Negative coordinates in POSCAR??

Message

helvetico · #1 Post by **helvetico** » Tue Nov 13, 2007 4:28 pm

I am new to VASP and now I am learning the programm with different calculations.. One of them is an ordinary optimization of an ethanol molecule.

The problem: I built the molecule with Molden and saved a xyz file. In this file, some of the coordinates are negative. (The values are at the end of this post). After the calculation I find all the atoms that had negative values shifted by the optimized cell parameters! With vmd and the pbctools it is not possible to wrap the atoms since they are in principle within the cell!

Are there settings in VASP to avoid this problem?

Initial POSCAR file (C, O, H):
Ethanol
1.00000
10.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 8.000000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 8.0000000000000000
2 1 6
Cartesian
0.000000 0.000000 0.000000
0.000000 0.000000 1.450000
1.319933 0.000000 -0.466667
-1.026719 0.000000 1.813000
0.513360 -0.889165 1.813000
0.513360 0.889165 1.813000
-0.513360 -0.889165 -0.363000
-0.513360 0.889165 -0.363000
1.319933 0.000000 -1.416667

Positions from OUTCAR file after the last ionic step:

POSITION
---------------------------------------------
-0.00891 0.00000 -0.02252
-0.00617 0.00000 1.48263
1.33142 0.00000 7.59274
8.91863 0.00000 1.84016
0.51356 6.99214 1.86341
0.51356 0.89037 1.86341
9.38945 6.98996 7.64872
9.38945 0.89255 7.64872
1.32361 0.00000 6.61408

The new cell parameters:
9.9476751895603463 0.0000000000000000 -0.0228153758958300
0.0000000000000000 7.8825078749918819 0.0000000000000000
-0.0183354070361986 0.0000000000000000 8.0714040759236098

#2 Post by **admin** » Wed Nov 14, 2007 9:20 am

supert · #3 Post by **supert** » Mon Mar 03, 2008 8:59 am

I think it may be better to set all the coordinates positive, by adding a constant to them.