Encapsulating graphene with hBN

Hi all,

Up to now, I’ve done simulations with graphene deposited on hBN.

To define my potentials, I used the following (1 is C, 2 is B and 3 is N):

pair_style hybrid/overlay rebo tersoff kolmogorov/crespi/z 22.6 kolmogorov/crespi/z 22.6

pair_coeff * * rebo CH.airebo C NULL NULL

pair_coeff * * tersoff BNC.tersoff NULL B N

pair_coeff 1 2 kolmogorov/crespi/z 1 CB.KC C B NULL

pair_coeff 1 3 kolmogorov/crespi/z 2 CN.KC C NULL N

Now, when adding an additional layer of hBN to encapsulate graphene, I have two options.

I did the first, where I just define the atoms from the top layer of hBN using the same ID’s 2 and 3 again.

However, I am afraid that in this case, the Tersoff potential, which seems to be still sizeable for distances of 7 Ang, will make the top layer interact with the bottom layer. Interlayer interactions are supposed to be captured by my KC potential, and not by the Tersoff potential.

Or should I not be worried about this long-range effect?

So I tried setting up the top layer using different group IDs (4 and 5).

I am having a hard time making this work, as tersoff only allows for one line of pair_coeff and one must use * *, as far as I understand the manual.

pair_style hybrid/overlay rebo tersoff kolmogorov/crespi/z 22.6 kolmogorov/crespi/z 22.6 kolmogorov/crespi/z 22.6 kolmogorov/crespi/z 22.6

pair_coeff * * rebo CH.airebo C NULL NULL NULL NULL

pair_coeff * * tersoff BNC.tersoff NULL B N B N

pair_coeff 1 2 kolmogorov/crespi/z 1 CB.KC C B NULL NULL NULL

pair_coeff 1 3 kolmogorov/crespi/z 2 CN.KC C NULL N NULL NULL

pair_coeff 1 4 kolmogorov/crespi/z 3 CB.KC C NULL NULL B NULL

pair_coeff 1 5 kolmogorov/crespi/z 4 CN.KC C NULL NULL NULL N

This runs, but I don’t think it solves my issue above.

Is the only way to make it work is by modifying my existing BNC.tersoff file?

Like changing

N B B 3.0 1.0 0.0 25000 4.3484 -0.89000 0.72751 1.25724e-7 2.199 340.00 1.95 0.05 3.568 1380.0

to something like

N1 B1 B1 3.0 1.0 0.0 25000 4.3484 -0.89000 0.72751 1.25724e-7 2.199 340.00 1.95 0.05 3.568 1380.0

N2 B2 B2 3.0 1.0 0.0 25000 4.3484 -0.89000 0.72751 1.25724e-7 2.199 340.00 1.95 0.05 3.568 1380.0

where I defined my pair_coeff as

pair_coeff * * tersoff BNC.tersoff NULL B1 N1 B2 N2

Unrelated question, is there a way to not print the whole input file to the output? I could only make the print command work to add output rather than remove output.

Thanks in advance,
Nicolas

Hi Nicolas,

I think you can do it in the first method you mentioned. B-N Tersoff is short ranged, with a cutoff distance of about 2 angstrom. There is definitely no interactions between the top and bottom BN layers. The only thing you need to worry about is the boundary conditions perpendicular to the layers. You should either use open boundary conditions or periodic boundary conditions with a proper “vacuum layer”.

Bruce

[...]

Or should I not be worried about this long-range effect?

So I tried setting up the top layer using different group IDs (4 and 5).

I am having a hard time making this work, as tersoff only allows for one line of pair_coeff and one must use * *, as far as I understand the manual.

yes, but you can specify tersoff multiple times on the pair_style line
(same as you do for the inter-layer potential) and that way you can
decouple the two tersoff segments completely.

pair_style hybrid/overlay rebo tersoff kolmogorov/crespi/z 22.6 kolmogorov/crespi/z 22.6 kolmogorov/crespi/z 22.6 kolmogorov/crespi/z 22.6

[...]

Is the only way to make it work is by modifying my existing BNC.tersoff file?

don't mess with the potential file. that is almost always causing
trouble, since the relations between different lines of parameters are
not as obvious as most people think.

axel.

Hi Bruce and Axel,

Thanks for such prompt response.

I’ll do it like that then.

What triggered my worry is that the total energy for two different alignments between top and bottom layer of hBN was giving such a small difference that maybe some locking was happening between them. Will look more into it.

Might also try Axel’s suggestion to gain the experience.

Won’t modify the potential file.

Regards,
Nicolas