[CPMD-list] Problems optimizing monoclinic unit cell
Ewald Pauwels
ewald.pauwels at UGent.be
Mon Dec 6 15:33:26 CET 2004
Dear all,
A few months ago, I posted a question regarding a problem with
optimizing the molecules in a monoclinic unit cell (original message
below). After some helpfull comments (for which I am very grateful,
btw), I did some further testing on this matter. I now suspect that
there is a bug in the CPMD code (both in 3.7.2 and in 3.9.1)! The
problem apparently lies in the treatment of the CELL keyword, as I will
try to discuss.
I first started out with the definition for my molecular system using
the keywords SYMMETRY and CELL:
&SYSTEM
ANGSTROM
SYMMETRY
12
CELL
5.0872 2.314338 1.073321 0.0 -0.374493314 0.0
As Bala Sundaram pointed out to me, this specific format (as given in
the manual) was incorrect and should be replaced by:
&SYSTEM
ANGSTROM
SYMMETRY
12
CELL
5.0872 2.314338 1.073321 -0.374493314 0.0 0.0
That is, the format of the CELL keyword for a monoclinic system is:
a b/a c/a cos(beta) 0 0
A notification of this format is also implemented in version 3.9.1 of
cpmd, although I think some mention in the manual would be helpfull.
Anyhow, the calculation still failed and the reason is quite apparent in
the output:
************************** SUPERCELL ***************************
SYMMETRY: MONOCLINIC
LATTICE CONSTANT(a.u.): 9.61341
CELL DIMENSION: 9.6134 2.3143 1.0733 -0.3745 0.0000 0.0000
VOLUME(OMEGA IN BOHR^3): 2046.33461
LATTICE VECTOR A1(BOHR): 9.6134 0.0000 0.0000
LATTICE VECTOR A2(BOHR): -8.3320 20.6296 0.0000
LATTICE VECTOR A3(BOHR): 0.0000 0.0000 10.3183
As the lattice vectors point out, the a and c axis are perpendicular!!
This is obviously not in accordance with the monoclinic symmetry!
As this did not solve my problem, I redid my calculation, this time with
the CELL VECTORS keyword:
&SYSTEM
ANGSTROM
CELL VECTORS
5.087 0.0 0.0
0.0 11.773 0.0
-2.04447 0.0 5.062781
This calculation was succesfull, but now in the output it appears that
my molecular symmetry is TRICLINIC!!
************************** SUPERCELL ***************************
SYMMETRY: TRICLINIC
LATTICE CONSTANT(a.u.): 9.61304
CELL DIMENSION: 9.6130 2.3143 1.0733 0.0000 -0.3744 0.0000
VOLUME(OMEGA IN BOHR^3): 2046.13616
LATTICE VECTOR A1(BOHR): 9.6130 0.0000 0.0000
LATTICE VECTOR A2(BOHR): 0.0000 22.2477 0.0000
LATTICE VECTOR A3(BOHR): -3.8635 0.0000 9.5673
Eventually, my calculations succeeded but I hope you agree that there
are certain issues with regard to the SYMMETRY and CELL keywords that
need to be looked at.
Sincerely,
Ewald Pauwels.
Ewald Pauwels wrote:
> Dear all,
>
> Does anybody have any experience with applying CPMD on monoclinic
> crystals?
> The system I am studying contains four (organic) molecules in the unit
> cell. When I optimize these molecules (only the atomic coordinates -
> not the cell constants), I find that the unit cell no longer satisfies
> the original symmetry. I.e. all four molecules have a substantially
> different conformation!
>
> I suspect that I must be doing something wrong here. Below is my
> inputfile.
>
>
> Thanks in advance,
> Ewald.
>
>
>
> -------------------------------------------------------
> inputfile
> -------------------------------------------------------
> &CPMD
> RESTART LATEST GEOFILE WAVEFUNCTION
> OPTIMIZE GEOMETRY
> LSD
> ODIIS NO_RESET=50
> 10
> TIMESTEP
> 20
> PRINT
> 10
> STRUCTURE BONDS ANGLES
> &END
>
> &SYSTEM
> ANGSTROM
> SYMMETRY
> 12
> CELL
> 5.0872 2.314338 1.073321 0.0 -0.374493314 0.0
> CUTOFF
> 25.
> MULTIPLICITY
> 1
> &END
>
> &ATOMS
> *C_VDB_U BINARY NEWF
> LMAX=P
> 8
> 0.218182218417937 1.473520453 0.333090471110274
> 0.732546560623109 1.717103823 -1.08374897292691
> 1.30308357695184 7.360020453 2.1982999039374
> 0.788719234746671 7.603603823 3.61513934797459
> -0.218182218417937 -1.473520453 -0.333090471110274
> -0.732546560623109 -1.717103823 1.08374897292691
> 1.73944801378772 4.412979547 2.86448084615795
> 2.25381235599289 4.169396177 1.44764140212077
> *N_VDB_U BINARY NEWF
> LMAX=P
> 4
> 2.03113800324446 1.040509513 -1.31410549705625
> -0.509872207874679 6.927009513 3.84549587210393
> -2.03113800324446 -1.040509513 1.31410549705625
> 3.55240379861424 4.845990487 1.21728487799143
> *O_VDB_U BINARY NEWF
> LMAX=P
> 8
> 1.05333382374583 1.104201443 1.19695275049904
> -1.00750738194352 1.676251513 0.53944941448416
> 0.467931971623952 6.990701443 1.33443762454863
> 2.5287731773133 7.562751513 1.99194096056352
> -1.05333382374583 -1.104201443 -1.19695275049904
> 1.00750738194352 -1.676251513 -0.53944941448416
> 2.57459961911561 4.782298557 3.72834312554672
> 0.513758413426261 4.210248487 3.07083978953184
> *H_VDB_U BINARY NEWF
> LMAX=S
> 20
> 2.33817143317496 1.18130282 -2.29299415508719
> 2.75812420631155 1.37791192 -0.67745575495101
> 1.93653355901116 0.03473035 -1.14763114043162
> 0.875530385150063 2.77889892 -1.22357285168305
> 0.0131007466085907 1.35318862 -1.81758891709173
> -0.816905637805185 7.06780282 4.82438453013487
> -1.23685841094177 7.26441192 3.20884612999869
> -0.415267763641377 5.92123035 3.67902151547929
> 0.645735410219716 8.66539892 3.75496322673072
> 1.50816504876119 7.23968862 4.34897929213941
> -2.33817143317496 -1.18130282 2.29299415508719
> -2.75812420631155 -1.37791192 0.67745575495101
> -1.93653355901116 -0.03473035 1.14763114043162
> -0.875530385150063 -2.77889892 1.22357285168305
> -0.0131007466085907 -1.35318862 1.81758891709173
> 3.85943722854474 4.70519718 0.23839621996049
> 4.27939000168133 4.50858808 1.85393462009667
> 3.45779935438094 5.85176965 1.38375923461606
> 2.39679618051984 3.10760108 1.30781752336463
> 1.53436654197837 4.53331138 0.713801457955944
> &END
>
> &DFT
> LDA CORRELATION PZ
> FUNCTIONAL BP
> GC-CUTOFF
> 5.D-5
> &END
> ------------------------------------------------------------------
>
>
>
>
>
>
--
Ewald Pauwels
Centre for Molecular Modelling
Laboratory of Theoretical Physics
Ghent University
Proeftuinstraat 86
B-9000 Gent
Belgium
ewald.pauwels at UGent.be
+32 9 264 65 76
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cpmd.org/pipermail/cpmd-list/attachments/20041206/40810341/attachment.html
More information about the CPMD-list
mailing list