry, ligand) triples.

Putting these together with the Jahn-Teller selection rule on the symmetric square, the central insight is that orbital degeneracy in non-linear systems is geometrically unstable: the molecule selects whichever distortion mode $\gamma \subseteq {\mathrm{Sym}}^2(\Gamma )\setminus A_{1g}$ lowers the energy most, lifting the degeneracy and identifying the new ground-state irrep with the symmetric-square component. The bridge is from the single-site CFT picture to the modern computational treatment (CASSCF / NEVPT2 / DFT) in which the same CFT vocabulary survives as the interpretive scaffold for the numerical result. The pattern generalises to organic systems via Hückel-Jahn-Teller in degenerate $\pi$-systems, to solid-state systems via vibronic Peierls distortions in low-dimensional metals, and to ferroelectric perovskites via pseudo-JT off-centring of heavy cations — each case a manifestation of the same group-theoretic mechanism that this unit's Master tier develops.

Connections [Master]

• Crystal-field splitting in octahedral complexes 16.03.02 pending. The direct prerequisite. The decomposition $V_d=E_g\oplus T_{2g}$ derived there is the input for the CFSE formula. The character-table machinery used to project $V_d$ onto its irreducible components builds toward the Tanabe-Sugano analysis of d-d transitions and the symmetric-square computation in the Jahn-Teller theorem of this unit.

• Coordination chemistry: geometries and isomerism 16.04.01. The parent topic of this unit. Werner's framework supplies the structural inputs (octahedral vs square-planar vs tetrahedral coordination, geometric and optical isomerism); the present unit supplies the electronic-structure inputs (CFSE, high-spin / low-spin, Jahn-Teller distortion) that decide which coordination geometry is energetically preferred. The chapter's structural-kinetic discussion of trans-effect substitution and Eigen-Wilkins mechanism uses the CFSE language developed here for the transition-state stabilisation argument.

• Bioinorganic chemistry: metalloenzymes 16.06.01. Hemoglobin, cytochromes, photosystem II, and other metalloprotein active sites have d-transition-metal centres whose function depends on the splitting pattern, spin state, and CFSE developed here. Oxygen binding to deoxyhemoglobin involves a $d^6$ Fe(II) spin-crossover from high-spin (deoxy) to low-spin (oxy), with the associated geometric change pulling the iron into the heme plane and triggering the cooperativity of the haemoglobin allosteric mechanism. Cytochrome electron transport relies on the $d^5\leftrightarrow d^6$ redox couple of low-spin Fe(III)/Fe(II) with redox potentials tuned by ligand field strength of the axial protein residues.

• Symmetry and group theory in chemistry 16.02.01. Supplies the point-group apparatus that this unit uses for the character-table reductions ($V_d=E_g\oplus T_{2g}$, ${\mathrm{Sym}}^2(E_g)=A_{1g}\oplus E_g$) and the irrep classification of vibrational modes that drive Jahn-Teller distortions. The Mulliken-notation conventions for the irreps of $O_h$ also originate from there.

• Atomic orbitals from H-atom QM 14.04.01 pending. Supplies the d-orbital basis itself — the five real $d_{z^2},d_{x^2-y^2},d_{xy},d_{xz},d_{yz}$ orbitals that the crystal-field perturbation acts on. Without the hydrogen-atom solutions and their angular-part Spherical-harmonic decomposition, the d-orbital basis would have to be introduced phenomenologically.

• Stern-Gerlach and spin-1/2 12.01.02 pending. Supplies the spin angular-momentum framework underlying the spin-only magnetic-moment formula $\mu =g_s\sqrt{S(S+1)}{\mu }_B$ used at intermediate tier and the spin-orbit-coupling discussion at master tier where it competes with the ligand-field splitting for 4d/5d series complexes.

• Organometallic 16- and 18-electron rules 16.05.01. The 18-electron rule is the closed-shell condition $t_{2g}^6$ (filled) + $\sigma$-bonding $a_{1g},e_g,t_{1u}$ (six pairs, mostly ligand-character) = 9 filled valence MOs = 18 electrons. The pi-acceptor ligands that dominate organometallic chemistry (CO, alkene, alkyne, phosphine) sit at the top of the spectrochemical series, justifying the systematic preference of organometallic catalysts for the 18-electron closed-shell configuration developed there.

• UV-Vis, IR, and NMR fundamentals 14.12.01. The crystal-field splitting ${\Delta }_o$ and the spin-allowed d-d transition structure developed here are the chromophoric basis for transition-metal-complex UV-Vis spectroscopy: Tanabe-Sugano diagrams parametrise the observed band positions in spectrochemical terms, the Laporte and spin selection rules set band intensities, and the magnetic moments derived from the spin-only formula and orbital contributions feed into the paramagnetic-shift analysis that distinguishes high-spin and low-spin assignments in solution NMR.

Historical & philosophical context [Master]

Crystal-field theory was developed by Hans Bethe in 1929 ^{[Bethe 1929]} in his Annalen der Physik paper on the splitting of atomic terms in crystals. Bethe applied the full apparatus of group-theoretic representation theory that had been worked out by Wigner, Schur, and others through the 1920s; his paper was among the first to use point-group character tables systematically for electronic-structure problems. The methodological move — that the energetic consequences of a chemical environment can be derived from the symmetry of that environment alone, prior to any detailed electrostatic calculation — became one of the foundational techniques of theoretical chemistry.

Van Vleck extended the theory through the 1930s to magnetic susceptibilities and paramagnetism ^{[Van Vleck 1932]}. His 1932 paper on paramagnetic anisotropy and the 1932 monograph The Theory of Electric and Magnetic Susceptibilities established the connection between crystal-field splittings and the observed magnetic behaviour of d- and f-block salts. Van Vleck also recognised that the pure-electrostatic CFT picture was incomplete: in his later work he introduced what would become ligand-field theory by acknowledging covalent admixture between metal d-orbitals and ligand orbitals. The Nobel Prize in Physics (1977) recognised his contributions to magnetism alongside Anderson and Mott.

The Jahn-Teller theorem appeared in 1937 ^{[Jahn-Teller 1937]} in the Proceedings of the Royal Society. Hermann Arthur Jahn was a young theoretical chemist working with Edward Teller in London; the paper presented the theorem in its full group-theoretic generality, with the symmetric-square selection rule for the distortion mode. The proof's central observation — that for non-linear molecules the set of allowed distortion irreps is non-empty for any degenerate ground state of dimension $\ge 2$ — relies only on the structure of finite-group representations, not on chemistry-specific input. The exception for linear molecules (which the theorem explicitly excludes) reflects the fact that the linear-molecule symmetry group $C_{\infty v}$ contains continuous rotations whose vibrational counterpart is not a normal-mode coordinate.

Tanabe and Sugano's 1954 papers in the Journal of the Physical Society of Japan ^{[Tanabe-Sugano 1954]} gave the systematic energy diagrams that now bear their name. Building on Racah's earlier work on the inter-electron-repulsion parameters $B$ and $C$ (Racah 1942), Tanabe and Sugano computed the full configuration-interaction problem for each $d^n$ configuration in an octahedral field and produced the universal plots of term energies vs ${\Delta }_o/B$. These diagrams remain the standard tool for assigning the visible-near-UV spectra of first-row TM complexes — a status they have held without significant modification for over seventy years.

The sigma / pi classification of ligands as donors and acceptors emerged from the Copenhagen school of inorganic chemistry around Schäffer and Jørgensen in the 1960s ^{[Schäffer-Jørgensen 1965, Schäffer 1968]}. The angular overlap model is the formal parameterisation of that picture; it has become the workhorse method for quantitative ligand-field analysis short of full DFT calculation. The contemporaneous emergence of organometallic chemistry as a major subfield (Wilkinson's catalyst 1965, the related cross-coupling chemistry of the 1970s) provided the experimental substrate on which the pi-acceptor / pi-donor classification was tested and refined.

The full LFT framework was consolidated by Ballhausen, Griffith, and Bersuker in monographs spanning the 1960s through 2010s — Griffith's 1961 Theory of Transition-Metal Ions, Ballhausen's 1962 Introduction to Ligand Field Theory, and Bersuker's 2010 Electronic Structure and Properties of Transition Metal Compounds (a successor and expansion of his earlier vibronic-coupling treatises). Cotton and Wilkinson's Advanced Inorganic Chemistry (six editions from 1962 to 1999) packaged the CFT / LFT framework into the standard graduate textbook treatment that organises most first-year-graduate inorganic curricula. The continuity from Bethe's 1929 paper through to the modern computational refinements is one of the clearest cases in chemistry of a theory whose interpretive vocabulary outlives several waves of methodological replacement — CFT remains the language even when DFT supplies the numbers.

Bibliography [Master]

@article{Bethe1929,
  author = {Bethe, Hans},
  title = {Termaufspaltung in Kristallen},
  journal = {Annalen der Physik},
  volume = {3},
  year = {1929},
  pages = {133--208},
}

@article{VanVleck1932,
  author = {Van Vleck, J. H.},
  title = {Theory of the variations in paramagnetic anisotropy among different salts of the iron group},
  journal = {Physical Review},
  volume = {41},
  year = {1932},
  pages = {208--215},
}

@book{VanVleck1932Book,
  author = {Van Vleck, J. H.},
  title = {The Theory of Electric and Magnetic Susceptibilities},
  publisher = {Oxford University Press},
  year = {1932},
}

@article{JahnTeller1937,
  author = {Jahn, H. A. and Teller, E.},
  title = {Stability of polyatomic molecules in degenerate electronic states. I. Orbital degeneracy},
  journal = {Proceedings of the Royal Society A},
  volume = {161},
  year = {1937},
  pages = {220--235},
}

@article{Racah1942,
  author = {Racah, G.},
  title = {Theory of complex spectra II},
  journal = {Physical Review},
  volume = {62},
  year = {1942},
  pages = {438--462},
}

@article{TanabeSugano1954,
  author = {Tanabe, Y. and Sugano, S.},
  title = {On the absorption spectra of complex ions, I and II},
  journal = {Journal of the Physical Society of Japan},
  volume = {9},
  year = {1954},
  pages = {753--766, 766--779},
}

@article{SchafferJorgensen1965,
  author = {Sch\"affer, C. E. and J\o{}rgensen, C. K.},
  title = {The angular overlap model: an attempt to revive the ligand field approaches},
  journal = {Molecular Physics},
  volume = {9},
  year = {1965},
  pages = {401--412},
}

@article{Schaffer1968,
  author = {Sch\"affer, C. E.},
  title = {A perturbation representation of weak covalent bonding},
  journal = {Structure and Bonding},
  volume = {5},
  year = {1968},
  pages = {68--95},
}

@book{Griffith1961,
  author = {Griffith, J. S.},
  title = {The Theory of Transition-Metal Ions},
  publisher = {Cambridge University Press},
  year = {1961},
}

@book{Ballhausen1962,
  author = {Ballhausen, C. J.},
  title = {Introduction to Ligand Field Theory},
  publisher = {McGraw-Hill},
  year = {1962},
}

@book{CottonWilkinson1999,
  author = {Cotton, F. A. and Wilkinson, G. and Murillo, C. A. and Bochmann, M.},
  title = {Advanced Inorganic Chemistry},
  edition = {6th},
  publisher = {Wiley-Interscience},
  year = {1999},
}

@book{HousecroftSharpe2018,
  author = {Housecroft, C. E. and Sharpe, A. G.},
  title = {Inorganic Chemistry},
  edition = {5th},
  publisher = {Pearson},
  year = {2018},
}

@book{Atkins2018,
  author = {Atkins, P. W. and Overton, T. and Rourke, J. and Weller, M. and Armstrong, F.},
  title = {Inorganic Chemistry},
  edition = {7th},
  publisher = {Oxford University Press},
  year = {2018},
}

@book{Bersuker2010,
  author = {Bersuker, I. B.},
  title = {Electronic Structure and Properties of Transition Metal Compounds},
  edition = {2nd},
  publisher = {Wiley},
  year = {2010},
}

@book{FiggisHitchman2000,
  author = {Figgis, B. N. and Hitchman, M. A.},
  title = {Ligand Field Theory and Its Applications},
  publisher = {Wiley-VCH},
  year = {2000},
}

@book{MiesslerFischerTarr2014,
  author = {Miessler, G. L. and Fischer, P. J. and Tarr, D. A.},
  title = {Inorganic Chemistry},
  edition = {5th},
  publisher = {Pearson},
  year = {2014},
}