[Pw_forum] bfgs optimization not going anywhere

[Matej Huš]

Dear PW community

I have a problem when trying to find optimized geometry of an adsorbed
intermediate in "benzene" hydrogenation on Ru surface (see input below).

I believe I am fairly close to optimised structure since I'm only adding one
hydrogen  atom to an already optimised reactant and only slight displacing
the adjacent atoms.

 

The problem is that BFGS doesn't seem to go anywhere. Energy only
fluctuates, which to my understanding is a sign of being close to a minimum.
However, forces remain large.

 

I've switched to the damp algorithm and the problem persists. I've searched
through the forum and the usual approaches of tackling this issue do not
seem to work.

 

Any ideas? Thank you!

Matej Hus

Laboratory of Catalysis and Chemcial Reaction Engineering

National Institute of Chemistry

Ljubljana, Slovenia



OUTPUT:

Total force = 0.377547 Total SCF correction = 0.000233

Total force = 0.290735 Total SCF correction = 0.000272

Total force = 0.292438 Total SCF correction = 0.000238

Total force = 0.338997 Total SCF correction = 0.000229

Total force = 0.358435 Total SCF correction = 0.000297

Total force = 0.337403 Total SCF correction = 0.28

Total force = 0.348615 Total SCF correction = 0.12

Total force = 0.353731 Total SCF correction = 0.27

Total force = 0.356168 Total SCF correction = 0.37

Total force = 0.357346 Total SCF correction = 0.22

Total force = 0.357909 Total SCF correction = 0.28

Total force = 0.358211 Total SCF correction = 0.40

Total force = 0.358327 Total SCF correction = 0.48

Total force = 0.358419 Total SCF correction = 0.29

Total force = 0.288584 Total SCF correction = 0.13

Total force = 0.281519 Total SCF correction = 0.24

  Total force = 0.325271 Total SCF correction = 0.22

  Total force = 0.342948 Total SCF correction = 0.49

  Total force = 0.350916 Total SCF correction = 0.26

  Total force = 0.354742 Total SCF correction = 0.15

 

total energy  =   -7426.57738045 Ry

total energy  =   -7426.51356677 Ry

total energy  =   -7426.57388244 Ry

total energy  =   -7426.57758011 Ry

total energy  =   -7426.57784907 Ry

total energy  =   -7426.57707917 Ry

total energy  =   -7426.57763184 Ry

total energy  =   -7426.5696 Ry

total energy  =   -7426.57782155 Ry

total energy  =   -7426.57783739 Ry

total energy  =   -7426.57784377 Ry

total energy  =   -7426.57784657 Ry

total energy  =   -7426.57784787 Ry

total energy  =   -7426.57784849 Ry

total energy  =   -7426.51197755 Ry

total energy  =   -7426.57231380 Ry

  total energy  =   -7426.57690438 Ry

  total energy  =   -7426.57767359 Ry

  total energy  =   -7426.57782320 Ry

  total energy  =   -7426.57785046 Ry

 

 

 

INPUT

 

&CONTROL

 calculation = 'relax' ,

restart_mode = 'from_scratch' ,

  outdir = './tmp/$filename/' ,

  pseudo_dir = '../../pseudo/' ,

  prefix = '$PROJECT.$PREFIX' ,

 tstress = .true. ,

 tprnfor = .true. ,

   tefield = .true.
,

   dipfield = .true.
,

 wf_collect = .true. ,

  forc_conv_thr = 1.0d-2 ,

  etot_conv_thr = 1.0d-3 ,

/

&SYSTEM

   ibrav = 0,

 nat = 169,

ntyp = 4,

 ecutwfc = 50. ,

 ecutrho = 400. ,

 occupations = 'smearing' ,

 degauss = 0.01 ,

smearing = 'gaussian' ,

   nspin = 2,

   starting_magnetization(3) = 0.1D0,


edir = 3 ,

 emaxpos = 0.70 ,

  eopreg = 0.10 ,

eamp = 0. ,

vdw_corr = 'grimme-d2' ,  

/

&ELECTRONS

conv_thr = 1.0d-8 ,

 mixing_mode = 'local-TF' ,

 mixing_beta = 0.7 ,

 diagonalization = 'david' ,

/

&IONS

   ion_dynamics = 'bfgs'

/

ATOMIC_SPECIES

  C  12.0107   C.pbe-van_ak.UPF

  O  15.9994   O.pbe-van_ak.UPF

  H  1.00794   H.pbe-van_ak.UPF 

  Ru 101.07  Ru.pbe-n-van.UPF 

CELL_PARAMETERS (angstrom)

  16.32120

Re: [Pw_forum] bfgs optimization not going anywhere

[Giuseppe Mattioli]

Dear Matej
I do not see anything particularly wrong in your input. But you are putting 
together a lot of things that do not make convergence easier...
1) open shell (are you sure you need it even if your total number of electrons 
is odd?)
2) dipole correction (maybe you can add the correction after a bit of 
optimization...)
3) k-points (block all the metal atoms and use gamma only. Then refine the 
optimization, if you are sure you need dispersion, or perform a nscf 
calculation to have DOS or whatever)
4) DFT-D2 (known huge overestimation of metal-molecule interaction, see 
Grimme's 2010 paper if I remember well. vdwdf is better for such systems, and 
the open-shell implementation has been recently added to the code).
You may try to switch off one or two of such features and see what happens. It 
is not a solution to your problem, I know...
HTH
Giuseppe
 

On Tuesday, November 24, 2015 09:43:02 AM Matej Huš wrote:
> Dear PW community
> 
> I have a problem when trying to find optimized geometry of an adsorbed
> intermediate in "benzene" hydrogenation on Ru surface (see input below).
> 
> I believe I am fairly close to optimised structure since I'm only adding one
> hydrogen  atom to an already optimised reactant and only slight displacing
> the adjacent atoms.
> 
> 
> 
> The problem is that BFGS doesn't seem to go anywhere. Energy only
> fluctuates, which to my understanding is a sign of being close to a minimum.
> However, forces remain large.
> 
> 
> 
> I've switched to the damp algorithm and the problem persists. I've searched
> through the forum and the usual approaches of tackling this issue do not
> seem to work.
> 
> 
> 
> Any ideas? Thank you!
> 
> Matej Hus
> 
> Laboratory of Catalysis and Chemcial Reaction Engineering
> 
> National Institute of Chemistry
> 
> Ljubljana, Slovenia
> 
> 
> 
> OUTPUT:
> 
> Total force = 0.377547 Total SCF correction = 0.000233
> 
> Total force = 0.290735 Total SCF correction = 0.000272
> 
> Total force = 0.292438 Total SCF correction = 0.000238
> 
> Total force = 0.338997 Total SCF correction = 0.000229
> 
> Total force = 0.358435 Total SCF correction = 0.000297
> 
> Total force = 0.337403 Total SCF correction = 0.28
> 
> Total force = 0.348615 Total SCF correction = 0.12
> 
> Total force = 0.353731 Total SCF correction = 0.27
> 
> Total force = 0.356168 Total SCF correction = 0.37
> 
> Total force = 0.357346 Total SCF correction = 0.22
> 
> Total force = 0.357909 Total SCF correction = 0.28
> 
> Total force = 0.358211 Total SCF correction = 0.40
> 
> Total force = 0.358327 Total SCF correction = 0.48
> 
> Total force = 0.358419 Total SCF correction = 0.29
> 
> Total force = 0.288584 Total SCF correction = 0.13
> 
> Total force = 0.281519 Total SCF correction = 0.24
> 
>   Total force = 0.325271 Total SCF correction = 0.22
> 
>   Total force = 0.342948 Total SCF correction = 0.49
> 
>   Total force = 0.350916 Total SCF correction = 0.26
> 
>   Total force = 0.354742 Total SCF correction = 0.15
> 
> 
> 
> total energy  =   -7426.57738045 Ry
> 
> total energy  =   -7426.51356677 Ry
> 
> total energy  =   -7426.57388244 Ry
> 
> total energy  =   -7426.57758011 Ry
> 
> total energy  =   -7426.57784907 Ry
> 
> total energy  =   -7426.57707917 Ry
> 
> total energy  =   -7426.57763184 Ry
> 
> total energy  =   -7426.5696 Ry
> 
> total energy  =   -7426.57782155 Ry
> 
> total energy  =   -7426.57783739 Ry
> 
> total energy  =   -7426.57784377 Ry
> 
> total energy  =   -7426.57784657 Ry
> 
> total energy  =   -7426.57784787 Ry
> 
> total energy  =   -7426.57784849 Ry
> 
> total energy  =   -7426.51197755 Ry
> 
> total energy  =   -7426.57231380 Ry
> 
>   total energy  =   -7426.57690438 Ry
> 
>   total energy  =   -7426.57767359 Ry
> 
>   total energy  =   -7426.57782320 Ry
> 
>   total energy  =   -7426.57785046 Ry
> 
> 
> 
> 
> 
> 
> 
> INPUT
> 
> 
> 
> &CONTROL
> 
>  calculation = 'relax' ,
> 
> restart_mode = 'from_scratch' ,
> 
>   outdir = './tmp/$filename/' ,
> 
>   pseudo_dir = '../../pseudo/' ,
> 
>   prefix = '$PROJECT.$PREFIX' ,
> 
>  tstress = .true. ,
> 
>  tprnfor = .true. ,
> 
>tefield = .true.
> ,
> 
>dipfield = .true.
> ,
> 
>  wf_collect

Re: [Pw_forum] bfgs optimization not going anywhere

[Jure Varlec]

On Tue, 24 Nov 2015 09:43:02 +0100
Matej Huš  wrote:

> I believe I am fairly close to optimised structure since I'm only
> adding one hydrogen  atom to an already optimised reactant and only
> slight displacing the adjacent atoms.

But have you converged the calculation with respect to ecutwfc
after adding the hydrogen atom? 50 Ry looks quite low to me. BTW, it's
good to find another QE user here at KI, I got the impression there
aren't (m)any. You're welcome to visit anytime :) 

Regards
Jure Varlec, L09

___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] bfgs optimization not going anywhere

[Giuseppe Mattioli]

> But have you converged the calculation with respect to ecutwfc
> after adding the hydrogen atom? 50 Ry looks quite low to me.

Rather high with US pseudopotentials. And you should not check again because H 
was already present in the system.

HTH
Giuseppe

On Tuesday, November 24, 2015 11:02:05 AM Jure Varlec wrote:
> On Tue, 24 Nov 2015 09:43:02 +0100
> 
> Matej Huš  wrote:
> > I believe I am fairly close to optimised structure since I'm only
> > adding one hydrogen  atom to an already optimised reactant and only
> > slight displacing the adjacent atoms.
> 
> But have you converged the calculation with respect to ecutwfc
> after adding the hydrogen atom? 50 Ry looks quite low to me. BTW, it's
> good to find another QE user here at KI, I got the impression there
> aren't (m)any. You're welcome to visit anytime :)
> 
> Regards
> Jure Varlec, L09
> 
> ___
> Pw_forum mailing list
> Pw_forum@pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum


- Article premier - Les hommes naissent et demeurent
libres et ègaux en droits. Les distinctions sociales
ne peuvent être fondèes que sur l'utilitè commune
- Article 2 - Le but de toute association politique
est la conservation des droits naturels et 
imprescriptibles de l'homme. Ces droits sont la libertè,
la propriètè, la sùretè et la rèsistance à l'oppression.


   Giuseppe Mattioli
   CNR - ISTITUTO DI STRUTTURA DELLA MATERIA   
   v. Salaria Km 29,300 - C.P. 10
   I 00015 - Monterotondo Stazione (RM), Italy
   Tel + 39 06 90672836 - Fax +39 06 90672316
   E-mail: 
   http://www.ism.cnr.it/english/staff/mattiolig
   ResearcherID: F-6308-2012

___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] bfgs optimization not going anywhere

[Ari P Seitsonen]

Dear Matej Hus,

  Adding to the previous comments (I think that it is ok to use the k 
points, even though you might gain some CPU time by first indeed using 
Gamma-only and then increasing, as your cell is quite large), have you 
checked the length of your vacuum/lattice vector along the surface normal? 
If I see correctly, you have only about six Ångströms, which is already 
very (too) little, plus the DFT-D2 (Grimme-D2) adds quite 
strong an interaction through the vacuum at such distances I believe, 
plus the dipole correction...


Greetings from Peaceful Montrouge/France,

   apsi

PS My personal opinion of the vdW-DF* is, yes, they might be better, but 
first check the literature about a "suitable empirical choice of the 
particular mixture" of the functional, as some of the choices give quite 
large lattice constants already (4-5 % larger than experimental, which, 
of course, is not necessarily the target, rather a good adsorption 
geometry and energy). And often the best agreement is found by tuning the 
exchange functional, which naturally should have not nothing to do with 
the vdW/London dispersion, which is pure correlation... So much about "ab 
initio". ;)  The good news of the day is that it seems that seldomly 
things seem to go completely wrong, no matter which choice of the vdW-DF* 
you take, but still, I would study a bit the literature first. If any one 
has a reference for a good review on the vdW-DF* et co applied to surfaces 
and adsorption, I would also be very thankful! :)


-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
  Ari Paavo Seitsonen / ari.p.seitso...@iki.fi / http://www.iki.fi/~apsi/
  Ecole Normale Supérieure (ENS), Département de Chimie, Paris
  Mobile (F) : +33 789 37 24 25(CH) : +41 79 71 90 935


On Tue, 24 Nov 2015, Matej Huš wrote:



Dear PW community

I have a problem when trying to find optimized geometry of an adsorbed
intermediate in "benzene" hydrogenation on Ru surface (see input below).

I believe I am fairly close to optimised structure since I'm only adding one
hydrogen  atom to an already optimised reactant and only slight displacing
the adjacent atoms.

 

The problem is that BFGS doesn't seem to go anywhere. Energy only
fluctuates, which to my understanding is a sign of being close to a minimum.
However, forces remain large.

 

I've switched to the damp algorithm and the problem persists. I've searched
through the forum and the usual approaches of tackling this issue do not
seem to work.

 

Any ideas? Thank you!

Matej Hus

Laboratory of Catalysis and Chemcial Reaction Engineering

National Institute of Chemistry

Ljubljana, Slovenia
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] bfgs optimization not going anywhere

[Matej Huš]

Thanks for very useful comments.

Yes, vacuum region is indeed too low in this example (whoops, seems like
I've attached an old input file). I also ran a calculation with 10 A of
vacuum without any improvement. Open shell is unnecessary as it quickly
converges to a closed shell solution. I've just wasted some CPU time with
that.

Ectuwfc, ecutrhc are large enough (converged). I believe K points should
also not be the source of a problem.

As for DFT-D2, I admit I am a bit in the dark here. How 'wrong' is it? Or to
put differently, is it more wrong that not using dispersion correction at
all? How about vdw-DF, how do I turn it on and does it perform better? Where
can I find more important about that?

Also, if I change/switch off the dispersion correction, then I will have to
re-optimize all already converged structures (metal, isolated molecules,
adsorbed intermediates), right? Looks rather tedious


Matej



-Original Message-
From: pw_forum-boun...@pwscf.org [mailto:pw_forum-boun...@pwscf.org] On
Behalf Of Ari P Seitsonen
Sent: Tuesday, November 24, 2015 11:22 AM
To: PWSCF Forum 
Subject: Re: [Pw_forum] bfgs optimization not going anywhere


Dear Matej Hus,

   Adding to the previous comments (I think that it is ok to use the k
points, even though you might gain some CPU time by first indeed using
Gamma-only and then increasing, as your cell is quite large), have you
checked the length of your vacuum/lattice vector along the surface normal? 
If I see correctly, you have only about six Ångströms, which is already very
(too) little, plus the DFT-D2 (Grimme-D2) adds quite strong an interaction
through the vacuum at such distances I believe, plus the dipole
correction...

 Greetings from Peaceful Montrouge/France,

apsi

PS My personal opinion of the vdW-DF* is, yes, they might be better, but
first check the literature about a "suitable empirical choice of the
particular mixture" of the functional, as some of the choices give quite
large lattice constants already (4-5 % larger than experimental, which, of
course, is not necessarily the target, rather a good adsorption geometry and
energy). And often the best agreement is found by tuning the exchange
functional, which naturally should have not nothing to do with the
vdW/London dispersion, which is pure correlation... So much about "ab
initio". ;)  The good news of the day is that it seems that seldomly things
seem to go completely wrong, no matter which choice of the vdW-DF* you take,
but still, I would study a bit the literature first. If any one has a
reference for a good review on the vdW-DF* et co applied to surfaces and
adsorption, I would also be very thankful! :)

-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=
-
   Ari Paavo Seitsonen / ari.p.seitso...@iki.fi / http://www.iki.fi/~apsi/
   Ecole Normale Supérieure (ENS), Département de Chimie, Paris
   Mobile (F) : +33 789 37 24 25(CH) : +41 79 71 90 935


On Tue, 24 Nov 2015, Matej Huš wrote:

> 
> Dear PW community
> 
> I have a problem when trying to find optimized geometry of an adsorbed 
> intermediate in "benzene" hydrogenation on Ru surface (see input below).
> 
> I believe I am fairly close to optimised structure since I'm only 
> adding one hydrogen  atom to an already optimised reactant and only 
> slight displacing the adjacent atoms.
> 
>  
> 
> The problem is that BFGS doesn't seem to go anywhere. Energy only 
> fluctuates, which to my understanding is a sign of being close to a
minimum.
> However, forces remain large.
> 
>  
> 
> I've switched to the damp algorithm and the problem persists. I've 
> searched through the forum and the usual approaches of tackling this 
> issue do not seem to work.
> 
>  
> 
> Any ideas? Thank you!
> 
> Matej Hus
> 
> Laboratory of Catalysis and Chemcial Reaction Engineering
> 
> National Institute of Chemistry
> 
> Ljubljana, Slovenia
>


___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

[Pw_forum] i-v characteristics

[Valentina Cantatore]

Dear QuantumEspresso users and developers,

I am interested in calculating i-v characteristics of graphene and 
doped-graphene. I am new in this field so any hint will be appreciated. Is 
there any tool that I can use?

Thank you very much for your help.

Valentina

Valentina Cantatore
PostDoc Researcher
Department of Chemistry and Chemical Engineering
Energy and Materials

+393471040381 (mobile)

[cid:image001.jpg@01CC2F3C.8093B9D0]
Chalmers University of Technology
Kemivägen 10
SE-412 96 Göteborg, Sweden
www.chalmers.se
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] bfgs optimization not going anywhere

[Paolo Giannozzi]

Just an observation: there are constrained atoms. I am not sure how well
current convergence criteria for BFGS deal with this case.

Paolo

On Tue, Nov 24, 2015 at 11:37 AM, Matej Huš  wrote:

> Thanks for very useful comments.
>
> Yes, vacuum region is indeed too low in this example (whoops, seems like
> I've attached an old input file). I also ran a calculation with 10 A of
> vacuum without any improvement. Open shell is unnecessary as it quickly
> converges to a closed shell solution. I've just wasted some CPU time with
> that.
>
> Ectuwfc, ecutrhc are large enough (converged). I believe K points should
> also not be the source of a problem.
>
> As for DFT-D2, I admit I am a bit in the dark here. How 'wrong' is it? Or
> to
> put differently, is it more wrong that not using dispersion correction at
> all? How about vdw-DF, how do I turn it on and does it perform better?
> Where
> can I find more important about that?
>
> Also, if I change/switch off the dispersion correction, then I will have to
> re-optimize all already converged structures (metal, isolated molecules,
> adsorbed intermediates), right? Looks rather tedious
>
>
> Matej
>
>
>
> -Original Message-
> From: pw_forum-boun...@pwscf.org [mailto:pw_forum-boun...@pwscf.org] On
> Behalf Of Ari P Seitsonen
> Sent: Tuesday, November 24, 2015 11:22 AM
> To: PWSCF Forum 
> Subject: Re: [Pw_forum] bfgs optimization not going anywhere
>
>
> Dear Matej Hus,
>
>Adding to the previous comments (I think that it is ok to use the k
> points, even though you might gain some CPU time by first indeed using
> Gamma-only and then increasing, as your cell is quite large), have you
> checked the length of your vacuum/lattice vector along the surface normal?
> If I see correctly, you have only about six Ångströms, which is already
> very
> (too) little, plus the DFT-D2 (Grimme-D2) adds quite strong an interaction
> through the vacuum at such distances I believe, plus the dipole
> correction...
>
>  Greetings from Peaceful Montrouge/France,
>
> apsi
>
> PS My personal opinion of the vdW-DF* is, yes, they might be better, but
> first check the literature about a "suitable empirical choice of the
> particular mixture" of the functional, as some of the choices give quite
> large lattice constants already (4-5 % larger than experimental, which, of
> course, is not necessarily the target, rather a good adsorption geometry
> and
> energy). And often the best agreement is found by tuning the exchange
> functional, which naturally should have not nothing to do with the
> vdW/London dispersion, which is pure correlation... So much about "ab
> initio". ;)  The good news of the day is that it seems that seldomly things
> seem to go completely wrong, no matter which choice of the vdW-DF* you
> take,
> but still, I would study a bit the literature first. If any one has a
> reference for a good review on the vdW-DF* et co applied to surfaces and
> adsorption, I would also be very thankful! :)
>
>
> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=
> -
>Ari Paavo Seitsonen / ari.p.seitso...@iki.fi / http://www.iki.fi/~apsi/
>Ecole Normale Supérieure (ENS), Département de Chimie, Paris
>Mobile (F) : +33 789 37 24 25(CH) : +41 79 71 90 935
>
>
> On Tue, 24 Nov 2015, Matej Huš wrote:
>
> >
> > Dear PW community
> >
> > I have a problem when trying to find optimized geometry of an adsorbed
> > intermediate in "benzene" hydrogenation on Ru surface (see input below).
> >
> > I believe I am fairly close to optimised structure since I'm only
> > adding one hydrogen  atom to an already optimised reactant and only
> > slight displacing the adjacent atoms.
> >
> >
> >
> > The problem is that BFGS doesn't seem to go anywhere. Energy only
> > fluctuates, which to my understanding is a sign of being close to a
> minimum.
> > However, forces remain large.
> >
> >
> >
> > I've switched to the damp algorithm and the problem persists. I've
> > searched through the forum and the usual approaches of tackling this
> > issue do not seem to work.
> >
> >
> >
> > Any ideas? Thank you!
> >
> > Matej Hus
> >
> > Laboratory of Catalysis and Chemcial Reaction Engineering
> >
> > National Institute of Chemistry
> >
> > Ljubljana, Slovenia
> >
>
>
> ___
> Pw_forum mailing list
> Pw_forum@pwscf.org
> http://pwscf.org/mailman/listinfo/pw_forum
>



-- 
Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum

Re: [Pw_forum] bfgs optimization not going anywhere

[Giuseppe Mattioli]

Dear Matej (and all)

> As for DFT-D2, I admit I am a bit in the dark here. How 'wrong' is it? Or to
> put differently, is it more wrong that not using dispersion correction at
> all?

Regarding the dispersion correction(s) this is the Stephan Grimme's paper which 
indicates that DFT-D2 doubles the DFT-D3 interaction energy between 
molecules and metal surfaces:

J. Chem. Phys. (2010) 132, 154104

In my experience of phthalocyanine/metal interactions (but if I remember well 
Ari has also performed calculations on porphyrin/phthalocyanine 
molecules adsorbed on metal surfaces and he can add his own knowledge), 
adsorption geometries and energies are very different, so that the results are 
also quite different. See for example this paper (Cu phthalocyanine interacting 
with Ag (100) with DFT-D2)

PRL 105, 115702 (2010)

And my paper (TiO phthalocyanine interacting with Ag (100) with vdW-DF)

J. Phys. Chem. C 2014, 118, 5255

And, yes, of course: if you use no dispersion correction you badly 
underestimate the interaction between the molecule and the surface...

2 more hints:

1) In the case of molecule/metal systems, ab initio functionals modeling long 
distance correlation between electrons should be (in my opinion) 
preferred to semiempirical functional with add pairwise forces based on 
tabulations (or partial calculations) of C6 coefficients. There is a nice 
perspective written by J. L. Bredas which addresses the issue

Nature Nanotechnology
8,
230–231
(2013)

2) Regarding lattice parameters, Ari is right in the case of vdwdf. But I've 
tried to simulate gold (bulk and) surfaces in contact with molecules by 
using the reparametrized vdwdf-c09 functional: lattice parameter of gold is ok, 
and also molecule-surface interactions are expected to be improved 
with respect to vdwdf!

> Also, if I change/switch off the dispersion correction, then I will have to
> re-optimize all already converged structures (metal, isolated molecules,
> adsorbed intermediates), right? Looks rather tedious

Yes, you should recalculate everything...:-(

HTH
Giuseppe

On Tuesday, November 24, 2015 11:37:45 AM Matej Huš wrote:
> Thanks for very useful comments.
> 
> Yes, vacuum region is indeed too low in this example (whoops, seems like
> I've attached an old input file). I also ran a calculation with 10 A of
> vacuum without any improvement. Open shell is unnecessary as it quickly
> converges to a closed shell solution. I've just wasted some CPU time with
> that.
> 
> Ectuwfc, ecutrhc are large enough (converged). I believe K points should
> also not be the source of a problem.
> 
> As for DFT-D2, I admit I am a bit in the dark here. How 'wrong' is it? Or to
> put differently, is it more wrong that not using dispersion correction at
> all? How about vdw-DF, how do I turn it on and does it perform better? Where
> can I find more important about that?
> 
> Also, if I change/switch off the dispersion correction, then I will have to
> re-optimize all already converged structures (metal, isolated molecules,
> adsorbed intermediates), right? Looks rather tedious
> 
> 
> Matej
> 
> 
> 
> -Original Message-
> From: pw_forum-boun...@pwscf.org [mailto:pw_forum-boun...@pwscf.org] On
> Behalf Of Ari P Seitsonen
> Sent: Tuesday, November 24, 2015 11:22 AM
> To: PWSCF Forum 
> Subject: Re: [Pw_forum] bfgs optimization not going anywhere
> 
> 
> Dear Matej Hus,
> 
>Adding to the previous comments (I think that it is ok to use the k
> points, even though you might gain some CPU time by first indeed using
> Gamma-only and then increasing, as your cell is quite large), have you
> checked the length of your vacuum/lattice vector along the surface normal?
> If I see correctly, you have only about six Ångströms, which is already very
> (too) little, plus the DFT-D2 (Grimme-D2) adds quite strong an interaction
> through the vacuum at such distances I believe, plus the dipole
> correction...
> 
>  Greetings from Peaceful Montrouge/France,
> 
> apsi
> 
> PS My personal opinion of the vdW-DF* is, yes, they might be better, but
> first check the literature about a "suitable empirical choice of the
> particular mixture" of the functional, as some of the choices give quite
> large lattice constants already (4-5 % larger than experimental, which, of
> course, is not necessarily the target, rather a good adsorption geometry and
> energy). And often the best agreement is found by tuning the exchange
> functional, which naturally should have not nothing to do with the
> vdW/London dispersion, which is pure correlation... So much about "ab
> initio". ;)  The good news of the day is that it seems that seldomly things
> seem to go completely wrong, no matter which choice of the vdW-DF* you take,
> but still, I would study a bit the literature first. If any one has a
> reference for a good review on the vdW-DF* et co applied to surfaces and
> adsorption, I would also be very thankful! :)
> 
> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=

Re: [Pw_forum] bfgs optimization not going anywhere

[Ilya Ryabinkin]

Colleagues,
to summarize: what would be 'the best' DFT correction on to of
*existing* functionals, e.g. PBE for molecule-metal-interaction? As
far as I understood, vdw-DF must be used with the corresponding
pseudopotential (BTW, where one can get them?).

--
I.


On Tue, Nov 24, 2015 at 8:30 AM, Giuseppe Mattioli
 wrote:
>
> Dear Matej (and all)
>
>> As for DFT-D2, I admit I am a bit in the dark here. How 'wrong' is it? Or to
>> put differently, is it more wrong that not using dispersion correction at
>> all?
>
> Regarding the dispersion correction(s) this is the Stephan Grimme's paper 
> which indicates that DFT-D2 doubles the DFT-D3 interaction energy between
> molecules and metal surfaces:
>
> J. Chem. Phys. (2010) 132, 154104
>
> In my experience of phthalocyanine/metal interactions (but if I remember well 
> Ari has also performed calculations on porphyrin/phthalocyanine
> molecules adsorbed on metal surfaces and he can add his own knowledge), 
> adsorption geometries and energies are very different, so that the results are
> also quite different. See for example this paper (Cu phthalocyanine 
> interacting with Ag (100) with DFT-D2)
>
> PRL 105, 115702 (2010)
>
> And my paper (TiO phthalocyanine interacting with Ag (100) with vdW-DF)
>
> J. Phys. Chem. C 2014, 118, 5255
>
> And, yes, of course: if you use no dispersion correction you badly 
> underestimate the interaction between the molecule and the surface...
>
> 2 more hints:
>
> 1) In the case of molecule/metal systems, ab initio functionals modeling long 
> distance correlation between electrons should be (in my opinion)
> preferred to semiempirical functional with add pairwise forces based on 
> tabulations (or partial calculations) of C6 coefficients. There is a nice
> perspective written by J. L. Bredas which addresses the issue
>
> Nature Nanotechnology
> 8,
> 230–231
> (2013)
>
> 2) Regarding lattice parameters, Ari is right in the case of vdwdf. But I've 
> tried to simulate gold (bulk and) surfaces in contact with molecules by
> using the reparametrized vdwdf-c09 functional: lattice parameter of gold is 
> ok, and also molecule-surface interactions are expected to be improved
> with respect to vdwdf!
>
>> Also, if I change/switch off the dispersion correction, then I will have to
>> re-optimize all already converged structures (metal, isolated molecules,
>> adsorbed intermediates), right? Looks rather tedious
>
> Yes, you should recalculate everything...:-(
>
> HTH
> Giuseppe
>
> On Tuesday, November 24, 2015 11:37:45 AM Matej Huš wrote:
>> Thanks for very useful comments.
>>
>> Yes, vacuum region is indeed too low in this example (whoops, seems like
>> I've attached an old input file). I also ran a calculation with 10 A of
>> vacuum without any improvement. Open shell is unnecessary as it quickly
>> converges to a closed shell solution. I've just wasted some CPU time with
>> that.
>>
>> Ectuwfc, ecutrhc are large enough (converged). I believe K points should
>> also not be the source of a problem.
>>
>> As for DFT-D2, I admit I am a bit in the dark here. How 'wrong' is it? Or to
>> put differently, is it more wrong that not using dispersion correction at
>> all? How about vdw-DF, how do I turn it on and does it perform better? Where
>> can I find more important about that?
>>
>> Also, if I change/switch off the dispersion correction, then I will have to
>> re-optimize all already converged structures (metal, isolated molecules,
>> adsorbed intermediates), right? Looks rather tedious
>>
>>
>> Matej
>>
>>
>>
>> -Original Message-
>> From: pw_forum-boun...@pwscf.org [mailto:pw_forum-boun...@pwscf.org] On
>> Behalf Of Ari P Seitsonen
>> Sent: Tuesday, November 24, 2015 11:22 AM
>> To: PWSCF Forum 
>> Subject: Re: [Pw_forum] bfgs optimization not going anywhere
>>
>>
>> Dear Matej Hus,
>>
>>Adding to the previous comments (I think that it is ok to use the k
>> points, even though you might gain some CPU time by first indeed using
>> Gamma-only and then increasing, as your cell is quite large), have you
>> checked the length of your vacuum/lattice vector along the surface normal?
>> If I see correctly, you have only about six Ångströms, which is already very
>> (too) little, plus the DFT-D2 (Grimme-D2) adds quite strong an interaction
>> through the vacuum at such distances I believe, plus the dipole
>> correction...
>>
>>  Greetings from Peaceful Montrouge/France,
>>
>> apsi
>>
>> PS My personal opinion of the vdW-DF* is, yes, they might be better, but
>> first check the literature about a "suitable empirical choice of the
>> particular mixture" of the functional, as some of the choices give quite
>> large lattice constants already (4-5 % larger than experimental, which, of
>> course, is not necessarily the target, rather a good adsorption geometry and
>> energy). And often the best agreement is found by tuning the exchange
>> functional, which naturally should have not nothing to do with the
>> vdW/Lon

[Pw_forum] Relaxation of hydrogenated amorphous carbon

[Alan J. R. Ambrozio]

Hi,

I'm trying to make a 'vc-relax' for an amorphous carbon structure (216
atoms 5% of H) obtained running a Quench from 6000K to 300K on LAMMPS. I'm
using gipaw pseudopotentials because I'll make some NMR calculations. To
have good results for NMR I need to have very good relaxed structures, but
to be honest, this is the first time that I'm running amorphous relaxation
on QE. I ran the vc-relax and scf normally, but when a try to run gipaw I
have an error message like this

"Error in routine cdiaghg (989):
 S matrix not positive definite"

I think this problem is because my structure isn't well relaxed... I
had this problem before with graphene... The think this problem can be
the numbers of atoms...


Could anyone give me some tricks to do good relaxation?

this is the input:

&CONTROL
  calculation='vc-relax'
  nstep=150
  tstress= .true.
  tprnfor= .true.
  prefix='50K'
  etot_conv_thr = 1.0D-4
  forc_conv_thr = 1.0D-3
  pseudo_dir ='/home/ambrozio/qe-gipaw-5.1/pseudo'
  outdir='./data/'

/
&SYSTEM
  ibrav = 0
  celldm(1)=  24.465150
  nat   = 216
  ntyp  = 2
  ecutwfc   = 100
  ecutrho   = 1200
  occupations ='smearing'  (I don't know if the system is metalic)
  spline_ps=.true.
  degauss=0.01
  london=.true.
  london_s6 = 0.75(use or not?  The structure has some parts
like graphite..)
/
&ELECTRONS
  electron_maxstep=5000
  conv_thr = 1.0D-6
  diagonalization='david'
  diago_thr_init = 1e-4
  mixing_mode = 'plain'
  mixing_beta = 0.7
/
&ions
upscale =100#this is default
/
&cell
Press = 0.0#this is default
/
ATOMIC_SPECIES
C  12.000  C.pbe-tm-gipaw.UPF
H   1.000   H.pbe-tm-gipaw.UPF

ATOMIC_POSITIONS {angstrom}

 C 0.7882546   0.6065626   0.7211832
 C 0.2881642  -0.0123605   0.2760093
 C 0.0228301   0.1905951   0.7547355
 C 0.2101520   0.4086151   0.1766798
 C 0.4457151   0.5643135   0.6661727
 C 0.7067007   0.7879631   0.4144110
 C 0.6855365   0.2258298   0.8198372
 C 0.1355208   0.0621604   0.8276883
 C 0.4363931   0.5141174   0.2279694
 C 0.5936407   0.7793334   0.4207085
 C 0.5859244   0.8848937   0.5767109
 C 0.5283677   0.3653187   0.3869548
 C 0.8312999   0.6246249   0.4091983
 C 0.9102840   0.1264440   0.3886683
 C 0.5469423   0.8867824   0.7639739
 C 0.2558971   0.4344034   0.4441222
 C 0.6662662   0.8590280   0.0741248
 C 0.6664248   0.5830016   0.8693308
 C 0.5466290   0.5737425   0.6999028
 H 0.4707029   0.2232765   0.6511726
 H 0.0072009   0.3950743   0.6588494
 C 0.8611044   0.4493360   0.3356426
 C 0.6113547   0.5271664   0.4131907
 C 0.1964175   0.7970353   0.8351151
 C 0.2675415   0.5302418   0.4948854
 C 0.7605494   0.7040310   0.6668560
 C 0.1249042   0.8634550   0.6759987
 C 0.7977951   0.2084323   0.2152986
 C 0.7664191   0.1562074   0.8031255
 C 0.6870201   0.1912958   0.3798247
 C 0.7330213   0.2870536   0.3933979
 H 0.7352675   0.6415087   0.0335573
 C 0.4422958   0.6995388   0.3650615
 C 0.5615385   0.5503336   0.5185775
 C 0.1080644   0.0077391   0.0778919
 C 0.0137435   0.4257022   0.9510966
 C 0.4321943   0.8571357   0.4714686
 C 0.4236154   0.0912046   0.8144852
 C 0.1254605   0.6280411   0.8449147
 C 0.0726748   0.4434211   0.4621965
 C 0.8205852   0.1096019   0.4662541
 C 0.3039631   0.1120866   0.6416651
 C 0.4248404   0.2169526   0.3711035
 C 0.4480209   0.1353904   0.2180935
 C 0.7788020   0.2023157   0.6274172
 C 0.6988667   0.3412404   0.0080584
 C 0.0707365   0.2193770   0.4688107
 C 0.5201701   0.2634695   0.3564601
 C 0.1846393   0.4134481   0.0784198
 C 0.4773098   0.6538072   0.9542338
 C 0.1421370   0.6011298   0.7244798
 C 0.4816710   0.8609266   0.9401489
 C 0.1243995   0.4561090   0.9586360
 C 0.6979409   0.3656521   0.5866604
 C 0.5678397   0.5603686   0.8130475
 C 0.7817233   0.4107388   0.0294312
 C 0.7626229   0.5919446   0.8315008
 H 0.1012895   0.7976846   0.2087955
 C 0.5516746   0.7019484   0.3614000
 C 0.1340584   0.0528037   0.7209917
 C 0.3689985   0.5777822   0.4903572
 C 0.4901815   0.3781462   0.8693084
 C 0.8851079   0.4363028   0.0200746
 C 0.1846087   0.5720996   0.5447761
 C 0.8238119   0.1755254   0.199
 C 0.0380825   0.0904789   0.0564483
 C 0.1225714   0.7981857   0.5002325
 C 0.4431491   0.4194498   0.9662942
 C 0.3424259   0.2698411   0.4176993
 C 0.6320468   0.4064309   0.0657004
 C 0.4744210   0.6037999   0.8630263
 C-0.0645495   0.0378500   0.5134204
 C-0.0251843   0.4104928   0.4205102
 C 0.3722911   0.7383781   0.0915926
 C 0.2422193   0.7343280   0.6339831
 C 0.1143970   0.1530761   0.6753776
 C 0.5932143   0.8382484   0.9562124
 C 0.0641976   0.6583118   0.0934076
 C 0.4117867   0.8358148   0.0525381
 C 0.4514283   0.5697306   0.5537673
 C 0.0411872   0.6205243   0.6

Re: [Pw_forum] Relaxation of hydrogenated amorphous carbon

[Paolo Giannozzi]

On Wed, Nov 25, 2015 at 2:49 AM, Alan J. R. Ambrozio 
wrote:

I think this problem is because my structure isn't well relaxed...
>
>
I think this problem is because your structure is not what you think: the
following line is not correct:
  ATOMIC_POSITIONS {angstrom}
With this:
  ATOMIC_POSITIONS {alat}
your structure looks much better (use xcrysden to visualize the input, or
auxiliary code "dist.x" to see the distances between atoms)

Paolo
-- 
Paolo Giannozzi, Dept. Chemistry&Physics&Environment,
Univ. Udine, via delle Scienze 208, 33100 Udine, Italy
Phone +39-0432-558216, fax +39-0432-558222
___
Pw_forum mailing list
Pw_forum@pwscf.org
http://pwscf.org/mailman/listinfo/pw_forum