Original Calculations X0494

The next step is to produce an original ELF plot on one of our synthesized compounds - [C3H12N2][V2TeO8] (x0494). Two sets of calculations will be performed, one with the organic cations, one without, to observe the effect of the amine on the lone pair localization.

The CTRL file was created using lminit.run. x0494 belongs to space group #14. Coordinates from the CIF were given for the space group p2_1/n. Inputting space group #14 during lminit.run yields the space group p2_1/c instead. Instead of typing in the space group number, type in the actual space group - p2_1/n. You can then use the coordinates and unit cell parameters found in the CIF. Using the p2_1/c unit cell contains the amines in the middle of the unit cell and the inorganic layers along the edges. Since we are interested in the inorganic layers, we want those to be centered instead, as it is using the p2_1/n unit cell.

Run calculations using the p2_1/n unit cell. Everything for the structure excluding the amines runs smoothly following the steps described in the tutorial. (RMAXS in Category STR had to be increased to 6 as with the SnWO4 structure). The CTRL file for the structure with the amine does not converge, so OMMAX1 and OMMAX2 in Category SCALE need to be increased (.18 .20 .22 and .45 .50 .55). RMAXS needs to be increased to 6 as well. Calculations can then be run.

IMPORTANT Before creating your initial CTRL file, make sure you are using coordinates and parameters for the unit cell that you want. Use Atoms first to choose the unit cell you want, and input those coordinates and parameters into lminit.run. When you create your CTRL file, you are stuck with that unit cell. In other words, when you open your files in DataExplorer, you will only see a single unit cell. You cannot choose to show multiple unit cells, alternate unit cells, or an extended view of the crystal lattice. What you put into the CTRL file is exactly what you see, so choose your unit cell wisely. A good self-check: create your dx structure file (cstruc.dx) before running lmstr.run. See what that structure looks like in DataExplorer. If you don't like what you see, you will need to use a different unit cell.

When generating the structure file, some bonds weren't drawn. You will have to go back to the structure file generation process to increase maximum allowed bond lengths in order to draw in those bonds.

Once the calculations are finished (use of a 108 x 108 x 163 grid takes about 2 days) and you view the ELF plot in DataExplorer by the process described in the SnWO4 plot replicaiton, you can clearly see the lone pair lobes around the tellurium atoms (set Isosurface value ~75). You will, however, notice other highly localized regions that are relatively isolated. These are artifacts of the basis set that we have chosen, reflective of the void spaces in our cell not occupied by atomic or empty spheres. To better illustrate this, you can run another set of calculations using different empty sphere radii and positions:

Take your CTRL file with the empty spheres already in place. Delete all of the "odd" empty sphere sites (ie E1, E3, E5…). Additionally, lower all the radii of the "even" empty sphere sites by ~.2 A. Rerun lmes.run - this will place new empty spheres in the void space between all atomic spheres and the remaining empty spheres. Compare the resulting CTRL file with the CTRL with the original empty spheres to make sure the new empty spheres are sufficiently different. Notice how the coordinates of the new spheres are shifted slightly - up to a .01 change. Since these coordinates are fractions of the unit cell, a .01 change corresponds to a ~.1 A shift in position. These shifts, plus the lowered radii of the "even" empty spheres provide a sufficient alteration to observe the fact that these small localized regions in the ELF plot are in fact artifacts. Notice in the altered plot that qualitatively, the lone pair lobes are the same (the shape is slightly different, which is to be expected), but the other localized regions have shifted.

You will also notice highly localized regions around the cations in the structure that includes the amines. This is to be expected - these are isolated molecules that are covalently bonding whose electrons aren't spread through an entire lattice, as they are in the inorganic layer. To better visualize the ELF of the organic compounds, see the How to make ELF plots of molecules with Gaussian page.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License