Neutron School: NPDF experiment
X. Qiu, Th. Proffen, J.F. Mitchell and S.J.L. Billinge,
Orbital correlations in the pseudo-cubic O and rhombohedral R phases of LaMnO3,
Phys. Rev. Lett. 94, 177203 (2005)
Presentations: Bragg talk - Diffuse talk - Modeling overview
Manuals: PDFgui - PDFgetN
Experiment: Experiment report - Archive with PDF data files
As mentioned in the lecture, LaMnO3 shows a Jahn-Teller distortion, which we can see structurally as a Mn-O long bond. Crystallographically, this long bond disappears at high temperature, and the goal of this experiment is to study the local behavior of the Mn-O long bond. Refinements are carried out using the program PDFgui (http://www.diffpy.org) which is installed on your computer. Here are the steps through this exercise:
The first step is to carry out a refinement of one of the datasets, e.g. the one measured at room temperature. Section 3.1 of the PDFgui users guide gives an example of how to set up a refinement.
|Structural parameters for LaMnO3|
|Reference||Proffen et al., Physical Review B60, 9973-9977 (1999).|
|Lattice parameters||a=5.732 Å, b=7.6832 Å, c=5.542 Å|
|Atom coordinates||Mn (0.50, 0.00, 0.00)|
La (0.0501, 0.25, -0.0068)
O (0.4873, 0.25, 0.0746)
O (0.304, 0.039, 0.7243)
If you are out of time, or want to see the solution, you may download lmo_rt.ddp, and load the file into PDFgui via File -> Open Project. In either case, converge the refinement to a minimum before proceeding to the next step.
Once the room temperature refinement is successfully converged, it is very simple to refine all the other temperatures as described in section 3.3.3 of the PDFgui users guide. Simply highlight the fit and select Fits -> Macros -> Temperature series from the menu. Click on Add to add all remaining data sets. Then highlight all the fits and click the refine button. This might be a good time for a cup of coffee !
Try it yourself. In case you are short on time, you might download lmo_temp.ddp, and load the file into PDFgui via File -> Open Project. This refinement was carried out from 1.7 Å to 19.5 Å.
Calculating bond lengths
One of the properties of interest is the distribution of Mn-O nearest neighbor bond lengths. Highlight one of the structure entries and select Phases -> Calculate bond lengths from the menu. Select Mn and O from the drop down menus and set the range to 1.0 to 2.5 Å. Note that there are three distinct Mn-O bond lengths. Make a spreadsheet with temperature and these three bond lengths. What do you observe ?
The plot on the right shows the resulting Mn-O bond lengths as a function of temperature obtained from Rietveld refinements. Clearly the Mn-O long bond disappears at high temperature in the average structure. Depending on the refinement range chosen for your PDF refinements above, a different Mn-O bond length behavior can be observed.
More systematic investigations of the Mn-O bond lengths as functions of refinement length scale in the structure are needed. PDFgui allows to set up refinements as function of r very similar to the setup of refinements as function of temperature. More details can be found in section 3.3.2 of the PDFgui users guide. Simply highlight the fit and select Fits -> Macros -> r-Series from the menu. Set the range to be 5 Å wide and start at 1.7 Å. Ideally one would do this for every temperature, but for the purpose of this exercise, choose two temperatures at the boundaries, e.g. room temperature and 1100 K. Once all the refinements are set up, click the run button and have some more coffee! Once finished, look at the resulting Mn-O bond lengths as a function of refinement range and temperature. Results for T=300K and T=1100 K for three different refinement ranges are shown below.
What next ?
If you made it all the way to the end - congratulations. Feel free to browse the list of publications to see the many applications of the PDF technique. If you are ever planning to carry out your own PDF experiment, feel free to contact me at email@example.com.
Artwork by Julie Coxe
NPDF summer student, 2007 & 2008