Question about hydrogenation in builder 12.2.1

[BlackBarrel]

Dear Quantumwise staff:

    Recently I'm trying to learn the new builder, it looks cool but does have some defect. I have scanned the tutorials with no answers, or maybe I just missed them...
    First, where can I choose the form of hydrogenation? I'm curious about silicene, in which the hydrogenation is probably sp2 like and only edge atoms get hydrogenated, but the hydrogenate button always gives an sp3 form and make every Si atom connected with at least an H.
    Second, edge Boron in h-BN can't be hydrogenated, but Nitride is no problem.
    Finally, how can I add a single atom in the builder without directly modifying the python script? I hope there is some way to add atoms and make their distance to edge reasonable. Imagine the edge of GNR is functionalized by atoms with a variety of bond lengths. Such system is hard to build using current builder.
    Last but not least, is there anyway to reset camera to xy, yz and xz plane? And may I save current work place as a whole project and reload all needed parts in the stash at a time?

    The new builder is really cool in the selection function and many other things. Tags are convenient. I have used Materials Studio and it has a really strong builder system, you can learn something from it. I think everything will be better in ATK:)

[Nordland]

I can answer most of your questions. I hope

    Finally, how can I add a single atom in the builder without directly modifying the python script? I hope there is some way to add atoms and make their distance to edge reasonable. Imagine the edge of GNR is functionalized by atoms with a variety of bond lengths. Such system is hard to build using current builder.

You can add atoms in a couple of ways, but what would suite you best, is to add a new atom through right clicking in the coordinate list.
You can then position the atom with the correct bond length using the z-matrix tool or the move tool.

    Last but not least, is there anyway to reset camera to xy, yz and xz plane?

I have seen that there is somebody that have written a plugin for setting the camera direction just like you ask.

And may I save current work place as a whole project and reload all needed parts in the stash at a time?

What I personally do, is that I select all the items in the stash, and select Save as, and then select a nc file. This call I call stash.nc.
Then when I continue, I just start by pressing add from file and selecting this file, and then you have all your configurations.

    The new builder is really cool in the selection function and many other things. Tags are convenient. I have used Materials Studio and it has a really strong builder system, you can learn something from it. I think everything will be better in ATK:)

It sure will

[BlackBarrel]

Thank you very much for your answer! I have some more questions about some other aspects:

1. I'm dealing with a system which contains Si and B, and I want to try the new Slater Koster module. However, it doesn't work. I inspect the reason and find that there is no B-Si interaction file (or something like that, I'm not quite familiar with SK method). Where can I download it?

2. Is it reasonable (or is there any precedent) to optimize the structure with semi-experimental method and calculate properties with DFT? Since SE method is usually much faster and optimization requires lower accuracy. I should do some test, but DFT optimization is too time-consuming for my system. Thus I have no way to compare it with the SE.

Any help will be appreciated!

[Anders Blom]

1. You can go to http://dftb.org and request to obtain their parameters. It's free for academics. Then you can use those with ATK.

2. It's completely system-specific, so there is no simple answer. The simple truth is that if your SE optimal geometry is close the DFT geometry, then your results are likely too. One way to check, at least, without running a full DFT optimization, is just to compute the DFT forces for the SE-optimized geometry. If they are small, then you are likely in a good situation.

[Anders Blom]

About your initial question about hydrogenation, it's not yet supported to choose the hybridization. The tool is quite clever and attempts to guess the hybridization, however, from the neighboring atoms. That's why graphene automatically is sp2-completed with H atoms, and a Si nanowire gets sp3. But in those cases you mention it can be hard for it to get it right automatically, and so in the future it would be nice to add functionality where one could specify the atom types. Also note that the best way to passivate is "as early as possible", before doing repetitions etc, then if you get something incorrect perhaps you just need to adjust 3-4 atoms by hand, rather than 10 or 100, in the already repeated structure.