Advanced Rietveld Techniques: Tricks and Tips with EXPGUI
This is a web/video version of a presentation made to the research group of John Mitchell at the Argonne Materials Science Division on February 22, 2006. It was originally planned to be released as a combined slide+video+audio (SofTV) recording of the presentation, but this did not come to pass due to incompatibilities between the Mac and Windows versions of PowerPoint. For this reason the slides have many more screen dumps than I would have done, if I had planned to provide recorded demos. I hope the extra detail is not a major hindrance.
Brian Toby, April 2006
Presentation Goal
- Introduce advanced & less obvious features of EXPGUI
- Format: Problems & Solutions
Prerequisites
- Crystallography
- Some experience with GSAS & EXPGUI
- Desire to learn new tricks & easier approaches
Outline
- Installing the software
- This section shows how GSAS, EXPGUI, CMPR & DRAWxtl (and GRACE) are obtained and the demo shows the installation of the software on a PC running Windows.
- How to get starting profile terms
-
Problem: Starting a refinement with bad profile coefficients & only an approximate structure is quite hard
Solution: In CMPR, fit peaks widths and then fit a profile function to the widths; then in EXPGUI write the profile function into an instrument parameter file
-
Problem: Starting a refinement with bad profile coefficients & only an approximate structure is quite hard
- Getting coordinates into GSAS
-
Problem: How do I get coordinates electronically into GSAS without retyping them?
Solution: EXPGUI can read a variety of coordinate formats, most importantly CIF; it can reset Origin 1 settings into Origin 2; atoms can also be displayed in an order other than the sequence they have been input.
-
Problem: How do I get coordinates electronically into GSAS without retyping them?
- Graphics options in EXPGUI
-
Problem: Are there good ways to graphically review my fit or structure?
Can
Solution: EXPGUI includes LIVEPLOT for review
-
Problem: Are there good ways to graphically review my fit or structure?
Can
- Simulation of diffraction patterns
-
Problem: How can I easily simulate a diffraction pattern, for example from a CIF?
Solution: Use EXPGUI to set up the simulation for GSAS.
-
Problem: How can I easily simulate a diffraction pattern, for example from a CIF?
- Bootstrapping the background
-
Problem: Fitting background is easy, except when the structure, profile
and initial background are all wrong (or when doing a Le Bail fit). How can I just put the background where I want it to be?
Solution: Use of fixed background points is not a great idea, but using such points to generate an initial background model that is later refined when the model is in good shape. BKGEDIT can be used to do exactly that.
-
Problem: Fitting background is easy, except when the structure, profile
and initial background are all wrong (or when doing a Le Bail fit). How can I just put the background where I want it to be?
- Returning to a prior refinement
-
Problem:Yesterday's Fit Was Better Than What I Have Done Since. How Do I Get It Back?
Solution:GSAS keeps a historical list of old experiment files. EXPGUI has a tool to browse them and load one.
-
Problem:Yesterday's Fit Was Better Than What I Have Done Since. How Do I Get It Back?
- Changing flags & settings for multiple histograms en mass
-
Problem: I have a 10 histogram refinement with 3 phases, but I don't want to have to edit 10 sets of background parameters or 30 sets of profile refinement flags. Can I set everything at the same time?
Solution: Yes - use multiple histogram selection.
-
Problem: I have a 10 histogram refinement with 3 phases, but I don't want to have to edit 10 sets of background parameters or 30 sets of profile refinement flags. Can I set everything at the same time?
- Creating and Viewing Fourier Maps
-
Problem: How do I create and analyze Fourier maps.
Solution: Fourier maps are set up in EXPEDT (not EXPGUI, sorry); GSAS has tools for finding peaks in the map, as well as plotting them; Another option is DRAWxtl, which plots structures and maps quite nicely.
-
Problem: How do I create and analyze Fourier maps.
- Constraining Atom Parameters
-
Problem: When I refine all the atomic parameters in my model,
the structure blows up. How do I deal with two atoms that share a site, etc.?
Solution: GSAS allows the number of refined parameters to be reduced by forcing the shifts on more than one parameter to be the same (or related by a ratio). EXPGUI offers some tools to make this easier to do, particularly when two atoms are placed at the same site and the fraction of the two atoms is to be varied while the total occupancy is to remain constant.
-
Problem: When I refine all the atomic parameters in my model,
the structure blows up. How do I deal with two atoms that share a site, etc.?
- Other Tips, Tools & Tricks
-
Problem: How do I learn about other obscure features in EXPGUI, without reading the tedious and humorless web-page documentation that
comes with the software?
Solution: You can learn some cute tricks here: controlling output in the .LST file; tricks in LSTVIEW, using detachable menus, benefits of the "Change Profile" window; the composition tool; editing excluded regions. Don't look for much humor here either.
-
Problem: How do I learn about other obscure features in EXPGUI, without reading the tedious and humorless web-page documentation that
comes with the software?
Links
Last modified: Thu Apr 6 10:07:23 CDT 2006