The use of the label charged-defect

[miaozhuang68]

Dear Shankar, Thank you very much for your serious reply in the previous post. Now I can use most of the functions I need in JDFTx proficiently! What I am currently calculating is the surface potential of the metal Slab. I need to deduct a part of the charge to simulate the polarization of the electrode. There is no doubt that I should use the label elec-initial-charge here. But according to what you mentioned in issue245 the Defect page in jdftx.org, my understanding is that if I want to place a charged defect in a relatively small area, I should use both tags charged-defect and charged-defect-correction. I don't know if I should use these two labels in my situation, if it will make my calculations more accurate, or is it enough to use coulomb-interaction slab and elec-initial-charge? Looking forward to hearing from you, and thank you in advance for your answers!

Best, Jam

[shankar1729]

For charging up a metal slab in the electrochemical context, i.e. with countercharge coming from an electrolyte, just follow the charged surfaces tutorials. No need for charged defect correction there.

You only need charged defect correction when you have localized charges due to defects, and your unit cell has a net charge. In the metal surfaces case for electrochemistry, the unit cell overall is neutral because your electrolyte countercharge neutralizes the electrode charge.

[miaozhuang68]

Dear Shankar, thank you for your quick reply! We used the NELECT label in our previous calculations in the vacuum in VASP, and some reviewers suggested that VASP would introduce a uniform background charge to keep the system electrically neutral, When the uniform background charge is large, the existing periodic background charge will interact with the uniform background charge and seriously affect the calculation. I wonder if the implicit solvation model of JDFTx can handle the effect of uniform background charge well. Because intuitively, the implicit solvation model can introduce periodic background charge more stably. And I urgently need to introduce a surface with a large electric polarization. Best, Jam

[shankar1729]

Implicit solvation models with an electrolyte will produce the compensating charge physically in the interface region (where the double layer should be). As long as you use fluid-cation and fluid-anion to specify an ionic concentration, as shown in all the tutorials involving charged systems, there will be no uniform background charge to deal with. (This is the whole point of the electrolyte solvation / GC-DFT work.)

Best, Shankar

[moyulyy]

Hi Shankar,

The ion concentration set in some literatures is 0.5M and 1.0M. How should I set the ion concentration? In other words, how does the difference in ion concentration affect the energy calculation of GC-DFT.

Best, lyy

[shankar1729]

See the documentation of commands fluid-cation and fluid-anion: the second parameter is the ionic concentration in mol/liter. For implicit solvation models, this changes the Debye screening length and the capacitance near PZC for nonlinear models, which in turn affects the charge at a given electrode potential in GC-DFT. However, effects like changes in explicit ion adsorption due to concentration cannot be captured by an implicit model.

Best, Shankar