Dear Neh,

The manuscript contained in the following link might be helpful:

https://arxiv.org/pdf/1207.5088.pdf

Rafik

You can construct a cubic unit cell (box) like 10A*10A*10A (or even bigger) in VESTA. Then, put the CO2 molecule in the center of the box and run the calculation.

Dear Neha,

I have quickly run a calculation using my own code (CASTEP) with PBE and get a much shorter bond length, about 1.139 A. LDA gave 1.130 A and RSCAN gave 1.126 A. A quick literature search suggests the experimental bond length is 1.128 A, so these DFT results look pretty good. I haven't done any convergence checks or anything so this number is not "the answer" but it looks reasonable to me and I would expect VASP to get pretty much the same result. I chose a 12Ax12Ax12A simulation cell with a single k-point at (1/4, 1/4, 1/4), and an initial C-O distance of 2 A.

It's very difficult to give informed advice without seeing your initial and final geometries, and knowing more about your input files generally. Perhaps you started the C & O too far apart and the forces are small enough to satisfy your force convergence tolerance almost immediately? Or if you started your atoms nearly half a lattice constant apart then the forces would be practically zero by symmetry (e.g. if you had a 7A x 7A x 7A simulation cell).

I assume your cell is larger than 5A per side, but if not then it's also possible you're measuring the "wrong" C-O distance, and one of the periodic images is closer -- however, the numbers you quote suggest this could only happen if you had a very small simulation cell.

Hope that helps,

Phil Hasnip

(CASTEP developer)

I believe your input file is wrong, maybe your initial configuration is unrealistic or your simulation cell is too small.