Change of basis vectors lead to different forces

[sciencemonk]

Hi vasp users,

I wonder if someone can help me to understand the behavior of my calculations for bulk anatase. I tried to find the internal structure and lattice parameters of anatase, complying with the procedure mentioned in the manual. However, when I obtained a converged atomic configuration, and change the basis vectors to a new equivalent set (while keeping the atomic positions), the result gives difference forces exerted on atoms and requires further relaxation.

To be specific, I used POSCAR as follows:

Code: Select all

anatase-bulk?????????????????????????????????????????????????????????????????????????????????????????????????????????????????
???1.00000000000000
?????3.8055302207018107????0.0000000000000000????0.0000000000000000
?????0.0000000000000000????3.8055302207018107????0.0000000000000000
?????1.9027651103509058????1.9027651103509058????4.8771429625523526
???2???4
Direct
??0.6250000000000000??0.8750000000000000??0.2500000000000000?
??0.3750000000000000??0.1250000000000000??0.7500000000000000
??0.8310745746220322??0.0810745746220322??0.8378508507559356
??0.1689254253779678??0.9189254253779678??0.1621491492440645
??0.4189254253779678??0.6689254253779678??0.6621491492440644
??0.5810745746220322??0.3310745746220322??0.3378508507559355

in which first two are Ti and the rest are O. The INCAR goes as

Code: Select all

SYSTEM?=?anatase-bulk?????????????????????????????????????????????????????????????????????????????????????

??ISTART?=????0????#?job???:?0-new??1-cont??2-samecut
??ICHARG?=????2????#?charge:?1-file?2-atom?10-const
??ISPIN??=????1????#?spin?polarized?calculation?
??ENCUT??=?520.
??NELM???=??100????#?maximum?number?of?electronic?SC?steps
??NELMIN?=????2????#?minimum?number?of?electronic?SC?steps
??EDIFF??=?1.0E-04?#?stopping-criterion?for?ELM

??EMIN???=?-10.0
??EMAX???=??10.0???#?energy-range?for?DOS?????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????
??NEDOS??=??2001???#?number?of?DOS?points
??ISMEAR?=???-5????#?how?the?partial?occupancies?are?set?
??SIGMA??=??0.01???#?broadening?in?eV?-4-tet?-1-fermi?0-gaus
??NGX????=???30
??NGY????=???30
??NGZ????=???42
??IALGO??=???48????#?algorithm,?38:?DAV;?48:?RMM-DIIS

??LWAVE??=????F????#?write?WAVECAR
??LCHARG?=????F????#?write?CHGCAR

??NSW????=????0????#?number?of?steps?for?IOM
??IBRION?=???-1????#?ionic?relax:?0-MD?1-quasi-New?2-CG
??ISYM???=????3????#?0-nonsym?1-usesym?2-fastsym

The resulting forces are converged as each atom has force less than 0.01 eV/Ang.

When I changed POSCAR as

Code: Select all

anatase-bulk
???1.00000000000000
?????1.9027651103509058????1.9027651103509058????4.8771429625523526
?????0.0000000000000000????3.8055302207018107????0.0000000000000000
????-3.8055302207018107????0.0000000000000000????0.0000000000000000
???2???4
Direct
?0.250000000000000??0.875000000000000??0.375000000000000
?0.750000000000000??0.125000000000000??0.625000000000000
?0.837850850755936??0.081074574622032??0.168925425377968
?0.162149149244065??0.918925425377968??0.831074574622032
?0.662149149244064??0.668925425377968??0.581074574622032
?0.337850850755936??0.331074574622032??0.418925425377968

i.e., essentially the same lattice shape and atomic positions, only different lattice vectors. The OUTCAR says with the new POSCAR, O atoms have forces as large as 0.06 eV/Ang and further relaxation is required.

Since vasp is periodic boundary condition based, same lattice shapes and atomic configurations should give same results, but why I got different? Could someone explain why?

Thanks a lot!

[support_vasp]

Hi,

We're sorry that we didn’t answer your question. This does not live up to the quality of support that we aim to provide. The team has since expanded. If we can still help with your problem, please ask again in a new post, linking to this one, and we will answer as quickly as possible.

Best wishes,

VASP