Dear experts,
In VASP, we can optimize the lattice (only xy direction) and atomic positions of a 2D material through modifying the constr_cell_relax.F file. However, it seems this trick does not work if we turn the SSNEB on to study the phase transition of a 2D material (the z direction would be optimized as well during the run). I wonder whether it is possible to use SSNEB method to study the 2D phase transition (without changing the lattice along z direction).
Best
Leiwy
SSNEB for 2D material
Moderator: moderators
Re: SSNEB for 2D material
Hi Leiwy,
I think that the best approach might be to run vasp through the ASE code. Vasp can be used as a calculator in ASE and the SSNEB is implemented there as well. ASE has a great deal of flexibility in terms of constraints that be applied for the calculations.
It would also not be difficult to modify the SSNEB code in vasp for 2D materials. This would likely be very similar to the modification in constr_cell_relax.F. One could either zero any change of the cell in z, or zero the stress component along z. I think that a '2D' tag could be used to turn this option on. Let me know if you want to proceed along these lines, and if you need any help - again, it should not be hard.
Graeme
I think that the best approach might be to run vasp through the ASE code. Vasp can be used as a calculator in ASE and the SSNEB is implemented there as well. ASE has a great deal of flexibility in terms of constraints that be applied for the calculations.
It would also not be difficult to modify the SSNEB code in vasp for 2D materials. This would likely be very similar to the modification in constr_cell_relax.F. One could either zero any change of the cell in z, or zero the stress component along z. I think that a '2D' tag could be used to turn this option on. Let me know if you want to proceed along these lines, and if you need any help - again, it should not be hard.
Graeme
Re: SSNEB for 2D material
Dear Graeme,
Thanks a lot for your reply. I will look into ASE for a solution following your first suggestion. As for the second one, it seems using NEB method provided by VASP itself works for 2D material if the constr_cell_relax.F is modified. But I'm also interested in if we can realize it by using ssneb provided by VTST code (LNEBCELL=T). Please shed some hints on how to do that, and I believe this can help a lot of researchers working on 2D phase transitions.
Best regards
Leiwy
Thanks a lot for your reply. I will look into ASE for a solution following your first suggestion. As for the second one, it seems using NEB method provided by VASP itself works for 2D material if the constr_cell_relax.F is modified. But I'm also interested in if we can realize it by using ssneb provided by VTST code (LNEBCELL=T). Please shed some hints on how to do that, and I believe this can help a lot of researchers working on 2D phase transitions.
Best regards
Leiwy
Re: SSNEB for 2D material
Dear Graeme,
By the way, I found a twodim_flag hiding in chain.F which set the z component of stress to zero. So this function has already been realized in VTST code?
Best
Leiwy
By the way, I found a twodim_flag hiding in chain.F which set the z component of stress to zero. So this function has already been realized in VTST code?
Best
Leiwy
Re: SSNEB for 2D material
Wow, yes, it is already implemented! This has to be thanks to Penghao Xiao.
Re: SSNEB for 2D material
Dear Graeme,
Thanks a lot for the efforts from you and your colleagues. We appreciate it so much!
Best wishes
Leiwy
Thanks a lot for the efforts from you and your colleagues. We appreciate it so much!
Best wishes
Leiwy