This example requires VTST tools version XXX or higher, download here.
Unit cell parameters were taken from the phonopy example folder. Silicon unit cell contains 8 atoms, thus 24 frequencies (phonons), three acoustic close to zero.
For this example vasp_raman.py script version [XXX] (https://raw.github.com/raman-sc/VASP/3cb3cdf0682609365c4b966472ef6eb5be1defc5/vasp_raman.py) was used.
Raman intensities for all modes will be computed in one go. For the multi-step procedure see Cyclopentadiene example.
Working directory should contain the following files (FAT file system has different soft linking ways):
INCAR - should contain LEPSILON=.TRUE. or LCALCEPS=.TRUE. because we want 'MACROSCOPIC STATIC DIELECTRIC TENSOR' in the OUTCAR
KPOINTS - just kpoints (soft link of the KPOINTS file from the PHONON folder, to prevent errors)
freq.dat - contains frequencies (soft link of the freq.dat file from the PHONON folder)
modes_sqrt_amu.dat - contains eigenvectors/sqrt(mass in amu) (soft link of the modes_sqrt_amu.dat file from the PHONON folder)
POSCAR.phon - VASP4 or VASP5 format is supported, at this point only positive scales are supported (soft link of the POSCAR file from the PHONON folder)
POTCAR - `PAW_PBE Si 05Jan2001` PP (soft link to the POTCAR file from the PHONON folder)
raman.sub - shell script for the scheduler to *do the deed*
- Run
dymselsph.plin thePHONONfolder to generateDISPLACECAR. Need to make sure that all atoms in the unit cell are found, in this case8:
perl ~/vtstscripts/dymselsph.pl POSCAR 8 3.87 0.001
----------------------------------------------------------------------
Using 1 central atom
Central Coordinates: 0.6250000000000000 0.6250000000000000 0.1250000000000000
Central atom 1: 0.6250000000000000 0.6250000000000000 0.1250000000000000
----------------------------------------------------------------------
8 atoms were found within a radius of 3.87 of atom 8,
leading to 24 degrees of freedom selected.
----------------------------------------------------------------------Be sure to check the INCAR file for the correct vtst settings (http://theory.cm.utexas.edu/vtsttools/dynmat/):
! phonons
ICHAIN = 1 ! Run the dynamical matrix code
! IMAGES M ! Number of parallel images, if desired as in step 2 above; otherwise, do not add.
NSW = 25 ! (DOF/M)+1 Number of ionic steps
IBRION = 3 ! Tell VASP to run dynamics,
POTIM = 0.0 ! with a time step of zero (ie, do nothing)
ISYM = 0 ! Make sure that symmetry is off,
EDIFFG = -1E-10 ! and that vasp does not quit due to low forces
Once phonons are obtained, we need to generate freq.dat and modes_sqrt_amu.dat, that are equivalent to the information obtained from the OUTCAR in the the case of phonons calculation using VASP:
python ~/vtstscripts/dymmatrix.py DISPLACECAR OUTCAR
Reading DISPLACECAR
Number of displacements: 24
Reading OUTCAR
Building dynamical matrix
Diagonalizing matrix
0.219483 cm^{-1} ... 1
0.023661 cm^{-1} ... 1
0.005978 cm^{-1} ... 1
145.585614 cm^{-1} ... 0
145.750931 cm^{-1} ... 0
145.783401 cm^{-1} ... 0
145.804204 cm^{-1} ... 0
145.823087 cm^{-1} ... 0
145.832016 cm^{-1} ... 0
402.135683 cm^{-1} ... 0
402.137959 cm^{-1} ... 0
402.145976 cm^{-1} ... 0
402.146508 cm^{-1} ... 0
402.150628 cm^{-1} ... 0
402.154078 cm^{-1} ... 0
447.963957 cm^{-1} ... 0
448.030348 cm^{-1} ... 0
448.143871 cm^{-1} ... 0
448.245366 cm^{-1} ... 0
448.402489 cm^{-1} ... 0
448.497029 cm^{-1} ... 0
504.414547 cm^{-1} ... 0
504.447782 cm^{-1} ... 0
504.483568 cm^{-1} ... 0
Phonons look good, similar to those obtained from VASP (Sibulk-VASP example).
Now we are ready to compute Raman intensities.
Contents of the raman.sub, note that modes are giving in the increasing order, contrary to VASP:
#!/bin/bash
#PBS -A ONRDC17403171
#PBS -l select=5:ncpus=32:mpiprocs=32
#PBS -l walltime=01:00:00
#PBS -q debug
#PBS -j oe
#PBS -N Si_bulk-Raman-VTST
#PBS -V
cd $PBS_O_WORKDIR
ulimit -s unlimited # remove limit on stack size
export VASP_RAMAN_RUN='aprun -B /u/afonari/vasp.5.3.2/vasp.5.3/vasp &> job.out'
export VASP_RAMAN_PARAMS='04_24_2_0.01'
python /u/afonari/vasp_raman.py > vasp_raman.out
Submit all calculations to the scheduler:
qsub raman.sub
Similar to the Sibulk-VASP example...
After the job finished, a lot of OUTCAR files have been created (strictly speaking 42... the answer??!), vasp_raman.dat will contain Raman activities for the future processing (for example with gnuplot).
From experiment, Raman spectrum of the bulk silicon contains only one intense peak at around 520 cm-1 (J.H. Parker, et al., Phys Rev, 155, 712 (1967)), vasp_raman.dat contains:
# mode freq(cm-1) alpha beta2 activity
004 145.58561 -0.0001635 0.0000028 0.0000206
005 145.75093 -0.0000817 0.0000009 0.0000064
006 145.78340 -0.0000000 0.0000008 0.0000054
007 145.80420 0.0002044 0.0000003 0.0000039
008 145.82309 0.0000409 0.0000002 0.0000018
009 145.83202 0.0002861 0.0000011 0.0000114
010 402.13568 -0.0016350 0.0015959 0.0112915
011 402.13796 -0.0105863 0.0002777 0.0069868
012 402.14598 -0.0084200 0.0000903 0.0038223
013 402.14651 -0.0000817 0.0000081 0.0000572
014 402.15063 0.0021254 0.0011882 0.0085204
015 402.15408 -0.0058450 0.0007709 0.0069335
016 447.96396 -0.0002861 0.0001196 0.0008405
017 448.03035 -0.0018802 0.0008435 0.0060639
018 448.14387 -0.0018393 0.0039953 0.0281196
019 448.24537 -0.0024116 0.0011679 0.0084368
020 448.40249 0.0025751 0.0001548 0.0013818
021 448.49703 -0.0000817 0.0002679 0.0018757
022 504.41455 -0.0006540 780.5241570 5463.6691183
023 504.44778 -0.0008175 780.5880251 5464.1162060
024 504.48357 -0.0215814 779.9106481 5459.3954955
Triple degenerate mode at 504 cm-1 has the largest intensity.
Shannon Stauffer (UT Austin, PI: G. Henkelman): Email.
Alexandr Fonari (Georgia Tech, PIs: J.-L. Bredas/V. Coropceanu): Email