Wait, What? Give the same ligand to a row of divalent first-row transition-metal ions and their complex stabilities often rise in a striking order: Mn(II) < Fe(II) < Co(II) < Ni(II) < Cu(II), then fall at Zn(II). That pattern is not a periodic-table trivia fact. It is an empirical thermodynamic regularity produced by several chemical contributions acting together.
The direct answer
The Irving–Williams series describes the common increase in stability of many complexes of first-row divalent transition-metal ions from Mn(II) through Cu(II), followed by lower stability for Zn(II): Mn²⁺ < Fe²⁺ < Co²⁺ < Ni²⁺ < Cu²⁺ > Zn²⁺. The trend reflects a combination of increasing effective metal–ligand attraction as ionic radius contracts, ligand-field stabilisation across the d series, covalent contributions and the special structural/electronic behaviour of Cu(II). It is a robust empirical tendency, not an absolute ordering valid for every ligand, solvent, oxidation state, geometry or pH.
What ‘stability’ means here
In coordination chemistry, stability in this context is thermodynamic, usually expressed through equilibrium formation constants. For a simple step such as M + L ⇌ ML, the formation constant is K = a(ML)/[a(M)a(L)] when written in terms of activities. A larger K means the complex is more favoured at equilibrium under the defined standard-state framework.
Do not confuse this with kinetic inertness. A complex can be thermodynamically stable yet exchange ligands rapidly, or thermodynamically less favoured yet persist because ligand substitution is slow. Irving–Williams is about equilibrium stability, not how quickly a complex forms or falls apart.
The empirical observation came first
Irving and Williams noticed that many ligands gave a similar ordering for divalent first-row transition metals. The original work and subsequent datasets made the surprising point that ligand identity changes the absolute constants enormously, yet a broad metal-ion ordering often survives. This is precisely the kind of pattern Chemistry values: not universal enough to be a law, but regular enough to demand explanation.
Contribution 1: shrinking ionic size
Across the first-row divalent series, increasing effective nuclear charge tends to contract the metal ion. For the same +2 charge, a smaller ion generally produces a stronger electrostatic interaction with donor atoms and often shorter metal–ligand distances. This contributes to increasing formation stability from the left side of the series toward the later metals.
But size alone cannot produce the full shape. If it did, a simple monotonic trend would be expected. The decline at Zn(II) and the pronounced Cu(II) maximum tell us electronic structure matters.
Contribution 2: ligand-field stabilisation
In an approximately octahedral ligand field, the five d orbitals split into lower-energy t₂g and higher-energy eg sets. Different d-electron counts therefore gain different ligand-field stabilisation energies. High-spin Mn(II), d⁵, has no net octahedral LFSE in the simplest treatment. Moving through Fe(II), Co(II) and Ni(II) generally increases the electronic stabilisation available from ligand-field splitting.
Zn(II) is d¹⁰. In the simplest octahedral crystal-field bookkeeping, both t₂g and eg levels are filled, giving no net LFSE relative to the barycentre. This helps explain why the series falls after Cu.
Why Cu(II) sits at the top
Cu(II) is d⁹. An ideal octahedral d⁹ arrangement is electronically degenerate and therefore strongly susceptible to Jahn–Teller distortion. Many Cu(II) complexes elongate along one axis or otherwise lower symmetry, redistributing bond strengths and electronic energy. The resulting stabilisation, together with size and increasing covalent character, contributes to the characteristic Cu maximum.
It is important not to turn this into a single-cause slogan. The Jahn–Teller effect is real, but the Irving–Williams trend emerges from the total free energy of ligand exchange: metal–ligand bonding, hydration/desolvation, ligand solvation, electronic stabilisation and entropy all contribute.
The water-to-ligand exchange picture
A clean way to think about the series is to imagine replacing coordinated water around M²⁺ with another ligand L. The observed equilibrium is not merely ‘metal loves ligand’. The system pays the cost of reorganising and desolvating species and gains the benefits of new metal–ligand bonding plus released solvent and any chelate entropy effects.
That is why absolute formation constants cannot be predicted from LFSE alone. The reference states matter. The ligand’s protonation state matters. Ionic strength and activity corrections matter for high-quality thermodynamic work.
Why pH can rewrite the apparent order
Many ligands are acids or bases. If a donor group must deprotonate before binding, then the concentration of the binding-competent ligand depends on pH. Metals can also hydrolyse. A reported ‘conditional formation constant’ therefore folds side equilibria into the apparent binding strength under particular conditions.
Two laboratories can obtain different apparent affinities without disagreeing about fundamental Chemistry if their pH, ionic medium, competing ions or ligand protonation states differ. This is why a naked log K value without conditions is incomplete evidence.
Connection to the chelate effect — but not ownership takeover
Multidentate ligands can form more stable complexes than comparable monodentate ligands because preorganisation, solvation and entropy change the total free energy. That is the chelate-effect job. Irving–Williams asks a different question: when the ligand framework is held broadly comparable, how does changing the metal ion reshape stability? The two ideas interact, but they are not the same canonical concept.
Connection to HSAB
Hard–soft acid–base reasoning compares polarizability and charge density to explain preferential interactions. Irving–Williams is a narrower empirical order for a particular oxidation-state row. HSAB can help explain why changing donor atoms creates exceptions or changes selectivity, especially with softer sulfur donors, but HSAB does not simply replace the series.
Observation versus inference
The observation is a set of equilibrium constants measured under defined conditions. The series is a recurring pattern in those data. Assigning percentages of the trend to ionic radius, LFSE, covalency, hydration and Jahn–Teller stabilisation is model-based interpretation. Those models become stronger when they simultaneously explain structural data, spectroscopy, thermodynamic cycles and families of ligands.
Why exceptions are scientifically valuable
An exception is not an embarrassment. If a ligand does not follow the expected sequence, ask what changed chemically. Is the donor atom softer? Did the coordination geometry switch? Is one metal in a different spin state? Is the ligand deprotonated for one ion but not another? Did redox chemistry intervene? Did precipitation or hydrolysis remove a species? Each exception can identify a hidden variable.
A Singapore learning progression
At lower-secondary level, learners first need the idea that ions differ in charge and size and can form compounds with characteristic properties. At O-Level/SEC Chemistry, transition metals, oxidation states, colour and bonding provide a bridge. At JC, equilibria, energetics, electronic structure and complex ions can be connected. Undergraduate inorganic chemistry adds ligand-field theory, stepwise formation constants and thermodynamic cycles. Professional coordination chemistry adds speciation models, activity corrections, calorimetry, spectroscopy, structural methods and computational bonding analysis.
Again, this is a conceptual progression rather than an exam checklist. The mature question is: under the actual chemical conditions, which metal–ligand species exist and what free-energy terms determine their populations?
Common misconceptions
- “Cu(II) always binds every ligand most strongly.” The series is a broad tendency with chemical boundaries.
- “Irving–Williams is just ionic radius.” Radius contributes, but electronic and solvation terms matter.
- “The Cu maximum is only Jahn–Teller.” Jahn–Teller contributes; total stability is a many-term free energy.
- “Zn(II) is weak because it has no d electrons.” Zn(II) is d¹⁰, not d⁰; its simple octahedral LFSE is zero.
- “A larger formation constant means a slower dissociation rate.” Thermodynamic stability and kinetic inertness are distinct.
- “A log K is meaningful without conditions.” pH, ionic medium, temperature and competing equilibria can be decisive.
A formation-constant checkpoint
If K₁ describes M + L ⇌ ML and K₂ describes ML + L ⇌ ML₂, the cumulative constant β₂ = K₁K₂. Stepwise and cumulative constants answer different bookkeeping questions. When comparing Irving–Williams behaviour, make sure the constants being compared refer to equivalent stoichiometries and conditions. Comparing K₁ for one metal with β₂ for another is not chemical insight; it is a category error.
How we know
Potentiometric titration can constrain protonation and complex-formation equilibria. Spectrophotometry can separate species when their spectra differ. Calorimetry can partition free-energy changes into enthalpic and entropic components. X-ray and spectroscopic methods reveal coordination geometry. Modern global speciation fitting tests whether one set of constants explains multiple observations simultaneously.
The original Irving–Williams work established the recurring order empirically. Later structural and electronic models explain why the pattern is plausible, while modern data on extraction, biological competition and unusual donors show where additional terms become important.
Transfer checks
- Why does a smaller M²⁺ ion often form stronger bonds with the same donor set, all else equal?
- Why does d¹⁰ Zn(II) not receive the same simple octahedral LFSE contribution as Ni(II)?
- If a ligand is mostly protonated at the working pH, why can its conditional metal-binding strength look much smaller?
- A Cu(II) complex forms rapidly but exchanges ligands rapidly. Does that contradict a large equilibrium formation constant?
Delayed independent reasoning check
Tomorrow, reconstruct the series from memory and then refuse to stop at memorisation. Give at least four contributors: ionic contraction, ligand-field stabilisation, Cu(II) Jahn–Teller/electronic effects, and the loss of LFSE at d¹⁰ Zn(II). Then name two boundaries such as pH/speciation and ligand donor type. That reconstruction is the difference between recalling an order and understanding why it is only conditionally predictive.
Practical interpretation
When you encounter metal-binding data, compare like with like: oxidation state, stoichiometry, temperature, ionic medium, pH and ligand form. Ask whether constants are thermodynamic or conditional. Check for hydrolysis, precipitation and redox side reactions. Then use Irving–Williams as a prior expectation, not as a replacement for measurement.
Model limits
The canonical series is most defensible for divalent first-row transition metals under broadly comparable coordination conditions. Strong changes in ligand softness, geometry, spin state, solvent, oxidation state, covalency or coupled chemistry can change the order. Cu(II) can adopt varied geometries; Fe(II) and Co(II) can change spin state; some ligands alter redox chemistry. A robust rule becomes brittle when its operating conditions are forgotten.
Evidence and further reading
- Irving & Williams, Nature (1948): Order of Stability of Metal Complexes
- Nature (1954): Electronegativity and the Stability of Metal Complexes
- Chemistry LibreTexts: the Irving–Williams series
- Peer-reviewed analysis of stability constants across the Irving–Williams order
Connect within eduKateSengkang
- How to Learn Hard and Soft Acids and Bases
- How to Learn the Chelate and Macrocyclic Effects
- How to Learn the Jahn–Teller Effect
- eduKateSengkang Science Hub
The quiet return
The Irving–Williams series is valuable because it is neither a random list nor an inviolable law. It is a stable chemical pattern that survives many ligand changes because several periodic and electronic trends push in the same broad direction. Its deepest lesson is methodological: recognise the pattern, explain the competing free-energy contributions, and then look deliberately for the conditions that make the pattern bend.
