Thermodynamic integration

[mf081]

Hello,

is it possible in VASP to perform a thermodynamic integration to obtain the total free energy via dF/d lamda=dU /d lambda (in plasma physics sometimes called "Aufladeformel", I don't know the english word), where lambda is a parameter of the interaction potential strength? Lambda must be varied e.g. from zero (ideal gas as reference system) to full (normal QMD) interaction strength. Or spoken more descriptive: The charge of the electrons (or ions) is varied from zero to e (or Z*e). How can one change this interaction strength in VASP?

Thank you

[admin]

please read the chapter s Monopole, Dipole and Quadrupole interactions (+ ff) in the online manual. In principle, it is possible to obtain U of a charged system in VASP, but the choice of the systems that can be handled is restricted (therefore please check the manual).

[mf081]

Thank you,

I read the chapters but I am not sure I understood it right. Does that mean, by adding electrons, VASP automatically adds a compensating background charge which effectively weakens the potential of the ions?

Is this the only way to manipulate the charges in the DFT? Or is it also possible to change the strength of e.g. a hydrogen pseudopotential directly from Z=1e to ... Z=0e? Or do the pseudopotentials not allow such scaling manipulations?

[mf081]

Nobody likes to answer?