I-V curve of original GNR and GNR cut from graphene sheet

[jim12565]

Hello again, I have simulated a 10 atoms wide zigzag graphene nanoribbon built from the nanoribbon builder and then simulated the (same?) GNR cut from the graphene sheet. The I-V curve of the original GNR is sort of sensible as attached but the I-V curve of the GNR cut from the graphene sheet seems incorrect as attached. When I visualise each structure in the Viewer, they seem the same as you can also see in the attachment. What can cause this difference of their I-V curve being different, or I could ask if the GNR cut from the graphene sheet is not perfectly right for som reason? Cheers, Jim.

[Petr Khomyakov]

The two structures have very different lattice parameter in the B-direction. Note that periodic boundary conditions are imposed in the A and B directions. It means that the repeating nanoribbon images in one of the two structures are very close to each other, ~3.5 Ang, whereas the vacuum padding is much larger for the other nanoribbon structure.

[jim12565]

Thanks for your fast reply Petr. So as a general rule, should I increase the lattice's B vector (something like 2 times) whenever I cut nanoribbon from a sheet structure?

[Petr Khomyakov]

The general rule is that you have to increase the vacuum region until all the physical quantities of your interest are not affected by the interaction between the images within some tolerance given by your target accuracy for a particular quantity. I would not make it smaller than 10 Ang.