12.3 Bonding: Valence Bond and Crystal Field Theory
Why is $\mathrm{[Ni(CN)_{4}]^{2-}}$ square planar and diamagnetic while $\mathrm{[NiCl_{4}]^{2-}}$ is tetrahedral and paramagnetic? Why is $\mathrm{[Ti(H_{2}O)_{6}]^{3+}}$ purple and $\mathrm{[Zn(H_{2}O)_{6}]^{2+}}$ colourless? Valence bond theory explains shapes and magnetism through hybridisation; crystal field theory splits the $d$ orbitals and explains spin, stability and colour.
Dissolve nickel(II) chloride in water and you get a green solution. Add excess ammonia and it turns blue-violet; add cyanide instead and it becomes yellow, and a magnet that slightly attracted the original salt now has no effect at all. Cobalt(III) with ammonia gives orange-yellow crystals that are repelled by a magnet; cobalt(III) with fluoride gives a blue salt that is strongly attracted. The metal ion is the same in each pair — only the ligands differ. Werner's theory (Section 12.1) tells us how many ligands surround the metal and how they are named and arranged (Section 12.2), but it says nothing about why the bonds form, why a four-coordinate complex is sometimes tetrahedral and sometimes square planar, or where the colours and the magnetism come from.
Two models answer these questions. Valence bond theory (VBT) extends the hybridisation idea of Section 3.3: each ligand donates a lone pair into an empty hybrid orbital of the metal, and the type of hybrid fixes the geometry and the number of unpaired electrons. Crystal field theory (CFT) takes a different view: the ligands are negative charges that raise the energy of some $d$ orbitals more than others. The resulting gap, $\Delta_{o}$, decides whether electrons pair up, how much extra stability the complex gains, and which colour of light it absorbs. JEE Main tests both models every year — almost always as a count: unpaired electrons, spin-only magnetic moment, CFSE in units of $\Delta_{o}$, or $d$ electrons on the metal.
Valence bond theory of complexes
A ligand is a Lewis base with at least one lone pair; a metal ion is a Lewis acid with empty orbitals. In VBT the metal–ligand bond is a coordinate (dative) σ bond: an orbital of the ligand holding a lone pair overlaps an empty orbital of the metal, and both electrons come from the ligand. To make a set of equivalent bonds pointing in definite directions, the metal first mixes the required number of its empty orbitals into hybrids, exactly as carbon does in methane.
According to valence bond theory applied to complexes (Pauling, 1931):
- The central metal atom or ion makes available a number of empty orbitals equal to its coordination number.
- These empty orbitals — chosen from $(n-1)d$, $ns$, $np$ and $nd$ — hybridise to give a set of equivalent hybrid orbitals of definite directions.
- Each ligand donates a lone pair into one hybrid orbital, forming a coordinate σ bond ($\mathrm{M\leftarrow L}$).
- The type of hybridisation fixes the geometry of the complex.
- Electrons already on the metal stay in its non-hybridised $d$ orbitals. Some ligands (strong-field ligands) force these electrons to pair up so that inner $d$ orbitals become free for hybridisation; others do not. The number of unpaired electrons left decides the magnetic behaviour.
| Coordination number | Orbitals used | Hybridisation | Geometry | Examples |
|---|---|---|---|---|
| 2 | $ns,\ np$ | $sp$ | linear | $\mathrm{[Ag(NH_{3})_{2}]^{+}}$, $\mathrm{[Ag(CN)_{2}]^{-}}$ |
| 4 | $ns,\ np_{x},np_{y},np_{z}$ | $sp^{3}$ | tetrahedral | $\mathrm{[NiCl_{4}]^{2-}}$, $\mathrm{[Zn(NH_{3})_{4}]^{2+}}$, $\mathrm{[Ni(CO)_{4}]}$ |
| 4 | $(n-1)d_{x^{2}-y^{2}},\ ns,\ np_{x},np_{y}$ | $dsp^{2}$ | square planar | $\mathrm{[Ni(CN)_{4}]^{2-}}$, $\mathrm{[PtCl_{4}]^{2-}}$ |
| 5 | $(n-1)d_{z^{2}},\ ns,\ np$ | $dsp^{3}$ | trigonal bipyramidal | $\mathrm{[Fe(CO)_{5}]}$ |
| 6 | $(n-1)d_{x^{2}-y^{2}},(n-1)d_{z^{2}},\ ns,\ np$ | $d^{2}sp^{3}$ | octahedral (inner orbital) | $\mathrm{[Co(NH_{3})_{6}]^{3+}}$, $\mathrm{[Fe(CN)_{6}]^{4-}}$ |
| 6 | $ns,\ np,\ nd_{x^{2}-y^{2}},nd_{z^{2}}$ | $sp^{3}d^{2}$ | octahedral (outer orbital) | $\mathrm{[CoF_{6}]^{3-}}$, $\mathrm{[Fe(H_{2}O)_{6}]^{2+}}$ |
The $d$ orbitals chosen are always those whose lobes point along the bond directions: $d_{x^{2}-y^{2}}$ for the four bonds in the $xy$ plane of a square planar complex, and both $d_{x^{2}-y^{2}}$ and $d_{z^{2}}$ for the six bonds along $\pm x$, $\pm y$, $\pm z$ of an octahedron. Writing $d$ first ($d^{2}sp^{3}$, $dsp^{2}$) signals inner $(n-1)d$ orbitals; writing it last ($sp^{3}d^{2}$) signals outer $nd$ orbitals.
An octahedral complex that uses the inner $(n-1)d$ orbitals ($d^{2}sp^{3}$) is an inner-orbital complex. Its $d$ electrons are squeezed into the remaining three $(n-1)d$ orbitals, so it has the fewest possible unpaired electrons: a low-spin (spin-paired) complex.
An octahedral complex that uses the outer $nd$ orbitals ($sp^{3}d^{2}$) is an outer-orbital complex. All five $(n-1)d$ orbitals stay available to the metal's own electrons, which spread out by Hund's rule: a high-spin (spin-free) complex. Outer $nd$ orbitals are larger and higher in energy, so outer-orbital bonds are longer and weaker, and these complexes are usually more labile (exchange ligands faster).
Which way does a given complex go? Strong-field ligands such as $\mathrm{CN^{-}}$, $\mathrm{CO}$, $\mathrm{NO_{2}^{-}}$ and ethane-1,2-diamine (en) — and $\mathrm{NH_{3}}$ or even $\mathrm{H_{2}O}$ with the small, highly charged $\mathrm{Co^{3+}}$ — push the metal's electrons into pairs to vacate two inner $d$ orbitals. Weak-field ligands such as $\mathrm{F^{-}}$, $\mathrm{Cl^{-}}$, $\mathrm{Br^{-}}$, $\mathrm{I^{-}}$ and usually $\mathrm{H_{2}O}$ cannot. In VBT this is an experimental input: we learn which happened from the measured magnetic moment, and the theory then tells us the hybridisation. Some ions have no choice: $d^{1}$, $d^{2}$ and $d^{3}$ ions always have two empty inner $d$ orbitals ($d^{2}sp^{3}$), while $d^{8}$, $d^{9}$ and $d^{10}$ ions in octahedral complexes cannot free two of them and are always $sp^{3}d^{2}$.
Magnetism is the link to experiment. A substance with unpaired electrons is paramagnetic — drawn into a magnetic field — because each unpaired electron behaves as a tiny magnet. With all electrons paired it is diamagnetic and weakly pushed out. The size of the effect is measured with a Gouy or Faraday balance (the sample appears heavier when the magnet is switched on) and expressed as the magnetic moment $\mu$ in Bohr magnetons (BM), $1\ \mathrm{BM}=\dfrac{eh}{4\pi m_{e}}=9.27\times10^{-24}\ \mathrm{J\,T^{-1}}$.
For a complex with $n$ unpaired electrons, if the contribution of orbital motion is neglected (it is largely "quenched" by the ligands in complexes of the first transition series),
$$\mu_{s}=\sqrt{n(n+2)}\ \ \mathrm{BM}.$$| $n$ | 0 | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|---|
| $\mu_{s}$ (BM) | 0 | $\sqrt3=1.73$ | $\sqrt8=2.83$ | $\sqrt{15}=3.87$ | $\sqrt{24}=4.90$ | $\sqrt{35}=5.92$ |
Read backwards: a measured moment fixes $n$ — take the nearest value in the table (measured values are often a little above the spin-only value because some orbital contribution survives).
- Oxidation state of the metal from the charge balance (ligand charges: $\mathrm{CN^{-}},\mathrm{Cl^{-}},\mathrm{F^{-}}$ −1; $\mathrm{C_{2}O_{4}^{2-}}$ −2; $\mathrm{H_{2}O},\mathrm{NH_{3}},\mathrm{CO}$, en 0).
- $d$ count: electrons lost come from $ns$ first, so $\mathrm{M^{x+}}$ of a 3d metal has $(\text{group number}-x)$ $d$ electrons ($\mathrm{Fe^{3+}}$: $8-3=5$).
- Pairing? Strong-field ligand (or a measured low moment) → pair the $d$ electrons as far as needed; weak-field ligand → Hund's rule.
- Count the empty orbitals needed (= coordination number), lowest energy first, and name the hybridisation and geometry from Rule 12.8.
- Unpaired electrons left in the non-hybridised $d$ orbitals → $\mu_{s}=\sqrt{n(n+2)}$.
What VBT cannot do. Valence bond theory gives a simple, pictorial account of geometry and magnetism, but it has serious gaps:
- It offers no explanation of colour — the absorption spectra of complexes are outside the model.
- It cannot predict whether a ligand will cause pairing; that is taken from the measured moment after the fact. It also cannot predict whether a four-coordinate complex will be tetrahedral or square planar.
- It does not distinguish weak and strong ligands quantitatively, and gives no energies — so it cannot explain the thermodynamic or kinetic stabilities of complexes, or why measured moments vary with temperature and exceed the spin-only values.
- Some descriptions are artificial. Square planar $\mathrm{[Cu(NH_{3})_{4}]^{2+}}$ ($3d^{9}$) would need one $3d$ electron promoted to a $4p$ orbital to vacate a $3d$ orbital for $dsp^{2}$; an electron that high in energy should be easily lost, yet the complex is not readily oxidised to copper(III).
The widget below builds the VBT orbital-box picture for a dozen complexes, step by step.
In the VBT builder choose $\mathrm{[Fe(CN)_{6}]^{3-}}$ (cyanide is a strong-field ligand). What will the readouts show for the hybridisation and the unpaired electrons?
$\mathrm{[Ni(NH_{3})_{6}]^{2+}}$ is octahedral. According to VBT it is
Crystal field theory: splitting of the d orbitals
Crystal field theory (Bethe and Van Vleck, 1929–35) treats the metal–ligand bond as purely electrostatic. Anionic ligands are point negative charges; neutral ligands such as $\mathrm{H_{2}O}$ and $\mathrm{NH_{3}}$ are point dipoles with their negative end (the lone pair) towards the metal. In a free gaseous ion all five $d$ orbitals have the same energy (they are degenerate). If the ion were surrounded by a spherical shell of negative charge, all five would be raised by the same amount. But real ligands sit in definite directions, and an electron in an orbital pointing at a ligand is repelled more than one in an orbital pointing between ligands. The degeneracy is lifted.
Crystal field theory regards a complex as a central metal ion surrounded by ligands that act as point charges (or point dipoles), and considers only the electrostatic repulsion between the ligands and the metal's $d$ electrons. Because the ligands lie in definite directions, the five $d$ orbitals no longer have equal energies: they split into groups.
The energy gap between the groups is the crystal field splitting energy: $\Delta_{o}$ (or $10\,Dq$) for an octahedral field and $\Delta_{t}$ for a tetrahedral field. It is usually quoted in $\mathrm{kJ\,mol^{-1}}$ or in wavenumbers ($\mathrm{cm^{-1}}$; $1000\ \mathrm{cm^{-1}}=11.96\ \mathrm{kJ\,mol^{-1}}$).
Octahedral field. Place the six ligands on the $\pm x$, $\pm y$ and $\pm z$ axes. The $d_{x^{2}-y^{2}}$ and $d_{z^{2}}$ orbitals have their lobes along the axes, so their electrons meet the ligands head-on and are raised in energy; this pair is called $e_{g}$. The $d_{xy}$, $d_{yz}$ and $d_{xz}$ orbitals point between the axes and are raised less; this trio is called $t_{2g}$. Relative to the energy the orbitals would have in a spherical field (the barycentre), the total energy is unchanged by the splitting: the two $e_{g}$ orbitals go up by $0.6\Delta_{o}$ and the three $t_{2g}$ orbitals go down by $0.4\Delta_{o}$, since $2(+0.6)+3(-0.4)=0$.
Tetrahedral field. The four ligands sit at alternate corners of a cube with the metal at its centre. Now no $d$ orbital points straight at a ligand, but $d_{xy}$, $d_{yz}$ and $d_{xz}$ (pointing to the midpoints of the cube's edges) come closer to the ligands than $d_{x^{2}-y^{2}}$ and $d_{z^{2}}$ (pointing to the face centres). The pattern is inverted: the three $t_{2}$ orbitals lie $0.4\Delta_{t}$ above the barycentre and the two $e$ orbitals $0.6\Delta_{t}$ below. (No subscript $g$: a tetrahedron has no centre of symmetry.)
- Octahedral: $e_{g}\ (d_{x^{2}-y^{2}},d_{z^{2}})$ at $+0.6\Delta_{o}$; $t_{2g}\ (d_{xy},d_{yz},d_{xz})$ at $-0.4\Delta_{o}$.
- Tetrahedral: $t_{2}\ (d_{xy},d_{yz},d_{xz})$ at $+0.4\Delta_{t}$; $e\ (d_{x^{2}-y^{2}},d_{z^{2}})$ at $-0.6\Delta_{t}$.
- For the same metal ion, the same ligands and the same metal–ligand distance, $$\Delta_{t}\approx\tfrac{4}{9}\,\Delta_{o}.$$ The splitting is smaller because there are only four ligands instead of six (a factor $\tfrac23$) and because no orbital points directly at a ligand (a further factor $\tfrac23$).
- The barycentre is preserved: the energy gained by electrons in the lower set is balanced by the energy lost by electrons in the upper set when all five orbitals are equally occupied ($d^{5}$ high spin and $d^{10}$ have no net gain).
The spectrochemical series. The size of $\Delta_{o}$ is measured from the absorption spectra of complexes (fourth subsection). Comparing complexes of one metal ion with different ligands gives an order of ligand field strength that is almost the same for every metal:
1. Ligand. In order of increasing field strength (increasing $\Delta_{o}$):
$$\mathrm{I^{-}\lt Br^{-}\lt SCN^{-}\lt Cl^{-}\lt S^{2-}}$$ $$\mathrm{\lt F^{-}\lt OH^{-}\lt C_{2}O_{4}^{2-}\lt H_{2}O}$$ $$\mathrm{\lt NCS^{-}\lt edta^{4-}\lt NH_{3}}$$ $$\mathrm{\lt en\lt CN^{-}\lt CO}$$(SCN⁻ bonded through S is weaker than NCS⁻ bonded through N.) Ligands to the left of $\mathrm{H_{2}O}$ are weak-field; $\mathrm{NH_{3}}$, en, $\mathrm{CN^{-}}$ and $\mathrm{CO}$ are strong-field.
2. Oxidation state of the metal. A higher charge pulls the ligands closer: $\Delta_{o}$ for $\mathrm{M^{3+}}$ is roughly one and a half to two times that for $\mathrm{M^{2+}}$ with the same ligands.
3. Row of the metal. $\Delta$ increases down a group, $3d\lt4d\lt5d$, by roughly 30–50% per step, because the larger $4d$ and $5d$ orbitals reach further towards the ligands. Complexes of $4d$ and $5d$ metals are therefore almost always low spin.
4. Geometry. $\Delta_{t}\approx\tfrac49\Delta_{o}$ (Rule 12.10).
The series is an experimental order, and here crystal field theory shows its own weakness: if the ligands were simply point charges, the anions $\mathrm{F^{-}}$ and $\mathrm{OH^{-}}$ should split the $d$ orbitals more than the neutral $\mathrm{NH_{3}}$, and neutral $\mathrm{CO}$ should be among the weakest — the opposite of what is observed. The real bonds have covalent character (and $\mathrm{CN^{-}}$ and $\mathrm{CO}$ accept metal $d$ electrons into empty π orbitals), which the purely electrostatic model ignores. CFT is nevertheless very successful in explaining magnetism, colour and stability once $\Delta$ is known.
In an octahedral complex, which $d$ orbitals are raised to the higher energy level?
High spin, low spin, CFSE and magnetism
Feed electrons one at a time into the split $d$ orbitals of an octahedral ion. The first three go singly into $t_{2g}$ (Hund's rule). The fourth faces a choice: it can enter $e_{g}$, climbing the gap $\Delta_{o}$ but staying unpaired, or it can pair with an electron in $t_{2g}$, avoiding the climb but paying the extra repulsion of two electrons in one orbital (and losing exchange energy). Whichever costs less wins.
The pairing energy $P$ is the energy needed to put a second electron, of opposite spin, into an orbital that already holds one. It is a property of the metal ion (typically $200$–$300\ \mathrm{kJ\,mol^{-1}}$ for $3d$ ions).
A high-spin (weak-field) complex has the largest possible number of unpaired electrons: electrons occupy the upper set before pairing. A low-spin (strong-field) complex has the smallest possible number: the lower set is filled completely before the upper set is used.
- Octahedral, $d^{4}$ to $d^{7}$: if $\Delta_{o}\lt P$ the complex is high spin (configuration $t_{2g}^{3}e_{g}^{1}$, $t_{2g}^{3}e_{g}^{2}$, $t_{2g}^{4}e_{g}^{2}$, $t_{2g}^{5}e_{g}^{2}$); if $\Delta_{o}\gt P$ it is low spin ($t_{2g}^{4}$, $t_{2g}^{5}$, $t_{2g}^{6}$, $t_{2g}^{6}e_{g}^{1}$).
- Octahedral $d^{1}$–$d^{3}$ and $d^{8}$–$d^{10}$ have only one possible arrangement, so "high spin or low spin" does not arise.
- Tetrahedral complexes are practically always high spin: $\Delta_{t}\approx\tfrac49\Delta_{o}$ is almost always smaller than $P$.
- In practice: weak-field ligands ($\mathrm{I^{-}}\ldots\mathrm{H_{2}O}$) give high-spin $3d$ complexes; $\mathrm{CN^{-}}$, $\mathrm{CO}$, en give low spin; $\mathrm{NH_{3}}$ and $\mathrm{H_{2}O}$ give low spin with $\mathrm{Co^{3+}}$ (large $\Delta_{o}$ of a $3+$ ion, unusually small $P$ for $d^{6}$).
This is the crystal-field picture behind VBT's inner- and outer-orbital complexes: low spin corresponds to inner-orbital $d^{2}sp^{3}$, high spin to outer-orbital $sp^{3}d^{2}$ — but CFT can now predict the outcome by comparing two energies.
The crystal field stabilisation energy is the net change in energy of the $d$ electrons caused by the splitting, measured from the barycentre. For an octahedral complex with configuration $t_{2g}^{x}e_{g}^{y}$,
$$\mathrm{CFSE}=(-0.4x+0.6y)\,\Delta_{o}\ \ [\,+\,mP\,],$$and for a tetrahedral complex $e^{x}t_{2}^{y}$,
$$\mathrm{CFSE}=(-0.6x+0.4y)\,\Delta_{t}.$$A negative CFSE means stabilisation. The optional term $mP$ counts the extra electron pairs ($m$) that the field has created compared with the free ion; it matters only when high-spin and low-spin arrangements are compared. In JEE questions "CFSE in units of $\Delta_{o}$" means the bracket alone, with pairing energy ignored unless the question asks for it.
| $d^{n}$ | Octahedral high spin | unpaired | CFSE ($\Delta_{o}$) | Octahedral low spin | unpaired | CFSE ($\Delta_{o}$) | Tetrahedral | CFSE ($\Delta_{t}$) |
|---|---|---|---|---|---|---|---|---|
| $d^{1}$ | $t_{2g}^{1}$ | 1 | $-0.4$ | same | $e^{1}$ | $-0.6$ | ||
| $d^{2}$ | $t_{2g}^{2}$ | 2 | $-0.8$ | same | $e^{2}$ | $-1.2$ | ||
| $d^{3}$ | $t_{2g}^{3}$ | 3 | $-1.2$ | same | $e^{2}t_{2}^{1}$ | $-0.8$ | ||
| $d^{4}$ | $t_{2g}^{3}e_{g}^{1}$ | 4 | $-0.6$ | $t_{2g}^{4}$ | 2 | $-1.6\ (+P)$ | $e^{2}t_{2}^{2}$ | $-0.4$ |
| $d^{5}$ | $t_{2g}^{3}e_{g}^{2}$ | 5 | $0$ | $t_{2g}^{5}$ | 1 | $-2.0\ (+2P)$ | $e^{2}t_{2}^{3}$ | $0$ |
| $d^{6}$ | $t_{2g}^{4}e_{g}^{2}$ | 4 | $-0.4$ | $t_{2g}^{6}$ | 0 | $-2.4\ (+2P)$ | $e^{3}t_{2}^{3}$ | $-0.6$ |
| $d^{7}$ | $t_{2g}^{5}e_{g}^{2}$ | 3 | $-0.8$ | $t_{2g}^{6}e_{g}^{1}$ | 1 | $-1.8\ (+P)$ | $e^{4}t_{2}^{3}$ | $-1.2$ |
| $d^{8}$ | $t_{2g}^{6}e_{g}^{2}$ | 2 | $-1.2$ | same | $e^{4}t_{2}^{4}$ | $-0.8$ | ||
| $d^{9}$ | $t_{2g}^{6}e_{g}^{3}$ | 1 | $-0.6$ | same | $e^{4}t_{2}^{5}$ | $-0.4$ | ||
| $d^{10}$ | $t_{2g}^{6}e_{g}^{4}$ | 0 | $0$ | same | $e^{4}t_{2}^{6}$ | $0$ |
Why the low-spin advantage is exactly $\Delta_{o}$ per pair. Going from high spin to low spin moves one electron from $e_{g}$ to $t_{2g}$ for each extra pair formed: the electron drops by $0.6\Delta_{o}+0.4\Delta_{o}=\Delta_{o}$ and the pairing costs $P$. So low spin wins exactly when $\Delta_{o}\gt P$ — the rule above. For $d^{6}$, for example, the low-spin arrangement is lower by $2(\Delta_{o}-P)$.
- Find the oxidation state and the $d$ count, as for VBT.
- Decide the geometry (octahedral, tetrahedral) and, for octahedral $d^{4}$–$d^{7}$, the spin state: given numbers → compare $\Delta_{o}$ with $P$; otherwise use the spectrochemical series (and the $\mathrm{Co^{3+}}$, $4d/5d$ exceptions).
- Write the configuration ($t_{2g}^{x}e_{g}^{y}$ or $e^{x}t_{2}^{y}$) and count unpaired electrons → $\mu_{s}=\sqrt{n(n+2)}\ \mathrm{BM}$.
- CFSE $=(-0.4x+0.6y)\Delta_{o}$ or $(-0.6x+0.4y)\Delta_{t}$; add $mP$ only if asked.
1. Counting d electrons wrongly. A transition metal loses its $4s$ electrons first: $\mathrm{Fe^{3+}}$ is $3d^{5}$ (not $3d^{3}4s^{2}$), $\mathrm{Cu^{2+}}$ is $3d^{9}$, $\mathrm{Co^{3+}}$ is $3d^{6}$. Find the oxidation state from the whole formula — in $\mathrm{K_{3}[Co(C_{2}O_{4})_{3}]}$ oxalate is $-2$, so cobalt is $+3$.
2. Treating H₂O as always weak. $\mathrm{[Co(H_{2}O)_{6}]^{3+}}$ and $\mathrm{[Co(NH_{3})_{6}]^{3+}}$ are low spin and diamagnetic; $\mathrm{[CoF_{6}]^{3-}}$ is high spin with 4 unpaired electrons. $\mathrm{Co^{3+}}$ is the classic exception; $\mathrm{[Co(NH_{3})_{6}]^{2+}}$, by contrast, is high spin.
3. "Strong field ⇒ low spin" for every geometry and every $d^{n}$. Tetrahedral complexes are high spin; octahedral $d^{8}$ always has 2 unpaired electrons (so $\mathrm{[Ni(NH_{3})_{6}]^{2+}}$ is $sp^{3}d^{2}$, outer orbital); $d^{3}$ always has 3.
4. Misusing $\Delta_{t}\approx\tfrac49\Delta_{o}$. It compares a tetrahedral and an octahedral complex of the same metal and ligand at the same distance — not any two complexes.
5. Colour seen = colour absorbed. A complex absorbing yellow light looks violet; a $d^{0}$ or $d^{10}$ ion (Sc³⁺, Ti⁴⁺, Zn²⁺, Cu⁺) has no $d$–$d$ transition and is colourless — unless an anhydrous salt lacks ligands altogether, as in white $\mathrm{CuSO_{4}}$.
For an octahedral $d^{5}$ ion with $\Delta_{o}\lt P$, the configuration and CFSE (pairing energy ignored) are
Colour of complexes and the importance of coordination compounds
White light contains all visible wavelengths, from violet (about $400\ \mathrm{nm}$) to red (about $750\ \mathrm{nm}$). A solution looks coloured when it absorbs part of this range; our eye sees the light that is left, which has the complementary colour of the light absorbed. Complementary colours lie opposite each other on the colour wheel.
| Wavelength absorbed (nm) | 400–430 | 430–490 | 490–560 | 560–590 | 590–620 | 620–750 |
|---|---|---|---|---|---|---|
| Colour absorbed | violet | blue | green | yellow | orange | red |
| Colour seen | yellow (-green) | orange | red-violet (purple) | violet | blue | blue-green |
In a complex with a partly filled $d$ subshell, a photon whose energy matches the gap can lift an electron from the lower set to the upper one — a $d$–$d$ transition. For the octahedral $d^{1}$ ion $\mathrm{[Ti(H_{2}O)_{6}]^{3+}}$ this is simply $t_{2g}^{1}e_{g}^{0}\rightarrow t_{2g}^{0}e_{g}^{1}$. The absorption band is centred near $498\ \mathrm{nm}$ and is broad, covering much of the green–yellow region; what passes through is red and violet light, which together look purple.
- The molar energy of the absorbed light equals the splitting: $$\Delta=\frac{N_{A}hc}{\lambda}=\frac{1.196\times10^{5}}{\lambda/\mathrm{nm}}\ \mathrm{kJ\,mol^{-1}}.$$ A larger $\Delta$ means absorption at a shorter wavelength.
- The colour seen is the complement of the colour absorbed.
- A $d$–$d$ transition needs at least one electron in the lower set and a vacancy in the upper set: $d^{0}$ and $d^{10}$ ions give no $d$–$d$ colour ($\mathrm{Sc^{3+}},\ \mathrm{Ti^{4+}},\ \mathrm{Zn^{2+}},\ \mathrm{Cu^{+}},\ \mathrm{Ag^{+}}$ complexes are colourless).
- Without ligands there is no splitting and no $d$–$d$ colour: anhydrous $\mathrm{CuSO_{4}}$ is white, $\mathrm{CuSO_{4}\cdot5H_{2}O}$ (with $\mathrm{[Cu(H_{2}O)_{4}]^{2+}}$ units) is blue.
Changing the ligand changes the colour. $\mathrm{[Ni(H_{2}O)_{6}]^{2+}}$ is green. Replace water with ethane-1,2-diamine and $\Delta_{o}$ increases; the absorption moves to shorter wavelength and $\mathrm{[Ni(en)_{3}]^{2+}}$ is violet. Similarly pale blue $\mathrm{[Cu(H_{2}O)_{4}]^{2+}}$ turns deep blue-violet $\mathrm{[Cu(NH_{3})_{4}]^{2+}}$ with ammonia. Intensely coloured species such as $\mathrm{MnO_{4}^{-}}$ and $\mathrm{CrO_{4}^{2-}}$ ($d^{0}$) owe their colour to a different process, charge transfer from ligand to metal, not to $d$–$d$ transitions; that is why they are so much more intense.
In the crystal field lab choose octahedral, $d^{6}$, $\Delta_{o}=270\ \mathrm{kJ\,mol^{-1}}$ and $P=250\ \mathrm{kJ\,mol^{-1}}$ (close to $\mathrm{[Co(NH_{3})_{6}]^{3+}}$). What will the readouts show?
Importance of coordination compounds. Complex formation is not a laboratory curiosity; it runs through analysis, metallurgy and life itself.
1. Qualitative and quantitative analysis. Many tests for metal ions are complex-forming reactions with a characteristic colour or solubility:
- $\mathrm{Ni^{2+}}$ gives a red precipitate with dimethylglyoxime ($\mathrm{H_{2}dmg}$, $\mathrm{C_{4}H_{8}N_{2}O_{2}}$) in ammoniacal solution: $$\begin{aligned}&\mathrm{Ni^{2+}(aq)+2C_{4}H_{8}N_{2}O_{2}(aq)}\\&\quad\rightarrow\mathrm{[Ni(C_{4}H_{7}N_{2}O_{2})_{2}](s)+2H^{+}(aq)}\end{aligned}$$ The chelate is square planar and diamagnetic.
- $\mathrm{Cu^{2+}}$ with excess ammonia gives deep blue $\mathrm{[Cu(NH_{3})_{4}]^{2+}}$: $$\begin{aligned}&\mathrm{[Cu(H_{2}O)_{4}]^{2+}(aq)+4NH_{3}(aq)}\\&\quad\rightarrow\mathrm{[Cu(NH_{3})_{4}]^{2+}(aq)+4H_{2}O(l)}\end{aligned}$$
- $\mathrm{Fe^{3+}}$ gives a blood-red colour with thiocyanate, $$\begin{aligned}&\mathrm{[Fe(H_{2}O)_{6}]^{3+}(aq)+SCN^{-}(aq)}\\&\quad\rightarrow\mathrm{[Fe(H_{2}O)_{5}(SCN)]^{2+}(aq)+H_{2}O(l)}\end{aligned}$$ and Prussian blue with hexacyanidoferrate(II): $$\begin{aligned}&\mathrm{4Fe^{3+}(aq)+3[Fe(CN)_{6}]^{4-}(aq)}\\&\quad\rightarrow\mathrm{Fe_{4}[Fe(CN)_{6}]_{3}(s)}\end{aligned}$$
- Silver chloride dissolves in aqueous ammonia as the linear $sp$ complex: $$\begin{aligned}&\mathrm{AgCl(s)+2NH_{3}(aq)}\\&\quad\rightarrow\mathrm{[Ag(NH_{3})_{2}]^{+}(aq)+Cl^{-}(aq)}\end{aligned}$$ This separates $\mathrm{Ag^{+}}$ from $\mathrm{Pb^{2+}}$ in the chloride group.
- The hardness of water is estimated by titration with EDTA, which forms stable 1 : 1 chelates with $\mathrm{Ca^{2+}}$ and $\mathrm{Mg^{2+}}$.
2. Extraction and purification of metals. Silver and gold are leached from their ores by aerated cyanide solution, which turns the metal into a soluble complex; zinc then displaces the noble metal:
$$\begin{aligned}&\mathrm{4Au(s)+8CN^{-}(aq)+2H_{2}O(l)+O_{2}(g)}\\&\quad\rightarrow\mathrm{4[Au(CN)_{2}]^{-}(aq)+4OH^{-}(aq)}\end{aligned}$$ $$\begin{aligned}&\mathrm{2[Au(CN)_{2}]^{-}(aq)+Zn(s)}\\&\quad\rightarrow\mathrm{[Zn(CN)_{4}]^{2-}(aq)+2Au(s)}\end{aligned}$$In the Mond process impure nickel is converted into the volatile complex $\mathrm{[Ni(CO)_{4}]}$ at about $330$–$350\ \mathrm{K}$, which is decomposed at about $450$–$470\ \mathrm{K}$ to give very pure nickel: $\mathrm{Ni(s)+4CO(g)\rightleftharpoons[Ni(CO)_{4}](g)}$.
3. Biological systems. Life depends on metal complexes with large chelating ring ligands:
- Haemoglobin, the oxygen carrier of the blood, contains iron(II) held by the four nitrogen atoms of a porphyrin ring (haem) and a fifth nitrogen from the protein; $\mathrm{O_{2}}$ binds reversibly at the sixth position. Carbon monoxide binds there far more strongly, which is why it is poisonous.
- Chlorophyll, the green pigment of photosynthesis, is a complex of magnesium(II) in a porphyrin-like (chlorin) ring.
- Vitamin B₁₂ (cyanocobalamin), needed to prevent pernicious anaemia, is a complex of cobalt(III) in a corrin ring.
- In medicine, EDTA (as its calcium salt) removes lead from the body in lead poisoning, and the square planar complex cis-$\mathrm{[PtCl_{2}(NH_{3})_{2}]}$ (cisplatin) is an anticancer drug.
An aqueous complex has its strongest absorption at about $600\ \mathrm{nm}$. Its colour is most likely
Exercises
Unless told otherwise: spin-only moments in BM to two decimal places; CFSE in units of $\Delta_{o}$ (or $\Delta_{t}$) with pairing energy ignored; $N_{A}hc=1.196\times10^{5}\ \mathrm{kJ\,mol^{-1}\,nm}$.
- 1 Find the oxidation state and the number of $d$ electrons of the metal in $\mathrm{K_{3}[Fe(CN)_{6}]}$.
- 2 Give the hybridisation, geometry and magnetic moment of $\mathrm{[Zn(NH_{3})_{4}]^{2+}}$.
- 3 $\mathrm{[Ni(NH_{3})_{6}]^{2+}}$: give the hybridisation, say whether it is an inner- or outer-orbital complex, and find the number of unpaired electrons.
- 4 Give the hybridisation of $\mathrm{[Cr(NH_{3})_{6}]^{3+}}$ and its spin-only magnetic moment.
- 5 The measured moment of $\mathrm{K_{3}[Mn(CN)_{6}]}$ is about $3.2\ \mathrm{BM}$. Find the number of unpaired electrons and the hybridisation of the manganese.
- 6 $\mathrm{[Ni(CO)_{4}]}$ is tetrahedral and diamagnetic. Give the oxidation state of nickel and describe the bonding in VBT terms.
- 7 $\mathrm{[PtCl_{4}]^{2-}}$ is diamagnetic. Deduce its hybridisation and geometry.
- 8 Give two experimental facts about $\mathrm{[Ti(H_{2}O)_{6}]^{3+}}$ or $\mathrm{[Cu(NH_{3})_{4}]^{2+}}$ that VBT cannot explain.
- 1 Name the $e_{g}$ and $t_{2g}$ orbitals of an octahedral complex and give the energy of each set relative to the barycentre, in units of $\Delta_{o}$.
- 2 An octahedral complex has $\Delta_{o}=18\,000\ \mathrm{cm^{-1}}$. Estimate $\Delta_{t}$ for the tetrahedral complex of the same metal ion and ligand at the same distance.
- 3 Arrange in order of increasing $\Delta_{o}$: $\mathrm{[Co(CN)_{6}]^{3-}}$, $\mathrm{[CoF_{6}]^{3-}}$, $\mathrm{[Co(NH_{3})_{6}]^{3+}}$, $\mathrm{[Co(H_{2}O)_{6}]^{3+}}$.
- 4 Explain why $\Delta_{o}$ is larger for $\mathrm{[Co(H_{2}O)_{6}]^{3+}}$ than for $\mathrm{[Co(H_{2}O)_{6}]^{2+}}$.
- 5 Why are complexes of $4d$ and $5d$ metals almost always low spin?
- 6 In a tetrahedral field, why do the $d_{xy}$, $d_{yz}$, $d_{xz}$ orbitals lie above $d_{x^{2}-y^{2}}$ and $d_{z^{2}}$?
- 7 Find the number of $d$ electrons on the metal in (a) $\mathrm{[Co(en)_{3}]Cl_{3}}$, (b) $\mathrm{K_{3}[Cr(C_{2}O_{4})_{3}]}$, (c) $\mathrm{[Cu(NH_{3})_{4}]SO_{4}}$.
- 8 Thiocyanate can bind through S or through N. Which end gives the larger $\Delta_{o}$? Why can the point-charge model not explain the position of CO at the top of the spectrochemical series?
- 1 Write the configuration and CFSE of an octahedral $d^{3}$ ion such as $\mathrm{Cr^{3+}}$.
- 2 For octahedral $\mathrm{Co^{2+}}$ ($d^{7}$), give the configuration, the number of unpaired electrons and the CFSE for the high-spin and the low-spin arrangements.
- 3 $\mathrm{[Fe(CN)_{6}]^{3-}}$: write the configuration and find the spin-only moment and the CFSE (show the pairing term).
- 4 Tetrahedral $\mathrm{[MnCl_{4}]^{2-}}$: find the configuration, CFSE and spin-only moment.
- 5 Tetrahedral $\mathrm{[CoCl_{4}]^{2-}}$: find the CFSE in units of $\Delta_{t}$.
- 6 For an $\mathrm{Mn^{3+}}$ complex, $\Delta_{o}=250\ \mathrm{kJ\,mol^{-1}}$ and $P=300\ \mathrm{kJ\,mol^{-1}}$. Decide the spin state and find the CFSE in $\mathrm{kJ\,mol^{-1}}$.
- 7 Why can an octahedral $d^{8}$ ion not be either high spin or low spin, while $d^{4}$–$d^{7}$ can?
- 8 An octahedral iron(II) complex has $\mu=4.90\ \mathrm{BM}$. Write its $t_{2g}/e_{g}$ configuration and the VBT hybridisation, and say what kind of ligand it has.
- 1 $\mathrm{[Cu(H_{2}O)_{6}]^{2+}}$ has a broad absorption band centred near $800\ \mathrm{nm}$ that extends well into the red. Find the corresponding $\Delta_{o}$ in $\mathrm{kJ\,mol^{-1}}$ and explain the pale blue colour.
- 2 Explain why $\mathrm{[Sc(H_{2}O)_{6}]^{3+}}$ and $\mathrm{[Zn(H_{2}O)_{6}]^{2+}}$ are colourless.
- 3 Green $\mathrm{[Ni(H_{2}O)_{6}]^{2+}}$ turns violet when ethane-1,2-diamine is added. Explain the change of colour.
- 4 A complex absorbs most strongly at $600\ \mathrm{nm}$. What colour does it appear, and what is $\Delta$ in $\mathrm{kJ\,mol^{-1}}$?
- 5 Anhydrous copper(II) sulphate is white, but the pentahydrate is blue. Explain.
- 6 Write the equation for the leaching of gold with aerated cyanide solution. How many moles of $\mathrm{CN^{-}}$ are consumed per mole of gold, and what is the role of zinc in the next step?
- 7 Name the metal and give its oxidation state in (a) haemoglobin, (b) chlorophyll, (c) vitamin B₁₂.
- 8 $\mathrm{Ni^{2+}}$ is detected with dimethylglyoxime. Give the colour of the product and the number of unpaired electrons in this square planar complex. Why does $\mathrm{AgCl}$ dissolve in aqueous ammonia, and what is the hybridisation of silver in the product?
E1. JEE Main · MCQ Which of the following is an outer-orbital complex? (A) $\mathrm{[Fe(CN)_{6}]^{4-}}$ (B) $\mathrm{[Co(NH_{3})_{6}]^{3+}}$ (C) $\mathrm{[Ni(NH_{3})_{6}]^{2+}}$ (D) $\mathrm{[Cr(NH_{3})_{6}]^{3+}}$
E2. JEE Main · Numerical The spin-only magnetic moment of $\mathrm{[Mn(H_{2}O)_{6}]^{2+}}$ is $\mu$ BM. Find the value of $\mu^{2}$. Give the answer to the nearest integer.
E3. JEE Main · MCQ Consider the octahedral cobalt(III) complexes $\mathrm{[Co(CN)_{6}]^{3-}}$, $\mathrm{[Co(NH_{3})_{6}]^{3+}}$, $\mathrm{[Co(H_{2}O)_{6}]^{3+}}$ and $\mathrm{[CoF_{6}]^{3-}}$. Writing each complex by its ligand, the correct order of the wavelength of light absorbed is (A) $\mathrm{CN^{-}\lt NH_{3}\lt H_{2}O\lt F^{-}}$ (B) $\mathrm{F^{-}\lt H_{2}O\lt NH_{3}\lt CN^{-}}$ (C) $\mathrm{NH_{3}\lt CN^{-}\lt F^{-}\lt H_{2}O}$ (D) $\mathrm{H_{2}O\lt F^{-}\lt CN^{-}\lt NH_{3}}$
E4. JEE Main · Numerical An octahedral titanium(III) complex has a single absorption band at $500\ \mathrm{nm}$. Find the magnitude of its crystal field stabilisation energy in $\mathrm{kJ\,mol^{-1}}$. ($N_{A}=6.02\times10^{23}\ \mathrm{mol^{-1}}$, $h=6.63\times10^{-34}\ \mathrm{J\,s}$, $c=3.00\times10^{8}\ \mathrm{m\,s^{-1}}$.) Give the answer to the nearest integer.
E5. JEE Main · MCQ Which statement about crystal field splitting is correct? (A) For the same metal, ligand and distance, $\Delta_{t}$ is larger than $\Delta_{o}$. (B) In a tetrahedral field the $d_{x^{2}-y^{2}}$ and $d_{z^{2}}$ orbitals are raised above the barycentre. (C) A low-spin octahedral $d^{6}$ complex has CFSE $-0.4\Delta_{o}$. (D) Tetrahedral complexes are almost always high spin because $\Delta_{t}$ is usually smaller than the pairing energy.
E6. JEE Main · Numerical How many of the following are paramagnetic? $\mathrm{[Fe(CN)_{6}]^{4-}}$, $\mathrm{[Fe(CN)_{6}]^{3-}}$, $\mathrm{[Co(NH_{3})_{6}]^{3+}}$, $\mathrm{[CoF_{6}]^{3-}}$, $\mathrm{[Ni(CN)_{4}]^{2-}}$, $\mathrm{[NiCl_{4}]^{2-}}$, $\mathrm{[Cu(NH_{3})_{4}]^{2+}}$, $\mathrm{[Zn(NH_{3})_{4}]^{2+}}$, $\mathrm{[Cr(H_{2}O)_{6}]^{3+}}$. Give the answer to the nearest integer.
Research question. As water ligands around $\mathrm{Cu^{2+}}$ are replaced step by step by ammonia, how does the colour of the solution change, and can a phone camera follow the shift towards stronger ligand fields?
Why it matters. Colorimetry — measuring the colour of a solution to find a concentration or a ligand environment — is used in water testing (copper, iron and nitrite kits), in clinical laboratories and in field analysis in places without expensive spectrometers. Low-cost "smartphone spectrometers" are an active research area for environmental monitoring in developing regions.
What to do.
- Under supervision, make $25\ \mathrm{mL}$ of about $0.05\ \mathrm{mol\,L^{-1}}$ copper(II) sulphate solution in each of six identical clear vials.
- Add $0,\ 0.5,\ 1,\ 2,\ 4$ and $8\ \mathrm{mL}$ of $2\ \mathrm{mol\,L^{-1}}$ aqueous ammonia (in a fume cupboard or well-ventilated room), topping up with water to equal volumes. Note any pale blue precipitate of copper(II) hydroxide that forms and redissolves.
- Photograph the vials against the same white screen in the same light, with the camera's white balance locked.
- Read the red, green and blue values at the centre of each vial with a free colour-picker app; compute the "absorbance" in each channel as $A=\log_{10}(I_\text{background}/I_\text{vial})$.
- Plot the three absorbances against the volume of ammonia. Decide which colour of light is absorbed most at each stage, and whether the absorbed band moves to shorter wavelength.
Tools: Copper(II) sulphate, dilute aqueous ammonia, measuring cylinders, six vials, a phone camera and colour-picker app, a spreadsheet; gloves and goggles.
Hand in: The photographs, a table of RGB values and absorbances, the plot, and one paragraph linking what you see to the spectrochemical series and Rule 12.13.
Going further. Repeat with nickel(II) sulphate and ethane-1,2-diamine (green → blue → violet as en replaces water). Which ligand produces the larger shift per molecule added, and why might a chelating ligand do so?
🧪 Self-check quiz
Q1. Which of the following is an inner-orbital octahedral complex?
Q2. The spin-only magnetic moment of $\mathrm{[Fe(H_{2}O)_{6}]^{3+}}$ is
Q3. The CFSE of an octahedral $d^{8}$ ion (pairing energy ignored) is
Q4. $\mathrm{[Ti(H_{2}O)_{6}]^{3+}}$ is purple because
Q5. Vitamin B₁₂ is a coordination compound of