Have you asked that question on the official accelrys forums?

https://community.accelrys.com/community/materials_studio

No I have not yet, but I will.

Thank you.

Dear Timur, dear Materials Studio user,

everything we need is given in Supplemental Information of this paper:

http://pubs.acs.org/doi/abs/10.1021/ja8096256

In fact, Table S2 contains all necessary information (and is only the generalization of the entries in the RCSR). Please, have a look. Then the following protocol should give the desired net:

a) choose a desired degree of interpenetration N

b) for convenience multiply all cell length by a factor of 10.

c) build the crystal in Materials Studio with the cell length of b) and the given space group of the table; important - choose origin 2!

d) place exactly one atom at the given fractional coordinates

e) now the decisive step: connect two atoms that do not belong to the same cell!

almost Done…

f) replace the vertices by your desired centroids (adamantanes, biphenyls, triptycenes…)

Please find attached an XSD file with a dia-c6 structure.

kind regards

Frank

Dear Dr. Hoffman,

I would like to thank you for all your help. I was able to make the 6 fold interpenetrated structure by following your instructions. Honestly at first I did not understand what do you mean by origin 2. I have done everything else but that origin 2 thing, and sure enough it did not work. It was placing 8 vertex in the unit cell. Then I have realized in the material studio there is an option says Option > Origin 1 right below where you input the space group. That took care of the problem and it placed 4 vertex in the unit cell.

My question now is what does that origin 2 does?

And since we have that reference paper which gives the space group based on degree of interpenetration it is easy for dia topology. How about if we want to do it with other topologies. How can we generalize it to other structures.

I will paste this comment to other platforms as well since it might help people who is following the discussion from only that platform.

Thank you again.

Best,

Timur.

Dear Timur,

some comments to your remaining questions:

1) The number of atoms in the unit cell do not differ, regardless of the choice of the origin of the cell (of course, you _see_ more vertices, but only because Materials Studio displays not half-atoms or quarter-atoms, when lying at faces or edges).

2) Origin-1 and Origin-2 can be, of course, transformed into each other. The reason, why O'Keeffe chooses Origin-2 is probably that the structure on which he carried out the topology analysis was solved with this chosen origin. The two main possibilities to choose the origin in this particular space group are: (a) Origin (0, 0, 0) lies at the site with symmetry elements 4-bar, 2, m or (b) Origin lies at 2/m.

3) If you place an atom at 0.75 0.25 0.25 in cell choice 1 then it lies at the site with 2/m (which is not desired), only in choice 2 the atom with 0.75 0.25 0.25 will lie on the desired position. Inversely, if you choose origin-1 then you have to put the atom at 0, 0, 0. It is, I think easier to connect the two atoms of neighboring cells, if you choose origin-2.

4) You ask what to do, if you want to generate catenated structures with other topologies: Well, exactly the same as for dia :-) Look at the RSCR for topologies for which catenated versions exist --> click the checkmark box "interpenetrating net" and hit simply search. For instance, I have chosen now nbo-3, the respective XSD file is attached.

http://rcsr.anu.edu.au/nets/nbo-c3

5) There is only one issue left: the _generalization_ (for an arbitrary degree of interpenetration) for the other topologies in the manner of Table S2 for dia. However, in the moment, I am not able to do this.

best wishes

Frank

Attachment Nbo-c3.xsd

Thank you a lot for all your helpful and educative discussion.

Regards,

Timur.

Frank Hoffmann

Dear Dr. Hoffmann,

I am very new in the simulation of 3D COFs. Your answer is very informative. However, I am unable to follow it completely. A step that you have mentioned as decisive step is related to connect two atoms that do not belong to the same cell. Would you like to let me know how to connect them?

Suvendu Karak ,

I do not remember the specific atoms; try all possible connections between two atoms that do not belong to the same cell. Of course, you have to display more than one cell... You can also download the XSD files and figure it out which concrete atoms are connected...

Frank Hoffmann

Dear Dr. Hoffmann,

Thank you very much for your kind note...