[xcrysden] How to display a whole molecule in unit cell with good connection?
liyincumt at gmail.com
liyincumt at gmail.com
Tue Oct 27 15:02:48 CET 2015
Dear Matic,
Thanks for your patient explanation!
There are only two very small negative values in the final coordinates. My crystal is a kind of amino acid. In one unit cell, there are two molecules. In my input file, there is no negative values in atomic coordinates, but the boundary still can apparently break the chemical bonds.
Input:
&CONTROL
title = '2015971-Asparticacid'
calculation = 'vc-relax'
restart_mode = 'from_scratch'
outdir = '/home/yinli/aspartic_acid/vc-relax/test1' ,
wfcdir = '/home/yinli/aspartic_acid/vc-relax/test1' ,
pseudo_dir = '/home/yinli/espresso-5.1.1/pseudo/' ,
prefix = '2015971-Asparticacid'
etot_conv_thr = 5.0d-7 ,
forc_conv_thr = 5.0d-4 ,
nstep = 1500
/
&SYSTEM
ibrav = -12
celldm(1) = 9.663114537963445
celldm(2) = 1.350523125061113, celldm(3) = 1.484795150092891
celldm(5) = -0.185014881113262
nat = 32
ntyp = 4
ecutwfc = 60
ecutrho = 600
occupations = 'fixed'
/
&ELECTRONS
electron_maxstep = 200
conv_thr = 1.0D-10
diago_thr_init = 1e-4
startingpot = 'atomic'
startingwfc = 'atomic'
mixing_mode = 'plain'
mixing_beta = 0.5
mixing_ndim = 8
diagonalization = 'david'
/
&IONS
ion_dynamics = 'bfgs'
/
&CELL
cell_dynamics = 'bfgs'
/
ATOMIC_SPECIES
H 1.00794 H.blyp-van_ak.UPF
C 12.01070 C.blyp-van_ak.UPF
N 14.00670 N.blyp-van_ak.UPF
O 15.99940 O.blyp-van_ak.UPF
ATOMIC_POSITIONS crystal
N 0.858300000000000 0.978000000000000 0.316200000000000
H 0.029000000000000 0.914000000000000 0.388000000000000
H 0.810000000000000 0.100000000000000 0.389000000000000
H 0.900000000000000 0.022000000000000 0.188000000000000
C 0.634800000000000 0.836000000000000 0.295200000000000
C 0.560500000000000 0.796000000000000 0.478600000000000
C 0.715600000000000 0.644000000000000 0.219800000000000
C 0.793000000000000 0.659000000000000 0.037800000000000
O 0.749700000000000 0.784000000000000 0.609500000000000
O 0.318400000000000 0.771000000000000 0.481700000000000
O 0.002000000000000 0.593000000000000 0.008500000000000
O 0.612500000000000 0.744000000000000 0.916100000000000
H 0.468000000000000 0.898000000000000 0.207000000000000
H 0.881000000000000 0.578000000000000 0.313000000000000
H 0.541000000000000 0.556000000000000 0.207000000000000
H 0.679000000000000 0.750000000000000 0.797000000000000
N 0.141700000000000 0.478000000000000 0.683800000000000
H 0.971000000000000 0.414000000000000 0.612000000000000
H 0.190000000000000 0.600000000000000 0.611000000000000
H 0.100000000000000 0.522000000000000 0.812000000000000
C 0.365200000000000 0.336000000000000 0.704800000000000
C 0.439500000000000 0.296000000000000 0.521400000000000
C 0.284400000000000 0.144000000000000 0.780200000000000
C 0.207000000000000 0.159000000000000 0.962200000000000
O 0.250300000000000 0.284000000000000 0.390500000000000
O 0.681600000000000 0.271000000000000 0.518300000000000
O 0.998000000000000 0.093000000000000 0.991500000000000
O 0.387500000000000 0.244000000000000 0.083900000000000
H 0.532000000000000 0.398000000000000 0.793000000000000
H 0.119000000000000 0.078000000000000 0.687000000000000
H 0.459000000000000 0.056000000000000 0.793000000000000
H 0.321000000000000 0.250000000000000 0.203000000000000
K_POINTS automatic
3 2 2 0 0 0
Dr. Yin Li
Department of Biophysics,Medical School, University of Pecs,
No.12 Szigeti Street, Pecs, H-7624, HUNGARY
Phone: +36-72-535271/36271
From: Matic Poberznik
Date: 2015-10-27 16:51
To: XCrySDen mailing list
Subject: Re: [xcrysden] How to display a whole molecule in unit cell with good connection?
>>Actually, I don't want to modify the relative distance between atoms and boundaries. I think journals >>would not accept this.
I think that modifying the relative distance, as far as the figure is concerned, should not be a problem. The only limitation I see is that all atoms defined in the ATOMIC_POSITIONS card should be translated by the same vector; due to periodicity both structures are equal, what will change is the way the program reads the structure. Without seeing the actual output I am assuming that your optimized coordinates contain atoms with negative values in coordinates, for example let's say the initial positions were 0.0, 0.0, 0.0, but after optimization they changed to -0.1,-0.1,-0.1, so assuming a cubic structure of length "1" the program draws the optimized coordinates at 0.9, 0.9, 0.9 (equivalent position in the crystal cell), the result being the molecule "breaks" in the unit cell. By shifiting all positions (to positive values) by a specific vector this problem should disappear and as stated both structures are equivalent, so this representation should be acceptable.
Both the xsf and pwscf formats are supported by openbabel, so converting it to other formats shouldn't be a problem, though I have no experience regarding the dynamic matrix, so I really have no idea if information will be lost or not.
Best regards,
Matic
On 10/27/2015 01:47 PM, liyincumt at gmail.com wrote:
Dear Matic,
Thank you very much for your help!
Actually, I don't want to modify the relative distance between atoms and boundaries. I think journals would not accept this. I was thinking on whether Molden could display it properly as the dynamic matrix file can be read by Molden as well. I tried many options in molden but it still can't work. So far I just found material studio (Display style: Default) can display it properly.
I wonder whether I can convert quantum espresso files including dynamic matrix file and pwo into cif or any other format which could be read by material studio. Even it can be converted, I am still worried about the loss of the information of the forces on atoms.
Best Regards,
Yin
Dr. Yin Li
Department of Biophysics,Medical School, University of Pecs,
No.12 Szigeti Street, Pecs, H-7624, HUNGARY
Phone: +36-72-535271/36271
From: Matic Poberznik
Date: 2015-10-27 14:27
To: XCrySDen mailing list
Subject: Re: [xcrysden] How to display a whole molecule in unit cell with good connection?
Modifying this in the pwscf output file is impossible, so I suggest you convert the output file to xsf format using the pwo2xsf.sh tool. It can be found in the Quantum Espresso installation directory in my case it is in
/esspreso-5.1.1/PW/tools/pwo2xsf.sh
To write the optimized coordinates to the xsf file the syntax would be :
bash pwo2xsf.sh -oc name_of_pwout.out > name_of_xsf.xsf
Once you have the xsf file I suggest you shift the coordinates of the atoms so that the molecule is in the center of the cell; If the molecule atoms are , for instance, at 0.0, 0.0, 0.0 you should shift them so that the coordinates correspond to 1/2a+1/2b+1/2c where a,b,c are vectors of the unit cell. You should shift all atomic positions accordingly and the result will hopefully be better.
Best regards,
Matic
On 10/27/2015 11:25 AM, liyincumt at gmail.com wrote:
Dear Matic,
Thanks for your kind help!
After setting the number of units drawn, some molecules inside the supercell are displayed correctly. But there are too many molecules inside the supercell. The figure would be hardly accepted by journal.
Is it possible to just display the molecules correctly in one unit cell? I have tried buttons of the display which are located in the bottom center of the screen, none of them can work properly.
Best Regards,
Yin
Dr. Yin Li
Department of Biophysics,Medical School, University of Pecs,
No.12 Szigeti Street, Pecs, H-7624, HUNGARY
Phone: +36-72-535271/36271
From: Matic Poberznik
Date: 2015-10-27 12:14
To: XCrySDen mailing list
Subject: Re: [xcrysden] How to display a whole molecule in unit cell with good connection?
Dear Yin,
If I understood your problem correctly, I think the best solution is to open the Modify menu and choose the Atomic Radius (Shift+R) option. Once there you can modify the Covalent Radius and Connectivity Factor for each atom type to suit your needs.
In order to load your custom parameters when xcrysden starts you should add a line in $HOME/.xcrysden/custom-definitions file, an example:
#
# this line changes the default atomic radius for aluminum to 1.6
#
set atmRad(13) 1.6
The custom definitions file is described on this link: http://www.xcrysden.org/doc/custom.html
If the problem stems from the periodic boundary conditions then first try the different display modes of unit cells, i. e. click the button for the display of the translational asymmetric unit (bottom center of the screen). Then try Modify => Number of units drawn (Shift+N), if the covalent radius is correct at least some molecules should be displayed correctly. If that is not enough let me know, there is one more workaround but it is a bit more complicated.
Best regards,
Matic
On 10/27/2015 09:19 AM, liyincumt at gmail.com wrote:
Dear Xcrysden developers,
Thanks for sharing this excellent visualization software for free!
I am using xcrysden to visualize the input and output files of quantum espresso. I have a problem with visualization of crystals. Due to the presence of periodic boundary condition, some of atoms which are outside of the boundary look like being non-covalently connected. Is it possible to display a whole molecule with good covalent connections in Xcrysden.
Thank you very much!
Best Regards,
Yin
Dr. Yin Li
Department of Biophysics,Medical School, University of Pecs,
No.12 Szigeti Street, Pecs, H-7624, HUNGARY
Phone: +36-72-535271/36271
_______________________________________________
XCrySDen mailing list
XCrySDen at democritos.it
http://www.democritos.it/mailman/listinfo/xcrysden
_______________________________________________
XCrySDen mailing list
XCrySDen at democritos.it
http://www.democritos.it/mailman/listinfo/xcrysden
--
Matic Poberznik
J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia
_______________________________________________
XCrySDen mailing list
XCrySDen at democritos.it
http://www.democritos.it/mailman/listinfo/xcrysden
--
Matic Poberznik
J. Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.democritos.it/pipermail/xcrysden/attachments/20151027/40bd63eb/attachment-0001.htm
More information about the XCrySDen
mailing list