Crystal Field, Ligand Field and MO Theory of Complexes
CSIR NET chemistry: d-orbital splitting, CFSE, high and low spin, spectrochemical series, Jahn-Teller effect, spinels, Irving-Williams order and MO theory.
- 19 min
On this page · 14 sections
Prerequisites
- Coordination Compounds: Geometry, Isomerism and Valence Bond Theory: coordination numbers, geometries and where VBT fails.
- Bonding in Heteronuclear Molecules: the MO picture of CO as a σ-donor and π-acceptor, used here for metal–ligand π bonding.
- The shapes of the five d orbitals: \( d_{xy}, d_{yz}, d_{xz} \) point between the axes; \( d_{x^2-y^2} \) and \( d_{z^2} \) point along them.
The Hook: Why Is Titanium(III) Purple?
Crystal Field Theory: The Model
Crystal field theory (CFT) treats ligands as point negative charges (or dipoles) and asks how their electrostatic field changes the energies of the metal d orbitals. It ignores covalency, which is its weakness, but gets the pattern of splitting right.
Octahedral splitting
Place six ligands on the ±x, ±y and ±z axes. The \( d_{x^2-y^2} \) and \( d_{z^2} \) orbitals point straight at them and are raised; \( d_{xy}, d_{xz}, d_{yz} \) point between them and are lowered relative to the average.
| Set | Orbitals | Energy relative to barycentre |
|---|---|---|
| \( e_g \) | \( d_{x^2-y^2}, d_{z^2} \) | \( +0.6\,\Delta_o \) (\( +6\,Dq \)) |
| \( t_{2g} \) | \( d_{xy}, d_{xz}, d_{yz} \) | \( -0.4\,\Delta_o \) (\( -4\,Dq \)) |
The barycentre rule: the weighted average energy is unchanged, \( 2(+0.6) + 3(-0.4) = 0 \). The splitting \( \Delta_o = 10\,Dq \).
Tetrahedral splitting
Four ligands at alternate corners of a cube point towards none of the d orbitals directly, but come closer to \( d_{xy}, d_{xz}, d_{yz} \). The order inverts, and there is no centre of symmetry, so the labels drop the g:
| Set | Orbitals | Energy |
|---|---|---|
| \( t_2 \) | \( d_{xy}, d_{xz}, d_{yz} \) | \( +0.4\,\Delta_t \) |
| \( e \) | \( d_{x^2-y^2}, d_{z^2} \) | \( -0.6\,\Delta_t \) |
for the same metal, ligands and distance (fewer ligands, and none pointing directly at an orbital). Because \( \Delta_t \) is small, it is almost always smaller than the pairing energy: tetrahedral complexes are high spin.
Square planar splitting
Remove the two z ligands from an octahedron (the limit of a tetragonal elongation). Everything with a z component falls:
$$d_{x^2-y^2} \gg d_{xy} > d_{z^2} > d_{xz} = d_{yz}$$The large gap to \( d_{x^2-y^2} \) is what matters: eight d electrons fill the four lower orbitals in pairs and leave \( d_{x^2-y^2} \) empty. That is why square planar \( d^8 \) complexes are diamagnetic. The exact order of the lower three orbitals depends on the complex; questions rely only on \( d_{x^2-y^2} \) highest and \( d_{xy} \) next.
graph LR
A[Free ion: 5 degenerate d] --> B[Octahedral: eg up, t2g down]
A --> C[Tetrahedral: t2 up, e down, Δt = 4/9 Δo]
B --> D[Tetragonal elongation: z orbitals fall]
D --> E[Square planar: dx2-y2 far above the rest]Crystal Field Stabilisation Energy
The CFSE is the energy of the d electrons in the field relative to the barycentre. For octahedral complexes:
$$\boxed{\text{CFSE} = (-0.4\,n_{t_{2g}} + 0.6\,n_{e_g})\,\Delta_o + (\text{extra pairs})\,P}$$where \( P \) is the mean pairing energy (the extra electron–electron repulsion plus loss of exchange energy when two electrons share an orbital).
High spin or low spin?
For \( d^4 \) to \( d^7 \) in octahedral fields there is a choice:
- \( \Delta_o < P \): high spin (weak field). Electrons spread out into \( e_g \) before pairing.
- \( \Delta_o > P \): low spin (strong field). Electrons pair in \( t_{2g} \) first.
| \( d^n \) | High spin | CFSE (HS) | Low spin | CFSE (LS) | Unpaired HS / LS |
|---|---|---|---|---|---|
| 1 | \( t_{2g}^1 \) | −0.4 Δo | same | 1 | |
| 2 | \( t_{2g}^2 \) | −0.8 Δo | same | 2 | |
| 3 | \( t_{2g}^3 \) | −1.2 Δo | same | 3 | |
| 4 | \( t_{2g}^3 e_g^1 \) | −0.6 Δo | \( t_{2g}^4 \) | −1.6 Δo + P | 4 / 2 |
| 5 | \( t_{2g}^3 e_g^2 \) | 0 | \( t_{2g}^5 \) | −2.0 Δo + 2P | 5 / 1 |
| 6 | \( t_{2g}^4 e_g^2 \) | −0.4 Δo | \( t_{2g}^6 \) | −2.4 Δo + 2P | 4 / 0 |
| 7 | \( t_{2g}^5 e_g^2 \) | −0.8 Δo | \( t_{2g}^6 e_g^1 \) | −1.8 Δo + P | 3 / 1 |
| 8 | \( t_{2g}^6 e_g^2 \) | −1.2 Δo | same | 2 | |
| 9 | \( t_{2g}^6 e_g^3 \) | −0.6 Δo | same | 1 | |
| 10 | \( t_{2g}^6 e_g^4 \) | 0 | same | 0 |
What decides the size of Δ
- Oxidation state: higher charge, shorter M–L distance, larger Δ. [Fe(H₂O)₆]³⁺ has a larger Δo than [Fe(H₂O)₆]²⁺.
- Row of the metal: Δ rises substantially from 3d to 4d to 5d, while pairing energy falls because the orbitals are larger. Result: 4d and 5d complexes are nearly always low spin.
- Geometry: \( \Delta_t \approx \tfrac{4}{9}\Delta_o \).
- The ligand: the spectrochemical series.
The spectrochemical series
$$\text{I}^- < \text{Br}^- < \text{S}^{2-} < \text{SCN}^-(\text{S}) < \text{Cl}^- < \text{NO}_3^- < \text{N}_3^- < \text{F}^- < \text{OH}^- < \text{C}_2\text{O}_4^{2-} < \text{H}_2\text{O}$$$$< \text{NCS}^-(\text{N}) < \text{CH}_3\text{CN} < \text{py} < \text{NH}_3 < \text{en} < \text{bipy} < \text{phen} < \text{NO}_2^- < \text{PPh}_3 < \text{CN}^- < \text{CO}$$Two things a point-charge model cannot explain: the anion F⁻ sits below neutral NH₃, and neutral CO is the strongest ligand of all. Both are explained by π bonding (MO section below): halides are π donors (lower Δ), CO and CN⁻ are π acceptors (raise Δ).
A matching series for metal ions, roughly: Mn²⁺ < Ni²⁺ < Co²⁺ < Fe²⁺ < V²⁺ < Fe³⁺ < Co³⁺ < Mo³⁺ < Rh³⁺ < Ru³⁺ < Pd⁴⁺ < Ir³⁺ < Pt⁴⁺. Note the trend: 3+ above 2+, and 4d/5d above 3d.
The Jahn–Teller Effect
In an octahedral complex the degeneracy comes from uneven occupation of a degenerate set:
| Degenerate set unevenly filled | Configurations | Effect |
|---|---|---|
| \( e_g \) (points at ligands) | high-spin \( d^4 \) (Cr²⁺, Mn³⁺), low-spin \( d^7 \) (low-spin Co²⁺, Ni³⁺), \( d^9 \) (Cu²⁺) | strong, distortion observed in structures |
| \( t_{2g} \) only | \( d^1, d^2 \), low-spin \( d^4, d^5 \), high-spin \( d^6, d^7 \) | weak, usually not seen in structures |
| none | \( d^3 \), high-spin \( d^5 \), low-spin \( d^6 \), \( d^8 \), \( d^{10} \) | no distortion |
Tetragonal elongation
Most strong Jahn–Teller ions elongate along z (two long, four short bonds). The levels split with the barycentre of each set preserved:
- \( e_g \): \( d_{z^2} \) falls by \( \delta_1/2 \), \( d_{x^2-y^2} \) rises by \( \delta_1/2 \).
- \( t_{2g} \): \( d_{xz}, d_{yz} \) fall by \( \delta_2/3 \), \( d_{xy} \) rises by \( 2\delta_2/3 \).
For Cu²⁺ (\( d^9 \)) the configuration becomes \( \ldots (d_{z^2})^2 (d_{x^2-y^2})^1 \). Two electrons are stabilised by \( \delta_1/2 \) and one destabilised by \( \delta_1/2 \): a net gain of \( \delta_1/2 \). That is why nearly every six-coordinate Cu(II) complex has four short and two long bonds, and why the limiting case, square planar Cu(II), is common.
Thermodynamic Evidence for CFSE
The double-humped hydration enthalpy curve
Plot the hydration enthalpy of the M²⁺ ions from Ca²⁺ to Zn²⁺. You would expect a smooth rise in magnitude as the ions shrink. Instead, only Ca²⁺ (\( d^0 \)), Mn²⁺ (high-spin \( d^5 \)) and Zn²⁺ (\( d^{10} \)) lie on the smooth line; all others are more exothermic than the line, with two humps peaking near V²⁺ (\( d^3 \)) and Ni²⁺ (\( d^8 \)). The extra stability is the CFSE of the high-spin hexaaqua ions, which is zero exactly for \( d^0, d^5, d^{10} \) and largest (−1.2 Δo) for \( d^3 \) and \( d^8 \). The same pattern appears in lattice energies of MCl₂ and in the ionic radii of high-spin M²⁺, which dip at V²⁺ and Ni²⁺ because electrons in \( t_{2g} \) shield the nucleus poorly along the bond axes.
The Irving–Williams series
For high-spin M²⁺ complexes with almost any ligand, stability constants follow
$$\text{Mn}^{2+} < \text{Fe}^{2+} < \text{Co}^{2+} < \text{Ni}^{2+} < \text{Cu}^{2+} > \text{Zn}^{2+}$$The rise to Ni²⁺ follows decreasing radius and increasing CFSE. Cu²⁺ comes out on top, although its CFSE (−0.6 Δo) is less than Ni²⁺’s, because the Jahn–Teller distortion gives four very strong equatorial bonds.
Spinels: normal or inverse?
A spinel AB₂O₄ has a cubic close-packed oxide lattice with A²⁺ in one-eighth of the tetrahedral holes and B³⁺ in half the octahedral holes (normal spinel). In an inverse spinel, B[AB]O₄, half the B³⁺ ions take the tetrahedral sites and A²⁺ goes octahedral.
The octahedral site preference energy (OSPE) is CFSE(octahedral) minus CFSE(tetrahedral), using \( \Delta_t = \tfrac{4}{9}\Delta_o \):
| \( d^n \) (high spin) | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| OSPE (units of Δo) | 0 | 0.13 | 0.27 | 0.84 | 0.42 | 0 | 0.13 | 0.27 | 0.84 | 0.42 | 0 |
Cr³⁺ (\( d^3 \)) and Ni²⁺ (\( d^8 \)) have the strongest preference for octahedral sites.
| Spinel | Ions | Type | Reason |
|---|---|---|---|
| Mn₃O₄ | Mn²⁺ \( d^5 \), Mn³⁺ \( d^4 \) | normal | Mn²⁺ has no preference; Mn³⁺ prefers octahedral |
| Fe₃O₄ (magnetite) | Fe²⁺ \( d^6 \), Fe³⁺ \( d^5 \) | inverse | Fe³⁺ has no preference; Fe²⁺ gains a little in octahedral sites |
| Co₃O₄ | Co²⁺ \( d^7 \), Co³⁺ low-spin \( d^6 \) | normal | low-spin Co³⁺ has a very large octahedral CFSE |
| NiFe₂O₄ | Ni²⁺ \( d^8 \), Fe³⁺ \( d^5 \) | inverse | Ni²⁺ strongly prefers octahedral |
| NiCr₂O₄ | Ni²⁺, Cr³⁺ \( d^3 \) | normal | Cr³⁺ takes the octahedral sites |
Ligand Field and MO Theory
Ligand field theory is crystal field theory with covalency allowed in; the full version is the molecular orbital treatment.
σ bonding in an octahedral complex
The metal’s valence orbitals (3d, 4s, 4p) are classified by symmetry and matched with ligand group orbitals (SALCs) built from the six σ lone pairs:
| Symmetry | Metal orbitals | Ligand SALCs? | Result |
|---|---|---|---|
| \( a_{1g} \) | 4s | yes | bonding + antibonding |
| \( t_{1u} \) | \( 4p_x, 4p_y, 4p_z \) | yes | bonding + antibonding |
| \( e_g \) | \( d_{x^2-y^2}, d_{z^2} \) | yes | bonding + antibonding \( e_g^* \) |
| \( t_{2g} \) | \( d_{xy}, d_{xz}, d_{yz} \) | no (σ only) | non-bonding |
The six ligand pairs (12 electrons) fill the bonding \( a_{1g}, t_{1u}, e_g \) MOs, which are mostly ligand in character. The metal’s d electrons go into the \( t_{2g} \) (non-bonding) and \( e_g^* \) (antibonding) levels. So the crystal field \( \Delta_o \) becomes the gap between \( t_{2g} \) and \( e_g^* \). Stronger σ donation pushes \( e_g^* \) higher and increases Δo.
π bonding: the reason for the spectrochemical series
The \( t_{2g} \) orbitals can overlap with ligand orbitals of π symmetry.
graph TD
A[Ligand pi orbitals] --> B{Filled or empty, and where?}
B -->|Filled, below metal t2g: halides, OH-, O2-| C[t2g pushed up, antibonding]
C --> D[Δo DECREASES: π donor, weak field]
B -->|Empty, above metal t2g: CO, CN-, PR3, bipy| E[t2g pulled down, bonding]
E --> F[Δo INCREASES: π acceptor, strong field]So the spectrochemical series is, roughly: π donors < weak π donors / pure σ donors < π acceptors. I⁻, Br⁻, Cl⁻ and F⁻ are weak-field partly because they are π donors; H₂O is a weaker π donor than OH⁻ and sits higher; NH₃ is σ only; bipy, phen, CN⁻ and CO accept π density into empty π* orbitals (back-bonding) and sit at the top.
The angular overlap model in one line each
The angular overlap model (AOM) assigns each σ interaction an energy \( e_\sigma \) and each π interaction \( e_\pi \), scaled by the angular overlap. For σ-donor, π-donor ligands:
- Octahedral: \( e_g \) raised by \( 3e_\sigma \), \( t_{2g} \) raised by \( 4e_\pi \), so \( \Delta_o = 3e_\sigma - 4e_\pi \).
- Tetrahedral: \( t_2 \) raised by \( \tfrac{4}{3}e_\sigma + \tfrac{8}{9}e_\pi \), \( e \) by \( \tfrac{8}{3}e_\pi \), so \( \Delta_t = \tfrac{4}{9}(3e_\sigma - 4e_\pi) \).
The \( \tfrac{4}{9} \) ratio drops out naturally. For a π acceptor, \( e_\pi \) is negative and Δo grows.
Memory Tricks
Don’t Confuse
| This | vs | That |
|---|---|---|
| \( \Delta_o \) between \( t_{2g} \) and \( e_g^* \) (MO) | \( \Delta_o \) between point-charge-split levels (CFT); same number, different picture | |
| CFSE (electronic stabilisation in the field) | total bond energy (CFSE is only a small part of it) | |
| Strong Jahn–Teller (\( e_g \) degeneracy) | weak Jahn–Teller (\( t_{2g} \) degeneracy) | |
| Normal spinel A[B₂]O₄ (Mn₃O₄, Co₃O₄) | inverse spinel B[AB]O₄ (Fe₃O₄, NiFe₂O₄) | |
| Irving–Williams: Cu²⁺ highest | CFSE: Ni²⁺ (−1.2 Δo) larger than Cu²⁺ (−0.6 Δo) | |
| π donor (F⁻, OH⁻): lowers Δ | π acceptor (CO, CN⁻): raises Δ |
Common Mistakes
Practice
All questions below are original practice questions written for this page. They are not from any CSIR NET paper.
Level 1: Foundation
- What is the CFSE of an octahedral high-spin \( d^7 \) ion, in units of Δo (ignore pairing)?
- Which d orbitals form the \( e_g \) set in an octahedral field?
- Name the d-electron counts of octahedral ions that show strong Jahn–Teller distortion.
Answers
- \( t_{2g}^5 e_g^2 \): \( 5(-0.4) + 2(0.6) = -0.8\,\Delta_o \).
- \( d_{x^2-y^2} \) and \( d_{z^2} \).
- High-spin \( d^4 \), low-spin \( d^7 \) and \( d^9 \).
Level 2: Part B Style
Practice Q1. The number of unpaired electrons in [Mn(CN)₆]³⁻ (low spin) and [MnF₆]³⁻ (high spin) respectively is:
(a) 2 and 4 (b) 0 and 4 (c) 2 and 5 (d) 1 and 4
Answer: (a)
Mn³⁺ is \( d^4 \). Low spin \( t_{2g}^4 \): two unpaired. High spin \( t_{2g}^3 e_g^1 \): four unpaired (and Jahn–Teller distorted).
Practice Q2. Which ion is expected to have the largest Δo?
(a) [Co(H₂O)₆]²⁺ (b) [Co(H₂O)₆]³⁺ (c) [Rh(H₂O)₆]³⁺ (d) [CoCl₄]²⁻
Answer: (c)
Higher oxidation state and a 4d metal both increase Δ. (d) is tetrahedral with a weak-field ligand and has the smallest splitting.
Practice Q3. Which spinel is inverse?
(a) MgAl₂O₄ (b) Mn₃O₄ (c) NiFe₂O₄ (d) ZnCr₂O₄
Answer: (c)
Ni²⁺ (\( d^8 \)) has a large octahedral site preference; Fe³⁺ (\( d^5 \)) has none. Fe³⁺ goes tetrahedral, giving Fe³⁺[Ni²⁺Fe³⁺]O₄. In (a) neither ion has d electrons and MgAl₂O₄ is the parent normal spinel; in (d) Cr³⁺ firmly occupies octahedral sites and Zn²⁺ (\( d^{10} \)) takes the tetrahedral ones.
Practice Q4. For an octahedral complex, which ligand acts as a π acceptor and raises Δo relative to a pure σ donor?
(a) Cl⁻ (b) OH⁻ (c) NH₃ (d) PPh₃
Answer: (d)
Phosphines accept metal \( t_{2g} \) density into empty orbitals of π symmetry (σ* P–C orbitals), lowering \( t_{2g} \) and widening Δo. Cl⁻ and OH⁻ are π donors; NH₃ is σ only.
Level 3: Part C Style
Practice Q5. For a hypothetical octahedral \( d^6 \) complex, Δo = 19,000 cm⁻¹ and the mean pairing energy P = 21,000 cm⁻¹. Find (i) the spin state, (ii) the CFSE in cm⁻¹ (using the extra-pairs convention), and (iii) the minimum Δo at which the low-spin form becomes the ground state.
Answer
(i) Δo < P, so high spin, \( t_{2g}^4 e_g^2 \).
(ii) CFSE = −0.4 × 19,000 = −7,600 cm⁻¹.
(iii) Low spin: −2.4 Δo + 2P. High spin: −0.4 Δo. Low spin is lower when \( -2.4\Delta_o + 2P < -0.4\Delta_o \), that is \( 2\Delta_o > 2P \), so \( \Delta_o > P = \) 21,000 cm⁻¹. At Δo = 19,000 cm⁻¹ the high-spin state is lower by \( 2(21{,}000 - 19{,}000) = 4{,}000 \) cm⁻¹.
Practice Q6. Consider the statements about high-spin hexaaqua ions of the first transition series:
- The hydration enthalpies of Ca²⁺, Mn²⁺ and Zn²⁺ lie on a smooth curve because their CFSE is zero.
- The octahedral site preference energy of Cr³⁺ is about 0.84 Δo.
- [Cu(H₂O)₆]²⁺ is expected to be a regular octahedron.
- The stability of M²⁺ complexes with en is greater for Cu²⁺ than for Ni²⁺.
Which are correct?
(a) 1, 2 and 4 only (b) 1 and 3 only (c) 2 and 4 only (d) 1, 2, 3 and 4
Answer: (a)
- Correct. \( d^0 \), high-spin \( d^5 \), \( d^{10} \) have zero CFSE.
- Correct. −1.2 Δo (oct) minus \( -0.8 \times \tfrac{4}{9} \)Δo = −0.356 Δo (tet) gives 0.844 Δo.
- Wrong. \( d^9 \) with \( e_g^3 \): strong Jahn–Teller elongation.
- Correct. Irving–Williams order peaks at Cu²⁺.
Practice Q7. Using the angular overlap model for a hypothetical metal ion with six identical ligands, \( e_\sigma = 7{,}000 \) cm⁻¹ and \( e_\pi = 900 \) cm⁻¹ (π donor). Calculate Δo, and Δt for the tetrahedral complex with the same ligands at the same distance. What happens to Δo if the ligand is replaced by a π acceptor with the same \( e_\sigma \) and \( e_\pi = -900 \) cm⁻¹?
Answer
\( \Delta_o = 3e_\sigma - 4e_\pi = 21{,}000 - 3{,}600 = \) 17,400 cm⁻¹.
\( \Delta_t = \tfrac{4}{9}\Delta_o = \tfrac{4}{9} \times 17{,}400 = \) 7,733 cm⁻¹ (about 7,730).
With a π acceptor: \( \Delta_o = 21{,}000 + 3{,}600 = \) 24,600 cm⁻¹: a rise of 7,200 cm⁻¹ from π effects alone, which is why CO and CN⁻ sit far above NH₃ in the series.
Quick Revision Table
| Idea | Fact |
|---|---|
| Octahedral | \( e_g \) +0.6 Δo, \( t_{2g} \) −0.4 Δo; Δo = 10 Dq |
| Tetrahedral | \( t_2 \) +0.4 Δt, \( e \) −0.6 Δt; Δt ≈ 4/9 Δo; high spin |
| Square planar | \( d_{x^2-y^2} \gg d_{xy} \) > rest; \( d^8 \) diamagnetic |
| Spin state | low spin if Δo > P; 4d/5d nearly always low spin |
| CFSE max | \( d^3 \) and \( d^8 \): −1.2 Δo (HS); LS \( d^6 \) −2.4 Δo |
| Spectrochemical | I⁻ < Br⁻ < Cl⁻ < F⁻ < OH⁻ < H₂O < NH₃ < en < bipy < CN⁻ < CO |
| π effects | donors lower Δ; acceptors raise Δ |
| Strong JT | HS \( d^4 \), LS \( d^7 \), \( d^9 \); usually z-elongation |
| Hydration curve | humps at \( d^3 \) and \( d^8 \); \( d^0, d^5, d^{10} \) on the line |
| Irving–Williams | Mn < Fe < Co < Ni < Cu > Zn |
| OSPE | 0.84 Δo for \( d^3 \), \( d^8 \); 0 for \( d^0, d^5, d^{10} \) |
| Spinels | Fe₃O₄, NiFe₂O₄ inverse; Mn₃O₄, Co₃O₄, NiCr₂O₄ normal |
| AOM | Δo = 3eσ − 4eπ; Δt = (4/9)Δo |
Key Takeaways
What’s Next
- Electronic Spectra of Complexes: measuring Δo and B from d–d bands, term symbols and Tanabe–Sugano diagrams.
- Magnetic Properties of Complexes: from unpaired-electron counts to orbital contributions.
- Reaction Mechanisms of Complexes: crystal-field activation energy and why Cr(III) and Co(III) are inert.
- Back to the chapter overview.