How to understand two minimum energy paths for one diffusion?
Posted: Thu May 21, 2009 3:31 am
Dear all,
I used CI-NEB calcualtions to study interstitial diffusion in fcc iron. Two different initial interpolations were done:
1)one is to interpolate 11 images linearly between the two adjacent octahedral sites(from 0 0 0.5 to 0.5 0.5 0.5 );
2)the another one is to include tetrahedral site, first did a linearly interpolating between one O-site to the T site (from 00 0.5 to 0.25 0.25 0.25), we added five images between these two end points; secondly we did a linearly interpolating between the T site to another O-site (from 0.25 0.25 0.25 to 0.5 0.5 0.5), we added five images between these two end points. Then it turned out we interpolate 11 images which include a T-site between the two adjacent O-site(from 0 0 0.5 to 0.5 0.5 0.5)
We did the CI-NEB calculations for these two interpolation with different initial guess (which means O-O and O-T-O initial guess), the corresponding chains of images all converged. The energy results of the 13 images in the MEP are as follows:
For O-O: 0, 0.1, 0.39, 1.06, 1.82, 2.40, 2.60, 2.39, 1.82, 1.06, 0.38, 0.1, 0 (non spin)
For O-T-O : 0, 0.12, 0.52, 1.37, 2.12, 2.56, 2.65, 2.57, 2.14, 1.39, 0.54, 0.12, 0 (non spin)
For spin calculation:
O-O: 0, 0.1, 0.25, 0.60, 0.85, 0.93, 0.92, 0.93, 0.87, 0.62, 0.26, 0.1, 0
O-T-O: 0, 0.15, 0.35, 0.92, 1.51, 2.01, 2.19, 2.02, 1.54, 0.93, 0.37, 0.16, 0
I have two questions:
1)If all the calculation converged, does it mean the two MEP (O-O, O-T-O) are possible, or only the smaller diffusion energy one is possible?(that is for non spin, the O-O which has 2.60 is possible)?
2)For the spin calcualtion, there is so large difference between the O-O MEP and O-T-O MEP, I don't know how to understand this?
Any comments are welcomed, Thanks!
<span class='smallblacktext'>[ Edited ]</span>
I used CI-NEB calcualtions to study interstitial diffusion in fcc iron. Two different initial interpolations were done:
1)one is to interpolate 11 images linearly between the two adjacent octahedral sites(from 0 0 0.5 to 0.5 0.5 0.5 );
2)the another one is to include tetrahedral site, first did a linearly interpolating between one O-site to the T site (from 00 0.5 to 0.25 0.25 0.25), we added five images between these two end points; secondly we did a linearly interpolating between the T site to another O-site (from 0.25 0.25 0.25 to 0.5 0.5 0.5), we added five images between these two end points. Then it turned out we interpolate 11 images which include a T-site between the two adjacent O-site(from 0 0 0.5 to 0.5 0.5 0.5)
We did the CI-NEB calculations for these two interpolation with different initial guess (which means O-O and O-T-O initial guess), the corresponding chains of images all converged. The energy results of the 13 images in the MEP are as follows:
For O-O: 0, 0.1, 0.39, 1.06, 1.82, 2.40, 2.60, 2.39, 1.82, 1.06, 0.38, 0.1, 0 (non spin)
For O-T-O : 0, 0.12, 0.52, 1.37, 2.12, 2.56, 2.65, 2.57, 2.14, 1.39, 0.54, 0.12, 0 (non spin)
For spin calculation:
O-O: 0, 0.1, 0.25, 0.60, 0.85, 0.93, 0.92, 0.93, 0.87, 0.62, 0.26, 0.1, 0
O-T-O: 0, 0.15, 0.35, 0.92, 1.51, 2.01, 2.19, 2.02, 1.54, 0.93, 0.37, 0.16, 0
I have two questions:
1)If all the calculation converged, does it mean the two MEP (O-O, O-T-O) are possible, or only the smaller diffusion energy one is possible?(that is for non spin, the O-O which has 2.60 is possible)?
2)For the spin calcualtion, there is so large difference between the O-O MEP and O-T-O MEP, I don't know how to understand this?
Any comments are welcomed, Thanks!
<span class='smallblacktext'>[ Edited ]</span>