Defect Equilibria in Functional Oxides: Kröger–Vink Notation, Mass Action, and Brouwer Diagrams
Point defects determine much of the functional behaviour of oxide ceramics. Oxygen vacancies can enable oxygen-ion transport; electronic carriers can produce leakage or mixed conduction; acceptor dopants can alter vacancy concentrations; and hydration can introduce mobile protons. These effects are coupled through defect equilibria, charge neutrality and the chemical potentials imposed by the surrounding atmosphere.
This article develops a quantitative framework for analysing point defects in oxides. It focuses on Kröger–Vink notation, mass-action expressions, charge neutrality, aliovalent doping, oxygen partial-pressure dependence and Brouwer diagrams. The aim is not to assign one universal defect model to every oxide, but to show how a model is constructed, tested and limited.
1. Point defects and Kröger–Vink notation
A Kröger–Vink symbol describes a defect species by its identity, lattice site and effective charge relative to the perfect crystal.
The general notation is
\[ \mathrm{M_{S}^{q}} \]
where \(M\) identifies the species, \(S\) identifies the lattice site, and \(q\) is the effective charge.
The superscripts commonly used are:
- \(x\): zero effective charge;
- \(\bullet\): one positive effective charge;
- \(\bullet\bullet\): two positive effective charges;
- \('\): one negative effective charge;
- \(''\): two negative effective charges.
For an oxide with a regular oxygen site and a regular cation site, common defects include:
| Symbol | Meaning |
|---|---|
| \(\mathrm{O_O^x}\) | Oxygen ion on a regular oxygen site |
| \(\mathrm{V_O^{\bullet\bullet}}\) | Doubly positively charged oxygen vacancy |
| \(\mathrm{M_M^x}\) | Host cation on its regular site |
| \(\mathrm{V_M^{''''}}\) | Cation vacancy with four negative effective charges, for a tetravalent cation site |
| \(\mathrm{e'}\) | Electron |
| \(\mathrm{h^{\bullet}}\) | Hole |
| \(\mathrm{OH_O^{\bullet}}\) | Proton associated with an oxygen ion on an oxygen site |
| \(\mathrm{O_i^{''}}\) | Doubly negatively charged oxygen interstitial |
The effective charge is defined relative to the perfect lattice site, not as the formal oxidation state of the atom alone. For example, an oxygen vacancy in a simple ionic oxide is often written \(\mathrm{V_O^{\bullet\bullet}}\) because removing an O²⁻ ion leaves a site with two positive effective charges relative to the perfect lattice.
The actual charge state can differ in materials where vacancies trap electrons or exhibit multiple charge states. The assumed charge state must therefore be consistent with the electronic structure and experimental conditions.
2. Writing defect reactions correctly
A defect reaction must conserve:
- each chemical element;
- lattice sites, where relevant;
- total effective charge.
Consider oxygen loss from an oxide:
\[ \mathrm{O_O^x \rightleftharpoons V_O^{\bullet\bullet} +\tfrac{1}{2}O_2(g)+2e'} \]
One lattice oxygen leaves as half an oxygen molecule. The oxygen vacancy has two positive effective charges, balanced by two electrons.
The reverse reaction describes oxygen incorporation:
\[ \mathrm{V_O^{\bullet\bullet} +\tfrac{1}{2}O_2(g)+2e' \rightleftharpoons O_O^x} \]
These equations are charge-balanced in effective-charge notation. They do not imply that the electrons must be free band electrons; in some oxides they may localize on cations or form small polarons. If localization is important, the electronic species should be represented accordingly.
A useful balancing check
For every proposed reaction, sum the effective charges on both sides. For the oxygen-loss reaction:
\[ 0=(+2)+2(-1) \]
The effective charge is conserved. Then verify oxygen atom balance and the number of lattice sites represented.
3. Mass-action expressions
For a reaction written as
\[ \sum_i \nu_i A_i=0 \]
the thermodynamic equilibrium constant is
\[ K=\prod_i a_i^{\nu_i} \]
where \(a_i\) is the activity of species \(i\) and the stoichiometric coefficient is positive for products and negative for reactants.
In dilute-defect approximations, activities of defects are often approximated by normalized concentrations. For a defect on a specific sublattice,
\[ a_i\approx\frac{[i]}{N_i} \]
where \(N_i\) is the concentration of available sites on that sublattice. Some treatments absorb site-density factors into an effective equilibrium constant and write concentrations directly. State which convention is being used.
For gaseous oxygen, the dimensionless activity is
\[ a_{\mathrm{O_2}}=\frac{p_{\mathrm{O_2}}}{p^\circ} \]
for an ideal gas, where \(p^\circ\) is the standard pressure. Using \(p_{\mathrm{O_2}}\) directly is a common shorthand when the pressure unit and reference are fixed, but a thermodynamic equilibrium constant is formally dimensionless.
Oxygen-loss equilibrium
For
\[ \mathrm{O_O^x \rightleftharpoons V_O^{\bullet\bullet} +\tfrac{1}{2}O_2(g)+2e'} \]
the dilute-defect mass-action expression is commonly written
\[ K_{\mathrm{red}}= [V_O^{\bullet\bullet}]\, n^2\, a_{\mathrm{O_2}}^{1/2} \]
where \(n\) denotes the normalized electron concentration under the adopted convention. The activity of the regular oxygen site is treated as approximately constant and absorbed into \(K_{\mathrm{red}}\).
This expression shows that oxygen pressure, vacancy concentration and electron concentration are coupled. A change in oxygen partial pressure cannot generally be interpreted by considering vacancies alone.
4. Charge neutrality
A bulk oxide must remain electrically neutral on a macroscopic scale. The charge-neutrality condition is obtained by summing the concentration of every defect multiplied by its effective charge.
For a simplified oxide containing doubly charged oxygen vacancies, electrons, holes and singly charged acceptors \(A'\),
\[ 2[V_O^{\bullet\bullet}]+[h^{\bullet}] = [n]+[A'] \]
The equation must be expanded if other charged species are present, such as cation vacancies, interstitials, donor dopants or protonic defects.
Charge neutrality is not an optional extra equation. It is needed alongside mass-action relationships and site-balance constraints to determine the defect concentrations.
Intrinsic electronic equilibrium
For electron–hole generation,
\[ \mathrm{\varnothing \rightleftharpoons e'+h^{\bullet}} \]
the mass-action relation in a simple non-degenerate semiconductor approximation is
\[ np=K_i(T) \]
where \(K_i(T)\) is the intrinsic carrier product. It is temperature-dependent and should not be treated as a universal constant.
The combination of \(np=K_i\) and charge neutrality determines the electron and hole concentrations when the relevant defect model is specified.
5. Oxygen partial pressure and defect regimes
The oxygen partial pressure can shift the dominant defect populations. The direction and magnitude of the response depend on the equilibria and charge-compensation regime.
Intrinsic reduction regime
Consider the oxygen-loss reaction and assume:
- doubly charged oxygen vacancies are the dominant ionic defects;
- electrons are the dominant electronic carriers;
- holes and fixed dopants are negligible;
- the dilute-defect approximation applies;
- charge neutrality is therefore \(n\approx2[V_O^{\bullet\bullet}]\).
Substituting \(n=2[V_O^{\bullet\bullet}]\) into the mass-action expression gives
\[ K_{\mathrm{red}} = 4[V_O^{\bullet\bullet}]^3 a_{\mathrm{O_2}}^{1/2} \]
Therefore,
\[ [V_O^{\bullet\bullet}] \propto a_{\mathrm{O_2}}^{-1/6} \]
and, because \(n\approx2[V_O^{\bullet\bullet}]\),
\[ n\propto a_{\mathrm{O_2}}^{-1/6} \]
The logarithmic slope is
\[ \frac{d\log[V_O^{\bullet\bullet}]} {d\log a_{\mathrm{O_2}}} =-\frac{1}{6} \]
This slope is not universal. It follows specifically from the stated reaction, vacancy charge state and charge-neutrality assumptions.
Acceptor-doped regime
If acceptor dopants are present and fully compensated by oxygen vacancies, charge neutrality may be approximated by
\[ 2[V_O^{\bullet\bullet}]\approx[A'] \]
If the acceptor concentration is fixed and the approximation remains valid, the vacancy concentration is approximately pinned by the dopant concentration over that regime. The electron concentration can still vary with oxygen pressure through the reduction equilibrium.
Using the same mass-action expression,
\[ n^2\propto a_{\mathrm{O_2}}^{-1/2} \]
so that
\[ n\propto a_{\mathrm{O_2}}^{-1/4} \]
This result assumes that the vacancy concentration is effectively constant and that electrons are the relevant electronic carriers. It does not apply automatically at very reducing conditions, where electronic compensation may become substantial, or at conditions where defect association changes the free-vacancy population.
6. Aliovalent doping and oxygen-vacancy formation
Substitution of a lower-valence cation on a higher-valence host site can introduce a negatively charged acceptor defect. In a simple tetravalent-cation oxide, trivalent dopant incorporation can be represented schematically as
\[ \mathrm{M_2O_3 \rightarrow 2M_M'+V_O^{\bullet\bullet}+3O_O^x} \]
Here, \(M_M'\) is a trivalent dopant on a tetravalent host-cation site. The reaction is a simplified incorporation expression; the actual precursor reaction and phase chemistry depend on the host oxide and processing conditions.
If the dopant is fully compensated by oxygen vacancies,
\[ [V_O^{\bullet\bullet}] \approx\frac{[M_M']}{2} \]
This is a charge-compensation limit, not proof that every nominal vacancy is mobile or unassociated. Dopant–vacancy association can reduce the concentration of free vacancies and influence ionic conductivity.
Donor doping
A higher-valence cation on a lower-valence host site can create a positively charged donor defect. In a simple oxide, electronic compensation may be represented as
\[ \mathrm{D_M^{\bullet}+e'} \]
as a charge-neutral defect pair. Alternatively, compensation may involve negatively charged cation vacancies or other defects, depending on composition and oxygen chemical potential.
Do not assume a donor dopant always produces mobile electrons. The compensation mechanism must be determined from the defect equilibria and experimental evidence.
7. Hydration and protonic defects
In many acceptor-doped perovskite oxides, water vapour can hydrate oxygen vacancies. A common hydration reaction is
\[ \mathrm{H_2O(g)+V_O^{\bullet\bullet} +O_O^x \rightleftharpoons 2OH_O^{\bullet}} \]
The proton is represented as an hydroxyl defect, where a proton is bound to a lattice oxygen.
A simplified mass-action expression is
\[ K_{\mathrm{hyd}}= \frac{[OH_O^{\bullet}]^2} {a_{\mathrm{H_2O}}[V_O^{\bullet\bullet}]} \]
with regular oxygen-site activity treated as approximately constant.
The reaction consumes oxygen vacancies and introduces protonic defects. Under conditions where hydration is extensive, the proton concentration can approach the vacancy-limited maximum, subject to charge neutrality, site availability and the total acceptor concentration.
The hydration equilibrium is temperature- and water-vapour-pressure-dependent. A high water-vapour pressure generally favours hydration in this simplified model, while increasing temperature often favours dehydration for exothermic hydration reactions. The actual thermodynamic response should be established for the material being studied.
8. Brouwer diagrams
A Brouwer diagram plots the logarithm of defect concentration against the logarithm of oxygen partial pressure. It is a graphical representation of the dominant defect regimes predicted by a chosen set of equilibria and constraints.
A straight-line segment appears when a particular set of assumptions produces a power-law relationship:
\[ [c_i]\propto p_{\mathrm{O_2}}^m \]
Taking logarithms gives
\[ \log[c_i]=m\log p_{\mathrm{O_2}}+C \]
where \(m\) is the slope and \(C\) is a constant over the regime considered.
Constructing a Brouwer diagram
- List plausible point defects and their effective charges.
- Write balanced defect reactions.
- Derive the corresponding mass-action expressions.
- Write charge neutrality and any required site-balance equations.
- Identify limiting regimes in which selected defects dominate.
- Solve the simplified equations in each regime.
- Calculate crossover conditions where two contributions become comparable.
- Plot the resulting concentrations and label the assumptions behind each segment.
A Brouwer diagram is therefore model-dependent. It is not a direct experimental plot unless defect concentrations have been independently measured.
Interpreting slopes
A slope can help identify a candidate defect regime, but only when the assumed defect charge states and compensation mechanisms are justified. Different defect models can produce similar slopes over a restricted pressure interval. Use complementary measurements such as thermogravimetry, conductivity, Seebeck coefficient, oxygen permeation or spectroscopic probes to distinguish them.
9. From defect concentration to ionic and electronic conductivity
The concentration of a defect does not by itself determine conductivity. Mobility, correlation, association and microstructure also matter.
For a carrier species \(i\) with charge number \(z_i\), concentration \(c_i\) and mobility \(\mu_i\), a simple drift expression is
\[ \sigma_i=|z_i|e c_i\mu_i \]
where \(e\) is the elementary charge. The total conductivity is the sum of contributions from mobile carriers under the assumptions of independent transport:
\[ \sigma_{\mathrm{total}}= \sigma_{\mathrm{ion}}+ \sigma_{\mathrm{electron}}+ \sigma_{\mathrm{hole}} \]
In real solids, correlation factors and coupled transport can modify this simple decomposition. For ionic conductors, the concentration of mobile vacancies may be smaller than the total structural vacancy concentration because some vacancies are trapped in defect associates or blocked by microstructural features.
For a polycrystalline ceramic, grain boundaries, pores, secondary phases and electrode interfaces can also affect the measured conductivity. A defect-chemistry model should therefore be connected to the measurement geometry and microstructure.
10. Defect association and non-ideal behaviour
At higher dopant or defect concentrations, the dilute, independent-defect approximation may fail. Defects can associate into electrically neutral or partially compensated complexes.
A schematic association reaction is
\[ \mathrm{A_M'+V_O^{\bullet\bullet} \rightleftharpoons (A_M'-V_O^{\bullet\bullet})^{\bullet}} \]
The complex shown has one net positive effective charge. Other association stoichiometries are possible.
Association changes the free-defect concentration and can alter migration barriers. It can also produce local structural distortions and defect ordering. In such cases, nominal dopant concentration is not a reliable proxy for mobile-vacancy concentration.
Evidence for association may come from deviations from ideal mass-action slopes, changes in activation energy, local-structure measurements, atomistic calculations or concentration-dependent transport. No single observation is usually decisive.
11. Worked reasoning: identifying a reduction regime
Suppose an oxide is measured over a range of oxygen partial pressures at fixed temperature. Its electronic conductivity increases as oxygen pressure decreases.
A defensible analysis proceeds as follows:
- Confirm that the gas composition has equilibrated and that the measured response is reversible.
- Determine whether the dominant electronic carrier is electron-like or hole-like using an appropriate method, such as Seebeck measurements or controlled electrochemical measurements.
- Measure oxygen nonstoichiometry, where possible, to test whether oxygen vacancies increase under reducing conditions.
- Compare the measured logarithmic slopes with predictions from candidate defect models.
- Check whether acceptor dopants, defect association or multiple vacancy charge states are important.
- Test whether the same model explains both conductivity and nonstoichiometry trends.
An increase in conductivity under reducing conditions is consistent with electron generation in many oxides, but it does not alone prove the oxygen-vacancy reduction model. Cation reduction, phase changes, surface conduction or changes in contact behaviour may also contribute.
12. Common modelling errors
- Unbalanced reactions: violate mass or effective-charge conservation.
- Confusing formal and effective charge: Kröger–Vink charges are relative to perfect lattice sites.
- Omitting charge neutrality: leaves defect concentrations underdetermined.
- Using concentrations as activities without stating assumptions: can fail at high defect concentrations.
- Treating nominal dopant content as free-vacancy content: ignores association and alternative compensation.
- Assuming one oxygen-vacancy charge state: multiple charge states may be relevant.
- Ignoring hydration: water vapour can change proton and vacancy populations.
- Extrapolating a Brouwer slope outside its regime: the dominant defects can change with oxygen pressure.
- Equating defect concentration with conductivity: mobility and microstructure also matter.
- Treating a fitted model as unique: several defect models may reproduce a limited dataset.
13. Minimum reporting checklist
A quantitative defect-chemistry analysis should report:
- Host composition, dopant content and assumed oxidation states.
- Kröger–Vink reactions and charge states used.
- Mass-action expressions and activity or concentration conventions.
- Charge-neutrality and site-balance equations.
- Assumed dominant defects and the pressure or temperature range of each regime.
- Equilibrium constants or thermodynamic parameters and their sources.
- Treatment of defect association and non-ideal behaviour.
- Oxygen and water-vapour partial pressures, temperature and equilibration protocol.
- Experimental evidence used to validate defect concentrations or carrier type.
- Uncertainty and limitations of the proposed model.
Further reading
- F. A. Kröger and H. J. Vink, “Relations between the concentrations of imperfections in crystalline solids,” in Solid State Physics, Vol. 3, Academic Press (1956), 307–435.
- A. R. West, Solid State Chemistry and its Applications, 2nd ed., Wiley (2014).
- J. Maier, Physical Chemistry of Ionic Materials: Ions and Electrons in Solids, Wiley (2004).
- R. Waser (ed.), Nanoelectronics and Information Technology: Advanced Electronic Materials and Novel Devices, 3rd ed., Wiley-VCH (2012), sections on defect chemistry and oxide materials.
- S. B. Adler, “Factors governing oxygen reduction in solid oxide fuel cell cathodes,” Chemical Reviews 104 (2004), 4791–4844. https://doi.org/10.1021/cr020724o.