[QE-users] Phonon: negative frequencies for adsorbate molecule using "nat-todo" option
Omer Mutasim
omermutasim at ymail.com
Wed Nov 18 16:51:42 CET 2020
Dear Dr. Tamas
I tried to use ASE base on your earlier suggestion. I want to calculate the vibration frequencies for slab with SO2 adsorbate , i want to vibrate only the adsorbate, while fixing the surface using finite difference method. I used the following script, similar to the examples but it doesn't work, i got the following error :
File "D:\Anaconda\lib\site-packages\ase\formula.py", line 402, in parse2 raise ValueError
ValueError
Please tell me what is wrong with this script shown below. or please can you write the correct script for this calculation
import ase.iofrom ase import Atomsfrom ase.io import read, writefrom ase.calculators.emt import EMTfrom ase.optimize import BFGSfrom ase.vibrations import Vibrationsslab = read("CuO_slab.cif")slab = Atoms('slab', calculator=EMT())BFGS(slab).run(fmax=0.01)vib = Vibrations(slab)vib.run()vib.summary()
what is the right procedure for reading cif file of the optimized structure
On Thursday, November 5, 2020, 03:57:18 PM GMT+4, Tamas Karpati <tkarpati at gmail.com> wrote:
Dear Omer,
Yes, i meant SO2 gas phase sim. This is an alternative to using the physisorbed
slab+SO2 complex as "reactant", R. Question of methodology and the nature of
materials. I cannot recall whether S+2O were together (as SO2) or decomposed in
your starting structure but in the second case you may use this approach.
Also, kinetic parameters are probably easier to derive for the gas
phase reactant (?)
If your SO2 is in a single piece on the slab in R, then you can drop this idea.
With XCrySDen, to visualize vibrations, you need to use dynmat.x and specify
filxsf, then open *.xsf with XCrySDen. Select to visualize forces and use the
animation controls to choose which normal mode is to be shown.
Expect arrows, not animated vibrations. Such arrows indicate the atomic
replacements to do in order to get 1 negative freq. better: move them
where the arrows
point to, or the exact opposite direction. One by one you can find a
real minimum str.
Note that unless you do an all-atom phonon, you can never trust those arrows
(neither direction nor frequency, even sign).
As for H+O (if distant enough, yet in the same simbox), you formally expect
the same as for the HO radical -a single bond with a single freq but since
this is not a real bond (strength is negligible), freq will be very
small (and yes, S=0).
t
On Thu, Nov 5, 2020 at 5:15 AM Omer Mutasim <omermutasim at ymail.com> wrote:
>
> Dear Dr. Tamas
>
> Sorry , what do you mean by “ you most probably need an SO2 simulation (optimization+phonons)
> rather than the same for a surface attached SO2 or SO+O. Big difference! “ do you mean i should do phonon for SO2 in gas phase ?
>
> I do agree with you , 3-atoms phonon is non-physical, i will include the top layer also.
>
> Regarding Xcrysden , I don’t find any axsf file in my output files , how do you visualise it ?
> Another question:
> For the reaction : H+O = OH , if i did phonon for the initial state( reactant ) , I should expect to get no vibrational modes at all , right ? I will get only 6 translational modes . So the vibrational entropy will be zero, Please correct me if i am wrong.
>
> By the way , i tried ASE for rate constants, as you recommended, it is really helpful.
>
> Thanks a lot for you unwavering help.
>
> Sent from Yahoo Mail for iPhone
>
> On Wednesday, November 4, 2020, 10:46 PM, Tamas Karpati <tkarpati at gmail.com> wrote:
>
> In addition to my earlier comments, i'd like to mention that for kinetics
> you most probably need an SO2 simulation (optimization+phonons)
> rather than the same for a surface attached SO2 or SO+O. Big difference!
>
> Back to the earliers, 3-atoms phonon is so unphysical that it is
> recommended to do an all-atom one, or add at least the directly
> bonded surface atoms (and extend towards all-atoms if you can).
>
> In addition, only all-atom phonon will show you whether your big negative
> freqs. indicate a non-minimum structure (ie. "freqs" ~ 2nd derivatives
> of the PEHS).
> With somewhat less atoms you can be lucky, and by visualizing normal
> modes (phonons) by eg. XCrysDen will show you where to move atoms
> to get into the local minimum. Kinetics theory builds on minima (and TS-es).
> Anyways, 3 atoms are too few (also see first section above).
>
> On Mon, Nov 2, 2020 at 11:29 AM Tamas Karpati <tkarpati at gmail.com> wrote:
> >
> > Dear Omer,
> >
> > I guess that your input is fine, your structure is not.
> > (By the way, tr2_ph could be lower.)
> >
> > You woud expect 6 near zero and 3N-6 positive freqs. for a completely
> > relaxed minimum structure (and 5 + 3N-5 for a TS).
> > This is ruined if you run PH.x on a different potential energy
> > hypersurface, PEHS.
> >
> > It is really very easy to spoil: use a different no. of k point or functional
> > for PW/vc-relax and PH and you're there. Another temptation is to
> > use experimental crystal structure and fix some/most atoms as such.
> > These all mean different PEHS'.
> > In addition, unconverged relaxation (or too loose convergence),
> > while moves on the same PEHS, provides you with inappropriate
> > freqs, as it does not bring your structure close enough to the local minimum.
> >
> > I would recommend to reconsider the "life" of your structure
> > (origin, optimization method, other parameters) and adjust if necessary.
> > t
> >
> > On Mon, Nov 2, 2020 at 7:43 AM Omer Mutasim <omermutasim at ymail.com> wrote:
> > >
> > > Dear all
> > >
> > > I'm doing phonon calculation at Gamma point (q) in order to estimate the reaction rate constants for a micro-kinetic model. I have perturbed only the adsorbate molecule with the 3 surface atoms, connected to adsorbate, using "nat-todo" option. However, i got 15 negative frequencies (should be 6 as i know ) ,with high absolute value.
> > >
> > > Can you please help me to know what is wrong with my input files ?
> > >
> > > Below are the output & input files:
> > >
> > >
> > >
> > > Mode symmetry, C_1 (1) point group:
> > >
> > > freq ( 1 - 1) = -2762.6 [cm-1] --> A I+R
> > > freq ( 2 - 2) = -2570.3 [cm-1] --> A I+R
> > > freq ( 3 - 3) = -2460.4 [cm-1] --> A I+R
> > > freq ( 4 - 4) = -2423.6 [cm-1] --> A I+R
> > > freq ( 5 - 5) = -2356.3 [cm-1] --> A I+R
> > > freq ( 6 - 6) = -2158.0 [cm-1] --> A I+R
> > > freq ( 7 - 7) = -2151.1 [cm-1] --> A I+R
> > > freq ( 8 - 8) = -2067.5 [cm-1] --> A I+R
> > > freq ( 9 - 9) = -2034.8 [cm-1] --> A I+R
> > > freq ( 10 - 10) = -2025.2 [cm-1] --> A I+R
> > > freq ( 11 - 11) = -1864.3 [cm-1] --> A I+R
> > > freq ( 12 - 12) = -1804.5 [cm-1] --> A I+R
> > > freq ( 13 - 13) = -1099.4 [cm-1] --> A I+R
> > > freq ( 14 - 14) = -947.6 [cm-1] --> A I+R
> > > freq ( 15 - 15) = -912.5 [cm-1] --> A I+R
> > > freq (316 -316) = 179.3 [cm-1] --> A I+R
> > > freq (317 -317) = 193.0 [cm-1] --> A I+R
> > > freq (318 -318) = 215.8 [cm-1] --> A I+R
> > > freq (319 -319) = 240.2 [cm-1] --> A I+R
> > > freq (320 -320) = 270.4 [cm-1] --> A I+R
> > > freq (321 -321) = 317.0 [cm-1] --> A I+R
> > > freq (322 -322) = 370.8 [cm-1] --> A I+R
> > > freq (323 -323) = 377.3 [cm-1] --> A I+R
> > > freq (324 -324) = 398.3 [cm-1] --> A I+R
> > > freq (325 -325) = 417.8 [cm-1] --> A I+R
> > > freq (326 -326) = 468.2 [cm-1] --> A I+R
> > > freq (327 -327) = 659.0 [cm-1] --> A I+R
> > > freq (328 -328) = 1096.6 [cm-1] --> A I+R
> > > freq (329 -329) = 1795.5 [cm-1] --> A I+R
> > > freq (330 -330) = 2199.3 [cm-1] --> A I+R
> > >
> > >
> > > ph.x input file:
> > >
> > > phonon calculation at Gamma point.
> > > &inputph
> > > outdir = './outdir'
> > > prefix = 'HS'
> > > tr2_ph = 1.0d-09
> > > epsil = .false.
> > > amass(1) = 58.69340
> > > amass(2) = 30.97376
> > > amass(3) = 1.00784
> > > amass(4) = 32.065
> > > fildyn = 'HS.dyn'
> > >
> > > alpha_mix(1)=0.1
> > >
> > > recover=.true
> > > nogg = .true
> > > nat_todo = 5
> > >
> > > /
> > > 0.0 0.0 0.0
> > >
> > > 1 2 37 46 54
> > >
> > >
> > >
> > > scf input file:
> > >
> > > &CONTROL
> > > calculation = "scf"
> > > prefix = 'HS'
> > > outdir = './outdir'
> > > pseudo_dir = '/home/'
> > > restart_mode = 'from_scratch'
> > > forc_conv_thr = 1.0e-03
> > > etot_conv_thr = 1e-04
> > > nstep = 999
> > > /
> > > &SYSTEM
> > > ibrav = 0
> > > ecutrho = 200
> > > ecutwfc = 25
> > > nat = 110
> > > ntyp = 4
> > > occupations='smearing',smearing='gaussian',degauss=0.005
> > > vdw_corr = 'DFT-D2'
> > > nspin = 2
> > > starting_magnetization(1)= 0.01
> > > /
> > > &ELECTRONS
> > > conv_thr = 1e-8
> > > electron_maxstep = 200
> > > mixing_mode ='local-TF'
> > > mixing_beta = 0.3
> > > /
> > > &IONS
> > > /
> > > K_POINTS {automatic}
> > > 3 3 1 0 0 0
> > > ATOMIC_SPECIES
> > > Ni 58.69340 Ni.pbe-n-rrkjus_psl.0.1.UPF
> > > P 30.97376 P.pbe-n-rrkjus_psl.1.0.0.UPF
> > > H 1.00784 H.pbe-rrkjus_psl.0.1.UPF
> > > S 32.065 S.pbe-n-rrkjus_psl.1.0.0.UPF
> > > CELL_PARAMETERS {angstrom}
> > > 11.76538354 0.0000000000 0.0000000000
> > > -5.8826917709 10.189121032 0.0000000000
> > > 0.0000000000 0.0000000000 30.993869056
> > > ATOMIC_POSITIONS (angstrom)
> > > H 0.879694621 3.392266427 10.708999692
> > > S 2.266698845 3.396363162 10.560733430
> > > Ni -2.744571590 4.755054131 0.244939179
> > > Ni 3.134031329 1.363792691 0.248008546
> > > .
> > > .
> > > .
> > > P -1.060403962 1.841094610 1.604930623
> > > P -3.921453199 6.792156181 0.000000000 0 0 0
> > > P 1.960697149 3.396027080 0.000000000 0 0 0
> > > P 7.842906399 0.000000000 0.000000000 0 0 0
> > >
> > >
> > > regards
> > >
> > >
> > > _______________________________________________
> > > Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> > > users mailing list users at lists.quantum-espresso.org
> > > https://lists.quantum-espresso.org/mailman/listinfo/users
> _______________________________________________
> Quantum ESPRESSO is supported by MaX (www.max-centre.eu)
> users mailing list users at lists.quantum-espresso.org
> https://lists.quantum-espresso.org/mailman/listinfo/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20201118/9ab7e203/attachment.html>
More information about the users
mailing list