Hi, Bernd,
I think if there exists an oxygen activity gradient across the oxide, then the A and B must move, because the system satisfies the thermodynamic relationship Gibbs-Duhem Equation:
In which, you have chosen one matrix component and used the diffusion potential to replace Chemical potential in the equation of diffusion flux, so that make it independent, right?
So A and B should move, do you think so ?
However, I have questions about how MICRESS will calculate the initial quasi-equilibrium at the interface? As I have tried, the initial composition of different phase is totally wrong output on the screen, for example, there should be no A and B existing in the gas phase, but the initial composition for quasi-equilibrium calculation output on the screen include A and B (which is the same problem as average composition) , so do you have some good solution to this problem ??
regards,
tatalemon
activity boundary condition
Re: activity boundary condition
Hi tatalemon,
I think when considering the three sublattices as you defined them, A and B would not move. It is just the relation of B+ and B2+ which would change and which you did not include in the Gibbs-Duhem relation.
If you replace the chemical potential with the diffusion potential, i.e. if you neglect VA on the sublattices, then of course a movement of O2- would require A or B to move to the opposite direction. But I think this is an artifact of using the diffusion potential (which is what MICRESS does).
In reality there are also interactions between O and A,B. They could be included e.g. by cross diffusion terms if you had a mobility database or construct a simple one yourself. Then, the composition difference in O2- would drive A and B. But the problem of VA and choosing a matrix component is still there...
With "initial composition" you mean the "Start Composition for iteration of quasi-equilibrium"? This output is just reflecting the database definition, and the default major compositions. Thermo-Calc creates them if you use "SET_DEFAULT_MAJOR_COMPOSITIONS".
If in this output A and B are included in the gas phase, then it would be an error of the thermodynamic database...
Bernd
I think when considering the three sublattices as you defined them, A and B would not move. It is just the relation of B+ and B2+ which would change and which you did not include in the Gibbs-Duhem relation.
If you replace the chemical potential with the diffusion potential, i.e. if you neglect VA on the sublattices, then of course a movement of O2- would require A or B to move to the opposite direction. But I think this is an artifact of using the diffusion potential (which is what MICRESS does).
In reality there are also interactions between O and A,B. They could be included e.g. by cross diffusion terms if you had a mobility database or construct a simple one yourself. Then, the composition difference in O2- would drive A and B. But the problem of VA and choosing a matrix component is still there...
With "initial composition" you mean the "Start Composition for iteration of quasi-equilibrium"? This output is just reflecting the database definition, and the default major compositions. Thermo-Calc creates them if you use "SET_DEFAULT_MAJOR_COMPOSITIONS".
If in this output A and B are included in the gas phase, then it would be an error of the thermodynamic database...
Bernd