Solid-State Defects and Non-Stoichiometry
A textbook diagram of NaCl shows a perfectly repeating lattice, every Na⁺ and Cl⁻ exactly where it should be. No real crystal above 0 K looks like that. Thermodynamics guarantees a finite population of point defects — missing ions, displaced ions, and in many transition-metal compounds, ions sitting in the "wrong" oxidation state — because those defects raise the crystal's entropy faster than they raise its enthalpy. This article goes past the simple Schottky/Frenkel headcount and builds the machinery chemists actually use to describe and quantify defect populations: Kröger-Vink notation, mass-action equilibria, and the non-stoichiometric compounds whose entire chemistry is controlled by how many defects they carry.
Writing defect reactions properly — Kröger-Vink notation
A defect reaction is written like any other chemical equation, but each species carries two extra pieces of information: the lattice site it occupies (as a subscript) and its effective charge relative to the perfect lattice at that site (as a superscript — a dot • for each unit of positive effective charge, a prime ′ for each unit of negative effective charge, and a cross ˣ for no effective charge).
Mi•• = an interstitial M²⁺ cation carrying effective charge +2
CaK• = a Ca²⁺ ion sitting on a K⁺ site, effective charge +1 (one unit more positive than the K⁺ it replaced)
Two rules make every Kröger-Vink equation checkable: mass balance (the same real atoms on both sides, vacancies contribute zero mass) and effective-charge balance (the sum of superscript charges must be zero on both sides — this is not optional, and it is what tells you how many vacancies a given dopant must create).
Intrinsic defects have an equilibrium constant, just like any reaction
Forming a Schottky pair (Fig.: a cation vacancy plus an anion vacancy) or a Frenkel pair (an ion displaced to an interstitial site) is written and treated exactly like a chemical equilibrium:
Frenkel (cation to interstitial): MM ⇌ Mi• + VM′ KF = [Mi•][VM′]
Because the defect concentrations are low, the "activity" of the host lattice does not appear (same reasoning as a solid appearing as 1 in a Ksp expression), and KS and KF are each temperature-dependent, growing with temperature exactly as a normal equilibrium constant does. This mass-action picture is the reason doped and undoped ionic solids behave so differently, as the third worked example below shows.
Non-stoichiometry — when the deviation is the chemistry
Most binary compounds are Daltonide — fixed, integer composition (NaCl is NaCl, full stop). A smaller but chemically important group of compounds, almost always built on a transition metal with an accessible second oxidation state, are Berthollide — their composition drifts continuously over a real range without a phase change. Two mechanisms produce this:
| Type | Mechanism | Charge compensation | Conduction type | Example |
|---|---|---|---|---|
| Metal excess | Anion vacancy traps an electron (F-centre), or extra interstitial cation | Trapped electron / extra electron | n-type |
The table cell above is intentionally the general statement — the two metal-excess routes and the metal-deficient route are laid out fully below, because each needs its own worked derivation.
| Type | What happens | Conductivity | Typical example |
|---|---|---|---|
| Metal excess (interstitial) | Extra M cations sit interstitially, each with a compensating extra electron nearby | n-type | Excess Zn²⁺ in ZnO, heated in Zn vapour |
| Metal excess (anion vacancy) | An anion vacancy exists where the electron that would have gone to the missing anion stays trapped in the cavity — an F-centre | n-type | NaCl heated in Na vapour (turns yellow) |
| Metal deficient | Cation vacancies are charge-compensated by oxidising some remaining cations to a higher state | p-type | Fe1−xO (wüstite), Cu2−xO |
Worked example 1 — writing and balancing an F-centre reaction. A crystal of NaCl is heated in sodium vapour, and some Na atoms deposit on the surface and diffuse in as Na⁺, leaving their electrons behind at anion vacancies.
Write the process starting from a Cl⁻ ion moving to the surface, leaving VCl•, and the incoming Na atom ionising to fill a surface cation site while its electron localises at the vacancy:
Na (vapour) → NaNaˣ + e′ (in F-centre) + ½Cl₂ (released)
Check the charge balance: NaNaˣ contributes 0 (a normal Na⁺ on a normal site), e′ contributes −1, and there is no cation vacancy created here — the trapped electron itself is the compensating negative species sitting in an anion vacancy that already existed or forms alongside it. The trapped electron behaves like a particle in a box, with discrete absorption levels in the visible range — which is exactly why F-centre NaCl is yellow and F-centre KCl is violet: the box size (lattice spacing) differs, so the absorbed wavelength differs.
Worked example 2 — the cation ratio inside Fe0.95O. Wüstite is always metal-deficient: Fe1−xO with x typically between about 0.05 and 0.12. Find what fraction of the iron is Fe³⁺ when x = 0.05.
Set up the two conservation laws. Let n₂ = mol Fe²⁺ and n₃ = mol Fe³⁺ per formula unit. Cation conservation: n₂ + n₃ = 1 − x. Charge balance against 1 mol of O²⁻ (charge −2): 2n₂ + 3n₃ = 2.
Substitute n₂ = (1 − x) − n₃ into the charge equation:
2[(1 − x) − n₃] + 3n₃ = 2
2(1 − x) + n₃ = 2
n₃ = 2x and n₂ = 1 − 3x
For x = 0.05: n₃ = 2(0.05) = 0.10 mol Fe³⁺, n₂ = 1 − 3(0.05) = 0.85 mol Fe²⁺. Check: 0.10 + 0.85 = 0.95 = 1 − x ✓, and 2(0.85) + 3(0.10) = 1.70 + 0.30 = 2.00 ✓.
Fraction of iron that is Fe³⁺ = 0.10 ÷ 0.95 = 10.5%. Every cation vacancy in the lattice is charge-compensated by exactly two Fe²⁺ ions being oxidised to Fe³⁺ — which is precisely why this deviation is metal-deficient and p-type: the extra positive charge sitting on the Fe³⁺ ions is compensated by holes that can hop from cation to cation, carrying current.
Extrinsic (doping-induced) defects — the aliovalent trick
Dissolving a foreign salt of a different cation charge into a host lattice is called aliovalent doping, and it creates vacancies deliberately, on demand, in a fixed ratio set purely by charge balance — no temperature-dependent equilibrium needed.
Worked example 3 — CaCl₂ dissolved into KCl. Ca²⁺ substitutes onto a K⁺ site. Write the Kröger-Vink reaction and find how many K⁺ vacancies form per mole of CaCl₂ dissolved.
CaCl₂ → CaK• + 2ClClˣ + VK′
CaK• carries effective charge +1 (Ca²⁺ replacing K⁺ is one unit more positive than the site "expects"). The two Cl⁻ occupy two normal Cl⁻ sites, contributing 0 effective charge each. Charge balance across the equation: left side is neutral (CaCl₂ is a neutral salt), so the right side must sum to zero: (+1) + 0 + 0 + (charge on VK) = 0, giving VK effective charge = −1, i.e. exactly one K⁺ vacancy per Ca²⁺ dissolved.
This is why the ratio is always 1:1, never 2:1 — one extra unit of positive charge from the substituted cation demands exactly one missing K⁺ to cancel it. At 1 mol% CaCl₂ doping, the extrinsic vacancy population is fixed at 1 mol% regardless of temperature, and at low-to-moderate temperature this number swamps the much smaller, exponentially temperature-dependent intrinsic Schottky population — the ionic conductivity of the doped crystal plateaus in this "extrinsic region" instead of rising steeply with T the way a pure crystal's does.
This same aliovalent-doping logic, scaled up, is what makes yttria-stabilised zirconia (Y₂O₃ substituted into ZrO₂) a practical solid electrolyte: each Y³⁺ replacing a Zr⁴⁺ leaves one O²⁻ vacancy, and at high temperature oxide ions hop between those vacancies fast enough to carry a usable ionic current — the working principle behind solid-oxide fuel cells and oxygen sensors.
Errors that appear most often
- Writing a Kröger-Vink species without checking effective-charge balance. An equation that does not sum to zero effective charge on both sides is not a valid defect reaction, however plausible it looks.
- Assuming every compound can be non-stoichiometric. Wide non-stoichiometry needs a metal with an accessible second oxidation state (Fe, Cu, Ti, Mn) or a lattice that tolerates interstitials; NaCl and MgO stay essentially Daltonide under ordinary conditions.
- Confusing intrinsic and extrinsic defects. Intrinsic (Schottky, Frenkel) populations obey a temperature-dependent K; extrinsic (dopant-induced) populations are fixed by the dopant concentration and do not grow exponentially with heating.
- Saying Frenkel defects change the crystal's density. Nothing leaves the crystal in a Frenkel defect — mass and volume are essentially unchanged, unlike a Schottky pair or a metal-deficient vacancy, both of which do lower the density.
- Treating "n-type" and "p-type" as properties only of covalent semiconductors. Ionic non-stoichiometric oxides carry exactly the same classification, driven by electron/hole excess rather than doped covalent bands.
Where this appears across postgraduate chemistry
| Context | What is actually asked |
|---|---|
| CSIR-NET / GATE inorganic | Kröger-Vink notation, Schottky/Frenkel mass-action expressions, classifying n-type vs p-type non-stoichiometry |
| IIT-JAM materials sections | Recognising which oxides are non-stoichiometric and why, F-centre colour explanations |
| Solid-state / materials chemistry research | Defect engineering for ionic conductors, catalysts and sensors — the aliovalent-doping logic above is the design principle |
Check the stoichiometry, not just the concept. Every defect-balance problem above reduces to careful mole and molar-mass arithmetic — exactly what the Molar Mass & Composition tool and the periodic table's oxidation-state data are built for.
Open the ABC Chemistry Calculator Suite →Preparing for CSIR-NET, GATE, IIT-JAM or CUET-PG? ABC Chemistry runs dedicated competitive-exam batches at its coaching centre and fully online for students across India — details at abcchemistry.in.