Thermodynamic Factors in Ionic Transport: Chemical Potential, Activity, and Chemical Diffusion in Solid Oxides
Thermodynamic Factors in Ionic Transport: Chemical Potential, Activity, and Chemical Diffusion in Solid Oxides
Ionic transport in a solid is driven by gradients in electrochemical potential, not by concentration gradients alone. Concentration is often used as the practical variable in transport equations, but this substitution is exact only for an ideal solution or when the non-ideality is explicitly included.
The thermodynamic factor quantifies how strongly chemical potential changes when the concentration of a mobile species changes. It connects defect thermodynamics to diffusion measurements, chemical relaxation, oxygen permeation, and the interpretation of transport in concentrated solid electrolytes.
This article derives the thermodynamic factor, distinguishes tracer diffusion from chemical diffusion, explains its relationship to ionic conductivity, and develops examples relevant to oxygen-ion and proton-conducting oxides.
1. Why concentration gradients are not the fundamental driving force
For a species (i) with charge number (z_i), the electrochemical potential is
[ _i = _i+z_iF ]
where (_i) is the chemical potential, (F) is Faraday’s constant, and () is the electric potential. The corresponding flux in a linear-response description is
[ _i = - _i ]
for a molar concentration (c_i), with (D_i) representing the relevant mobility-related diffusivity under the assumptions of this expression.
The chemical potential can be written in terms of activity:
[ _i=_i^+RTa_i ]
and therefore
[ _i=RTa_i ]
A concentration gradient produces a chemical-potential gradient through the activity–concentration relationship. If activity is proportional to concentration, the familiar Fickian expression follows. If the material is non-ideal, the activity coefficient changes with concentration and modifies the effective diffusion response.
In ionic solids, the electric-potential term also matters. Charged defects generally cannot diffuse independently over macroscopic distances without satisfying local charge neutrality or developing an internal electric field.
2. Definition of the thermodynamic factor
For a single mobile species, define the thermodynamic factor as
[ ]
Using (_i=_i^+RTa_i),
[ ]
The derivative is evaluated at constant temperature and pressure, with the composition constraints stated for the system. In a multicomponent solid, the thermodynamic response is generally a matrix of derivatives rather than one scalar factor.
The thermodynamic factor is dimensionless. It measures the logarithmic change in activity associated with a logarithmic change in concentration.
For an ideal dilute species, (a_ic_i), so
[ _i=1 ]
For a non-ideal solution, (_i) can differ from unity. Its value and concentration dependence reflect interactions, defect association, site exclusion, ordering, and other non-ideal contributions.
3. Activity coefficients and non-ideal defect populations
Write the activity as
[ a_i=_i ]
where (_i) is the activity coefficient and (c^) is the chosen standard concentration. Then
[ a_i = _i+() ]
Differentiating with respect to (c_i),
[ ]
This form separates the ideal contribution, equal to one, from the non-ideal contribution.
In an oxide, defect activities need not equal their site fractions. At low defect concentrations, an ideal dilute approximation may be useful. At higher concentrations, electrostatic interactions, defect association, elastic interactions, and site-fraction constraints can cause substantial deviations.
For example, an oxygen vacancy may associate with an acceptor dopant:
[ ]
The precise effective charge and association stoichiometry depend on the defect complex. Association changes the fraction of vacancies that are mobile and changes the relationship between total vacancy concentration and vacancy activity.
4. From chemical-potential gradients to a chemical diffusion coefficient
Consider an isothermal, one-dimensional system in which the electric-field and coupled-species constraints have either been accounted for or are absent from the reduced description. The flux can be written
[ J_i = - ]
Applying the definition of (_i),
[ = ]
Substitution gives
[ J_i=-D_i_i ]
Comparing with Fick’s first law,
[ J_i=-D_{} ]
yields, for this single-component reduced model,
[ ]
This relation is useful conceptually, but its interpretation requires care. The diffusivity (D_i) must correspond to the mobility or tracer-like quantity used in the flux law, and the system must satisfy the assumptions used to eliminate coupled transport and electric-field effects. In a real ionic solid, the measured chemical diffusion coefficient can also contain ambipolar, correlation, and defect-reaction contributions.
5. Thermodynamic factor in a lattice with finite site occupancy
A simple lattice model illustrates why the ideal dilute approximation can fail. Consider mobile defects occupying a fraction () of equivalent sites, with vacant host sites occupying fraction (1-). For an ideal lattice gas,
[ a= ]
apart from the chosen standard-state convention. Since concentration is proportional to (),
[ a=-(1-) ]
and therefore
[ ]
At low occupancy, (), the factor approaches unity. As the occupied fraction increases, the thermodynamic factor rises because adding another defect becomes progressively less favourable due to site exclusion.
This is an ideal lattice model, not a universal expression for oxide defects. Real materials may additionally exhibit defect interactions, ordering, multiple sublattices, and association.
6. Thermodynamic factor and ionic conductivity
For independent charge carriers in the Nernst–Einstein approximation, the conductivity is
[ ]
where (D_{,i}) is the diffusivity appropriate to collective charge transport. In a strictly uncorrelated ideal model it can be related to the tracer diffusivity. In real solids, correlated jumps and defect interactions can make these quantities different.
The thermodynamic factor does not simply multiply the measured conductivity in this equation. Conductivity describes response to an electrochemical-potential gradient, whereas the thermodynamic factor converts a chemical-potential gradient into a concentration gradient. It therefore enters the relation between conductivity-related mobility and chemical diffusion.
Combining the Nernst–Einstein expression with the reduced single-species relation gives
[ D_{} = _i ]
provided the same carrier population and compatible diffusivity definitions are used, and coupled-species effects are negligible or have been incorporated consistently.
A mismatch between a diffusivity inferred from conductivity and one measured by tracer exchange does not automatically indicate an incorrect thermodynamic factor. Correlation effects, carrier concentration estimates, grain boundaries, and microstructural limitations must also be considered.
7. Tracer diffusion, self-diffusion, and chemical diffusion
These diffusivities answer different experimental questions:
- Tracer diffusion describes the spreading of an isotopically labelled species without requiring a macroscopic chemical-composition gradient.
- Self-diffusion describes equilibrium random motion of identical particles and is often measured using isotope tracers.
- Chemical diffusion describes relaxation of a chemical-composition gradient and includes the thermodynamic response of the material.
- Charge diffusivity is the collective transport quantity associated with electrical current and can differ from tracer diffusivity because of correlations.
In a simple single-species model, chemical diffusion is the product of a mobility-related diffusivity and the thermodynamic factor. In a multicomponent ionic solid, the relation can involve a matrix of kinetic coefficients and thermodynamic derivatives.
Consequently, a chemical diffusion coefficient measured by oxygen-isotope exchange, conductivity relaxation, or a permeation experiment should not be compared directly with a tracer diffusivity without identifying the relevant thermodynamic and kinetic assumptions.
8. Multicomponent thermodynamics: a matrix of factors
For several independent composition variables, the chemical-potential response is described by a matrix:
[ {ij} = ( ){T,P} ]
The matrix () is related to the curvature of the free energy with respect to composition. Its off-diagonal elements describe how changing one component affects the chemical potential of another.
For a multicomponent solid, the fluxes can be represented schematically as
[ = - ]
where () is a kinetic-coefficient matrix and () is the vector of electrochemical potentials.
The concentration-based diffusion response therefore depends on both the kinetic matrix and the thermodynamic derivative matrix. This is why a single scalar thermodynamic factor is insufficient for many mixed ionic–electronic conductors and multidefect oxides.
9. Application to oxygen nonstoichiometry in perovskite oxides
For a perovskite written as
[ ]
the oxygen deficiency () is related to the oxygen-vacancy population. A change in oxygen chemical potential changes the equilibrium value of (), and the curvature of the free energy with respect to () controls the thermodynamic response.
For oxygen exchange with the gas phase, a common reduction reaction is
[ ]
The equilibrium oxygen content depends on temperature, oxygen partial pressure, electronic compensation, and defect interactions.
A useful scalar thermodynamic factor can be defined with respect to oxygen nonstoichiometry, subject to a clearly specified convention. For example,
[ _ = ]
where (_{}) must be defined consistently with the chosen reduction or oxygen-exchange reaction. This definition is not interchangeable with the single-defect (_i) unless the reaction stoichiometry and composition variable are mapped explicitly.
The slope of an equilibrium () versus (p_{}) is therefore a thermodynamic observable. It can help identify defect regimes, but it is not by itself a direct measurement of an oxygen tracer diffusivity.
10. Chemical diffusion and oxygen chemical-potential gradients
Chemical diffusion is particularly important in mixed ionic–electronic conducting oxides. A spatial variation in oxygen content creates a gradient in oxygen chemical potential. Oxygen exchange at the surface and transport through the bulk then act together to relax the gradient.
For a simplified chemical-diffusion description,
[ = ( D_{}c ) ]
If (D_{}) depends on composition, it must remain inside the divergence operator. Replacing it by a constant is justified only over a sufficiently narrow composition range or when its variation is negligible.
For oxygen nonstoichiometry, the concentration variable may be represented by (), oxygen-site fraction, or oxygen content. The numerical value and units of the reported chemical diffusion coefficient depend on that choice and on the normalization convention.
11. Example: interpreting a non-ideal thermodynamic factor
Suppose an experimentally or computationally determined activity coefficient follows the illustrative relation
[ =Ac ]
where (A) is a constant over the concentration range considered. Since
[ a=()+Ac ]
the thermodynamic factor becomes
[ = = 1+Ac ]
If (Ac=0.30), then
[ ]
Under the reduced single-species assumptions, the chemical diffusion coefficient is 30% greater than the mobility-related diffusivity used in the flux expression:
[ D_{}=1.30D_i ]
This numerical example illustrates the thermodynamic correction only. It is not a measured value for a particular oxide, and the assumed activity-coefficient model should not be extrapolated beyond the concentration range in which it is valid.
12. Thermodynamic factor and defect association
Defect association can make the total defect concentration a poor proxy for the concentration of mobile carriers. Consider a simplified association equilibrium:
[ ]
The total vacancy content includes both free and associated vacancies:
[ [V_O]{} = [V_O]{} + [(AV_O)] ]
If only free vacancies contribute efficiently to long-range transport, the conductivity depends more directly on the free-vacancy population than on the total oxygen deficiency.
The free-defect activity can vary nonlinearly with total defect content as the association equilibrium shifts. This contributes to non-ideal thermodynamic behaviour and can also alter the kinetic coefficients through changes in the mobile fraction.
Thermodynamic non-ideality and kinetic trapping are related but distinct: the thermodynamic factor describes the free-energy response, while the mobility and correlation terms describe how defects move.
13. Temperature dependence
The thermodynamic factor is generally temperature dependent because both the free-energy curvature and defect populations change with temperature.
For an activity coefficient that depends on temperature and composition,
[ (T,c) = 1+ ( )_T ]
Thus, even if the functional form of () is unchanged, the factor can vary because the interaction parameters vary with temperature.
When chemical diffusion is measured at several temperatures, its activation energy may therefore contain contributions from both defect mobility and thermodynamic response. Treating the chemical diffusion coefficient as a pure migration coefficient can lead to an incorrect interpretation of the activation energy.
14. How the thermodynamic factor is determined
Several complementary approaches are available:
- Activity or chemical-potential measurements: determine the relationship between chemical potential and composition, then evaluate the logarithmic derivative.
- Equilibrium nonstoichiometry measurements: measure () as a function of oxygen partial pressure and temperature, then derive the chemical-potential response using the reaction stoichiometry.
- Calorimetry and thermodynamic modelling: constrain enthalpic and entropic contributions to the free energy.
- First-principles calculations: calculate defect energetics and, with suitable statistical-mechanical treatment, composition-dependent free energies.
- Combined transport analysis: compare chemical diffusion, tracer diffusion, and conductivity using a consistent carrier concentration and correlation model.
Numerical differentiation of experimental data can amplify noise. A physically justified fit to the equilibrium composition curve is often preferable to taking point-by-point finite differences.
15. Practical interpretation for proton-conducting oxides
In acceptor-doped proton conductors, hydration changes the concentrations of oxygen vacancies and protonic defects. The proton concentration is controlled by the hydration equilibrium, while proton mobility depends on local bonding, lattice dynamics, and migration barriers.
A concentration-dependent proton activity means that the chemical-potential response need not be ideal. If proton–proton interactions, dopant–proton association, or site exclusion become important, the thermodynamic factor can differ from unity.
This distinction matters when comparing:
- Proton conductivity with proton tracer or self-diffusion.
- Water uptake with proton concentration inferred from electrical measurements.
- Chemical relaxation coefficients with microscopic migration coefficients.
- Samples with different acceptor concentrations or hydration histories.
A conductivity increase can arise from a larger carrier concentration, a higher mobility, or both. Thermodynamic-factor analysis helps separate the equilibrium contribution from the kinetic contribution, but it does not replace independent evidence about the mobile defect population.
16. Common interpretation errors
- Assuming () universally: this is an ideal-solution approximation, not a general property of solid electrolytes.
- Equating chemical and tracer diffusion: they differ through thermodynamic and correlation effects.
- Multiplying conductivity by () without a derivation: the thermodynamic factor modifies the conversion between chemical-potential and concentration gradients; it is not a universal conductivity multiplier.
- Ignoring electroneutrality: charged defects in a solid generally have coupled transport and internal electric fields.
- Using total defect concentration as mobile-carrier concentration: associated or trapped defects may contribute little to long-range transport.
- Mixing concentration conventions: site fraction, molar concentration, and nonstoichiometry parameter are not interchangeable without conversion.
- Differentiating noisy equilibrium data directly: the derivative can be much less reliable than the original measurements.
17. Connecting thermodynamics, kinetics, and measured transport
A useful conceptual separation is:
- Thermodynamics: determines equilibrium defect populations and the response of chemical potential to composition.
- Kinetics: determines migration barriers, jump frequencies, correlation, and mobility.
- Microstructure: determines grain-boundary resistance, porosity, tortuosity, and interfacial contributions.
- Electrochemical constraints: determine how charged species and electronic carriers move together while maintaining charge balance.
A rigorous transport interpretation therefore combines equilibrium defect chemistry, thermodynamic factors, kinetic coefficients, and microstructural characterization. No single conductivity or diffusion measurement usually identifies all of these contributions independently.
18. Key equations
The thermodynamic factor is
[ ]
For an activity-coefficient representation,
[ ]
For a reduced single-species diffusion model,
[ ]
For independent charge carriers in the Nernst–Einstein approximation,
[ ]
These equations connect activity, chemical potential, diffusion, and electrical transport. Their application requires consistent definitions of concentration, diffusivity, charge state, and the constraints imposed by the solid’s defect chemistry.
19. Further reading
This article is part of the thermodynamics and transport resources on this website:
- Chemical Potential and Activity in Solid Materials
- Thermodynamic Equilibrium and the Mass Action Law in Solids
- Defect Formation Energy and Equilibrium Concentrations in Oxides
- Oxygen Chemical Potential and (pO_2)-Dependent Defect Chemistry
- Hydration Thermodynamics in Proton-Conducting Oxides
- Ionic Transport in Solid Oxides
- Oxygen Nonstoichiometry Measurement
- Transport Number: Measurement and Calculation
20. Final perspective
The thermodynamic factor is the bridge between a material’s equilibrium free-energy landscape and the concentration-based description of transport. It explains why chemical diffusion can differ from tracer diffusion and why non-ideal defect populations matter when interpreting conductivity and relaxation experiments.
For functional oxides, the most reliable transport analysis combines thermodynamic information with kinetic measurements and a defect model appropriate to the composition and operating atmosphere. This is particularly important in mixed ionic–electronic conductors and proton-conducting ceramics, where defect populations and transport constraints are strongly coupled.