phonon frequencies outcar shows a different trend

Message

satish2414 · #1 Post by **satish2414** » Sat Jan 28, 2012 8:58 am

I have optimized the structure of TiO2 with PREC = accurate and NSW = 1, NFREE=2 POTIM=0.015,IBRION=5 in the INCAR file. But the output file was like the following:

13 f = 3.187285 THz 20.026299 2PiTHz 106.316364 cm-1 13.181554 meV
X Y Z dx dy dz
0.000000 0.000000 7.766956 0.000000 0.082216 0.000006
1.888000 0.000000 5.401368 0.000000 -0.082321 0.000001
1.888000 1.888000 4.863536 0.000000 -0.078900 -0.000025
1.888000 0.000000 7.343037 0.000000 0.069110 0.000006
0.000000 1.888000 8.304788 0.000000 0.078722 -0.000009
0.000000 0.000000 5.825287 0.000000 -0.069241 0.000008

14 f = 0.007498 THz 0.047109 2PiTHz 0.250096 cm-1 0.031008 meV
X Y Z dx dy dz
0.000000 0.000000 7.766956 0.000123 -0.079234 0.001442
1.888000 0.000000 5.401368 0.000123 -0.079090 0.001431
1.888000 1.888000 4.863536 0.000123 -0.078855 0.001414
1.888000 0.000000 7.343037 0.000123 -0.079166 0.001429
0.000000 1.888000 8.304788 0.000123 -0.078958 0.001459
0.000000 0.000000 5.825287 0.000123 -0.079081 0.001440

15 f/i= 0.006191 THz 0.038898 2PiTHz 0.206501 cm-1 0.025603 meV
X Y Z dx dy dz
0.000000 0.000000 7.766956 0.000965 0.001440 0.079089
1.888000 0.000000 5.401368 0.000964 0.001437 0.079104
1.888000 1.888000 4.863536 0.000965 0.001433 0.079196
1.888000 0.000000 7.343037 0.000966 0.001438 0.079028
0.000000 1.888000 8.304788 0.000966 0.001435 0.079102
0.000000 0.000000 5.825287 0.000966 0.001437 0.079070

16 f/i= 0.009040 THz 0.056799 2PiTHz 0.301539 cm-1 0.037386 meV
X Y Z dx dy dz
0.000000 0.000000 7.766956 0.079072 0.000106 -0.000971
1.888000 0.000000 5.401368 0.079053 0.000105 -0.000964
1.888000 1.888000 4.863536 0.079145 0.000105 -0.000964
1.888000 0.000000 7.343037 0.079196 0.000106 -0.000962
0.000000 1.888000 8.304788 0.079211 0.000105 -0.000972
0.000000 0.000000 5.825287 0.079175 0.000105 -0.000969

17 f/i= 7.680155 THz 48.255835 2PiTHz 256.182374 cm-1 31.762578 meV
X Y Z dx dy dz
0.000000 0.000000 7.766956 0.000000 0.059460 0.000000
1.888000 0.000000 5.401368 0.000000 0.059570 0.000000
1.888000 1.888000 4.863536 0.000000 -0.137857 -0.000004
1.888000 0.000000 7.343037 0.000000 -0.040715 -0.000001
0.000000 1.888000 8.304788 0.000000 -0.137741 0.000004
0.000000 0.000000 5.825287 0.000000 -0.040813 0.000002

18 f/i= 9.488318 THz 59.616863 2PiTHz 316.496226 cm-1 39.240545 meV
X Y Z dx dy dz
0.000000 0.000000 7.766956 0.000000 -0.059900 0.000001
1.888000 0.000000 5.401368 0.000000 0.059841 0.000001
1.888000 1.888000 4.863536 0.000000 -0.123805 -0.000002
1.888000 0.000000 7.343037 0.000000 0.071937 0.000000
0.000000 1.888000 8.304788 0.000000 0.124008 -0.000004
0.000000 0.000000 5.825287 0.000000 -0.071877 0.000000

As far as I know, the frequencies should decrease monotonously but its not the case here as the last few frequencies have increased.

Anybody knows why this happens or am I wrong in the concept?

Also the NSW=1 again does several ionic steps (> 1). What does this mean? according to the nsw tag it should be only 1 ionic step (or frozen calculaions) . So wh does this happen.

Thank you

vistawanted · #2 Post by **vistawanted** » Tue Jan 31, 2012 5:27 pm

13 f = 3.187285 THz
14 f = 0.007498 THz
15 f/i= 0.006191 THz
16 f/i= 0.009040 THz
17 f/i= 7.680155 THz
18 f/i= 9.488318 THz
Sorry, I do not know about VASP, so I cannot guarantee you my conjecture is right. But anyway, by comparing the results you shown, I may guess that the "f/i" means imaginary frequency.
Which q-point did you calculate? I tried Gamma point for anatase-TiO2 to obtain the IR and Raman spectrum by Quantum-ESPRESSO, but result was not as good as I expected (may because I did not generate the good pseudo-potential for Ti):
*************************************************************
Mode symmetry, D_4h(4/mmm) point group:

omega( 1 - 2) = -84.1 [cm-1] --> E_u X_5' M_5' I
omega( 3 - 3) = 23.1 [cm-1] --> A_2u X_4' M_4' I
omega( 4 - 5) = 152.9 [cm-1] --> E_g X_5 M_5 R
omega( 6 - 7) = 159.2 [cm-1] --> E_g X_5 M_5 R
omega( 8 - 9) = 231.5 [cm-1] --> E_u X_5' M_5' I
omega( 10 - 10) = 328.3 [cm-1] --> A_2u X_4' M_4' I
omega( 11 - 11) = 377.2 [cm-1] --> B_1g X_2 M_2 R
omega( 12 - 13) = 439.8 [cm-1] --> E_u X_5' M_5' I
omega( 14 - 14) = 490.1 [cm-1] --> B_1g X_2 M_2 R
omega( 15 - 15) = 512.5 [cm-1] --> A_1g X_1 M_1 R
omega( 16 - 16) = 535.4 [cm-1] --> B_2u X_3' M_3'
omega( 17 - 18) = 631.2 [cm-1] --> E_g X_5 M_5 R

satish2414 · #3 Post by **satish2414** » Mon Feb 06, 2012 7:05 pm

I did not understand how you assigned the E_u,A_2u etc.. Can I get some explanation on that please? I have 18 phonon frequencies in my OUTCAR for 6 atoms, how can I assign these parameters to the values.

I do understand that q points refer to something related to brillouin zone but I am not sure about its details. It will really help me if u could comment on the same. And my KPOINTS file was 8x8x1 k-mesh. So I would appreciate if I can get some information on finding the qpoints or assigning the same...

Thank you. Simple language may be easy to understand.

vistawanted · #4 Post by **vistawanted** » Wed Feb 08, 2012 6:05 pm

[quote="satish2414"]
I have 18 phonon frequencies in my OUTCAR for 6 atoms, how can I assign these parameters to the values.
[/quote]
And my KPOINTS file was 8x8x1 k-mesh. So I would appreciate if I can get some information on finding the qpoints or assigning the same...
[/quote]</span>