My Community

Hi there,

I am having trouble correctly implementing a POSCAR to represent a group of atoms. I want to represent a group of 14 (or more) atoms in an FCC structure. I have tried to do this in many ways, but am always getting errors and the OUTCAR gives co-ordinates that don't really make sense.

Can anyone give me an example of how to define a large number of atoms in a crystal structure?

thanks!

Hi Al,

could you send an example please?

Thx

alex

for example to make an fcc box with 14 atoms I am trying to do it like this, unit scaling and lattice vectors in X,Y,Z, then directly place the atoms where they should be. Still not sure about Cartesian or direct in this case, but I always get atoms overlapping on top of each other hence an error.

Al
1.0
4.05 0.0 0.0
0.0 4.05 0.0
0.0 0.0 4.05
14
Cartesian
0.0 0.0 0.0
0.5 0.0 0.5
0.5 0.5 0.0
0.0 0.5 0.5
1.0 1.0 1.0
1.0 0.0 1.0
1.0 1.0 0.0
0.0 1.0 1.0
1.0 0.0 0.0
0.0 1.0 0.0
0.0 0.0 1.0
1.0 0.5 0.5
0.5 1.0 0.5
0.5 0.5 1.0

after some playing around with visualisation tools, I managed to get closer to what I want. This POSCAR gives a cube of what looks like 4X full unit cells of fcc Al. This is achieved with 32 atoms as follows. Does this make sense to use as an extended solid?

Al FCC
2.0
4.05 0.0 0.0
0.0 4.05 0.0
0.0 0.0 4.05
Al
32
Direct
0.0 0.0 0.0 # 1
0.5 0.0 0.5 # 2
0.5 0.5 0.0 # 3
0.0 0.5 0.5 # 4
0.5 0.0 0.0 # 5
0.0 0.5 0.0 # 6
0.0 0.0 0.5 # 7
0.25 0.0 0.25 # 8
0.25 0.25 0.0 # 9
0.0 0.25 0.25 # 10
0.75 0.0 0.25 # 11
0.25 0.75 0.0 # 12
0.0 0.25 0.75 # 13
0.25 0.0 0.75 # 14
0.75 0.25 0.0 # 15
0.0 0.75 0.25 # 16
0.75 0.0 0.75 # 17
0.75 0.75 0.0 # 18
0.0 0.75 0.75 # 19
0.5 0.5 0.5 # 20
0.25 0.5 0.25 # 21
0.25 0.25 0.5 # 22
0.5 0.25 0.25 # 23
0.25 0.5 0.75 # 24
0.75 0.25 0.5 # 25
0.5 0.75 0.25 # 26
0.75 0.5 0.25 # 27
0.25 0.75 0.5 # 28
0.5 0.25 0.75 # 29
0.75 0.5 0.75 # 30
0.75 0.75 0.5 # 31
0.5 0.75 0.75 # 32

Hi Al,

your 1x1 should look like this

Cu
1.00000000000000
4.0500000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 4.0500000000000000 0.0000000000000000
0.0000000000000000 0.0000000000000000 4.0500000000000000
4
Direct
0.0000000000000000 0.0000000000000000 0.0000000000000000
0.0000000000000000 0.5000000000000000 0.5000000000000000
0.5000000000000000 0.0000000000000000 0.5000000000000000
0.5000000000000000 0.5000000000000000 0.0000000000000000

Cartesian does not make sense in your case. If you write it like that, it would be 'direct'.
There is no double occupation (like 0 0 0 and 1 1 1 is the _same_ position for fractional coordinates: translational symmetry!), too.

However, the 2x2 is alright! If you'd like to have a smaller super cell, you could go for multiples of the primitive fcc, which contains a single atom only.

Cheers,

alex