[CPMD-list] Liquid calculation (fwd)
Chiranjib Majumder
chimaju at magnum.barc.ernet.in
Tue Nov 12 10:38:19 CET 2002
Dear apsi
Thanks a lot for your reply. I am pleased to see that the fluctuations
in the temperature is around 55 to 60 as you have predicted.
I shall appreciate if can tell me how i should impose a periodic boundary
condition. What parameters are required for that.
It needs to be mentioned that I wanted to do a calculation for Na liquid
(which you are mentioning as cluster). I suppose you have guessed this
idea from my input. As I told you that this is my first calculation for a
liquid system so I don't know what parameters I should fix for liquid
calculation as special. What I have done is that the cell length is fixed
according to the XYZ coordinates.
Above all what exactly I want to know is what parameters I should use to
do a calculation for an extended system like liquid or solid with only
gamma point.
Thanking you in advance
Chiranjib
On Tue, 5 Nov 2002 Ari.P.Seitsonen at iki.fi wrote:
>
> Dear Chiranjib,
>
> Please excuse me for the very late reply, I have had such a busy
> time in the last days!! (E.g. yesterday cleaning my "new" apartment
> for six hours etc, very scientific ;)
>
> > For the first time I am trying to run a liquid system. I have taken 32
> > atoms of Na and positioned them in a fcc lattice points. To learn the
> > calculations, I thought it is better to try with single gamma point.
> > When we do single gamma point for isolated systems like
> > cluster we keep some vacuum between two clusters but I thought for an
> > extended liquid I should not keep any vacuum. So my cell length is fixed
> > accordingly. I really don't know if I am wrong. Pl. correct me. I didn't
> > fix NOSE parameters and would like to take the default vales.
> >
> > Overall I told you about my problem and I need to know how best I can
> > perform a finite temperature calculation using this CPMD. Also I don't
> > know whether I need to use some parameters to vary the occupancies.
> >
> > My input file is attached herewith. This input file preceeds another input
> > where I took the optimized wavefunctions for this atomic positions.
> >
> > I hope this information will be sufficient for you to answere my querries
> > so that I could proceed further.
> >
> > Thanking you
> >
> > Chiranjib
> >
> > &CPMD
> > MOLECULAR DYNAMICS
> > RESTART WAVEFUNCTION COORDINATES VELOCITIES LATEST
> > NOSE IONS
> > 400 2000
> > NOSE ELECTRONS
> > 0.0001 6000
> > MAXSTEP
> > 20000
> > TRAJECTORY SAMPLE
> > 10
> > &END
> >
> > &SYSTEM
> > ANGSTORM
> > SYMMETRY
> > 1
> > CELL
> > 10.97 1.0 1.0 0 0 0
> > CUTOFF
> > 15.0
> > &END
> > &ATOMS
> > ISOTOPE
> > 23.0
> >
> > *NA_SGS KLEINMAN-BYLANDER
> > LMAX=S
> > 32
> > 0.000000000 0.000000000 0.000000000
> > 2.743362646 2.743362646 0.000000000
> > 0.000000000 2.743362646 2.743362646
> > 2.743362646 0.000000000 2.743362646
> > 5.486725291 0.000000000 0.000000000
> > 8.230087937 2.743362646 0.000000000
> > 5.486725291 2.743362646 2.743362646
> > 8.230087937 0.000000000 2.743362646
> > 0.000000000 5.486725291 0.000000000
> > 2.743362646 8.230087937 0.000000000
> > 0.000000000 8.230087937 2.743362646
> > 2.743362646 5.486725291 2.743362646
> > 5.486725291 5.486725291 0.000000000
> > 8.230087937 8.230087937 0.000000000
> > 5.486725291 8.230087937 2.743362646
> > 8.230087937 5.486725291 2.743362646
> > 0.000000000 0.000000000 5.486725291
> > 2.743362646 2.743362646 5.486725291
> > 0.000000000 2.743362646 8.230087937
> > 2.743362646 0.000000000 8.230087937
> > 5.486725291 0.000000000 5.486725291
> > 8.230087937 2.743362646 5.486725291
> > 5.486725291 2.743362646 8.230087937
> > 8.230087937 0.000000000 8.230087937
> > 0.000000000 5.486725291 5.486725291
> > 2.743362646 8.230087937 5.486725291
> > 0.000000000 8.230087937 8.230087937
> > 2.743362646 5.486725291 8.230087937
> > 5.486725291 5.486725291 5.486725291
> > 8.230087937 8.230087937 5.486725291
> > 5.486725291 8.230087937 8.230087937
> > 8.230087937 5.486725291 8.230087937
> > END
> >
> > &DFT
> > NEWCODE
> > FUNCTIONAL LDA
> > LDA CORRELATION
> > &END
>
> Hmm, I don't remember your original question well, was it why the
> temperature fluctuates strongly even though you assign a Nose
> thermostat for the ions? From what I guess, the characteristic
> frequency for the ions is too high, I don't think that you'd have the
> ions oscillating at 2000 wave numbers in a sodium cluster, that's
> namely the frequency of strong, covalent bonds like C-O (carbon
> monoxide) or heavy water (deuterated water). Do you have some
> estimate, e.g. from smaller clusters? Then you could take e.g. 90 % of
> the highest frequency, or some frequency which is close to the maximum
> in the vibrational density of states.
>
> The second point is more physical: In such a small cluster the
> temperature necessarily would oscillate, and even though the Nose
> thermostat would dampen it, it's not necessarily physical. For
> example, the trajectory of the ions is somewhat affected. The
> thermostat means that there's an external "force" or ensemble which
> transfers energy to or from the systems, e.g. a molecule inserted in
> water. However, if you just would like to have the properties of a
> single molecule in vacuum the constant temperature is not well
> defined, it's more an average temperature, also in experiments. A
> collegue once showed (it's certainly something trivial to do the
> derivation) me that the mean oscillation of a system with N atoms is
>
> dT = sqrt [ 2 / ( 3 N ) ] * T_wanted
>
> so if you have 32 atoms at 400 K the fluctuations would be about 58 K.
>
> Hopefully this helps you somewhat!
>
> Greetings from Zurich!,
>
> apsi
>
> --
> -=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-=*=-
> Ari Paavo Seitsonen / Ari.P.Seitsonen at iki.fi / http://www.iki.fi/~apsi/
> Tel +41 1 635 44 97 / Fax +41 1 635 68 38 / GSM +41 79 719 09 35
> Anschrift: Physikalisch Chemisches Institut (PCI), Universität Zürich (UniZh)
> Indirizzo: Winterthurerstraße 190, CH-8057 Zürich
> Address: Schweiz / Svizzera / Suisse / Svizra / Switzerland
>
More information about the CPMD-list
mailing list