BNL Materials Studio User Group: 10 may 2002

We discussed file input and output; and demonstrated how diffraction can be simulated, and monitored while atom positions are moved. For an example structure we started with BaTiO3 from the structure library. One way to build a distorted structure is to create a supercell, and move or rotate selected atoms. This can be done using mouse shortcuts or by explicitly specifying new positions for atoms. The point was brought up as to how the program controls what symmetries you get when tilting atoms around. The answer is, that it's up to the user to determine how they'd like the symmetry to be relaxed. The symmetry finding tool can be a help here. For example, you could construct a supercell with a carefully defined distortion, then ask the program to find and/or impose the symmetry that describes it. Then, subsequent atom-pulling will carry symmetry copies along with the motion. Let me know if someone has a particular application that would make a good demonstration of this.

Diffraction simulation is part of the "Reflex" package, which includes Rietveld refinement for solving powder data, as well as a powder simulation function. First, call up the "Reflex, powder diffraction" dialog box. Examine the different forms, you can specify the radiation type, line broadening algorithms, diffracted angle range. There is a menu called "display" in which you specify how it shall give you a "chart" (the graph of the diffracted curve) and/or the "grid" (table of reflected intensities). Examples for the BaTiO3 example are shown below. The graph has two data curves, which was done by specifying "add" for the chart display (add = add another curve to the graph, not add the datat together).

These images were made by selecting the chart window in materials studio (or by selecting the entire data grid in the case of the table), using "copy" from the materials studio edit menu, and using "paste" in adobe illustrator. Generally, the objects in materials studio can be pasted into nearly any windows program, with the format dependend on the context. For example, if you "paste" table values into a word processor or text document, you get an unformatted list of values. Pasting into a spreadsheet puts the data into columns. From the program you're pasting into, look in the edit menu for a "paste special" command. For example, if you copy the x-ray graph and look under "paste special" in microsoft word, you see several choices, which allow you to obtain a bitmap, a graphic image in windows meta format (where you can edit the axis labels, etc), or a table of values for intensity and two-theta that were generated by the simulation.

Question, what is the best way to input structure data?

If your structure is already in a standard data format, like SHELX (*.res file), go ahead and use it that way. Materials studio reads many standard formats including protein database structure files. The SHELX format is a simple ascii file that lists cell parameters and symmetry, and then the unique atom positions, occupancy and thermal parameter. Looks like this:

TITL BaTiO3



CELL   0.0000   4.0100   4.0100   4.0100  90.0000  90.0000  90.0000



LATT 1



SYMM -X,-Y,Z



SYMM -X,Y,-Z



SYMM X,-Y,-Z



SYMM Z,X,Y



SYMM Z,-X,-Y



SYMM -Z,-X,Y



SYMM -Z,X,-Y



SYMM Y,Z,X



SYMM -Y,Z,-X



SYMM Y,-Z,-X



SYMM -Y,-Z,X



SYMM Y,X,-Z



SYMM -Y,-X,-Z



SYMM Y,-X,Z



SYMM -Y,X,Z



SYMM X,Z,-Y



SYMM -X,Z,Y



SYMM -X,-Z,-Y



SYMM X,-Z,Y



SYMM Z,Y,-X



SYMM Z,-Y,X



SYMM -Z,Y,X



SYMM -Z,-Y,-X



SFAC Ba  O Ti



TI1   3    0.50000    0.50000    0.50000    0.02083   0.00000



BA1   1    0.00000    0.00000    0.00000    0.02083   0.00000



O1    2    0.00000    0.50000    0.50000    0.06250   0.00000



END

or this, for the example of the superstructure where symmetry was removed and atoms were tweaked

TITL BaTiO3



CELL   0.0000   8.0200   4.0100   8.0200  90.0000  90.0000  90.0000



LATT -1



SFAC Ba  O Ti



TI1   3    0.00000    0.50000    0.50000    1.00000   0.00000



TI2   3    0.50000    0.50000    0.50000    1.00000   0.00000



TI3   3    0.00000    0.50000   -0.00000    1.00000   0.00000



TI4   3    0.50000    0.50000   -0.00000    1.00000   0.00000



BA5   1    0.25000    0.00000    0.25000    0.50000   0.00000



BA6   1    0.25000    0.00000    0.75000    0.50000   0.00000



BA7   1    0.75000    0.00000    0.75000    0.50000   0.00000



BA8   1    0.75000    0.00000    0.25000    0.50000   0.00000



O9    2    0.00000    0.00000    0.50000    1.00000   0.00000



O10   2    0.50000    0.00000    0.50000    1.00000   0.00000



O11   2    0.00000    0.00000   -0.00000    1.00000   0.00000



O12   2    0.50000    0.00000   -0.00000    1.00000   0.00000



O13   2    0.31362    0.41010    0.53580    1.00000   0.00000



O14   2    0.25000    0.50000   -0.00000    1.00000   0.00000



O15   2    0.81660    0.45120    0.50183    1.00000   0.00000



O16   2    0.75000    0.50000    0.00000    1.00000   0.00000



O17   2    0.00000    0.50000    0.25000    1.00000   0.00000



O18   2    0.50000    0.50000    0.25000    1.00000   0.00000



O19   2   -0.00000    0.50000    0.75000    1.00000   0.00000



O20   2    0.50000    0.50000    0.75000    1.00000   0.00000



END

An example application raised at the meeting was to dope a material with 5% impurities and then model the static disorder. (To model averaged disorder, by specifying a nonunit occupancy and/or mixtures of atoms on sites, search the help files under "occupancy".) One route to such structures is to build the substructure crystal; make a supercell and relax symmetry to P1; output the structure in SHELX format; write the unix script to randomly change the species on selected sites; read the SHELX file back in. You can then look at simulated Bragg diffraction from the resulting structure. It is also possible to run a molecular dynamics simulation to relax the structure, using the "Discover" module, and find out what moved and what the diffraction now looks like. Maybe we can do this next time.

10 may 2002	The basics: intro to the software, model building, basic questions
17 may 2002	Powder diffraction package, input/output of models and files
31 may 2002	Notes about charges and bonds; tables of bond lengths and angles
28 march 2003	Diffraction from interfaces; Cerius2 faulted and single crystals and diffraction

BNL Materials Studio User Group: 17 may 2002

Meeting summary:

Other questions:



	"build, symmetry, supercell" to create 2a,b,2c superstructure. Select "make P1" to remove the symmetry constraints.	Select the oxygens around a particular cation, and rotate relative (holding down shift and right-click is one way) to make the distorted structure.
BaTiO₃ cell from database.