• View
• Edit
• History
• Print

Tricks For Background Fitting

If your data is dirty, as in radial diffraction data in the diamond anvil cell, background fitting can become fairly tricky. Let's be clear, there is not miracle and I do not know any perfect method. Here are three methods you can plan with

• Global polynomial
• Individual polynomial
• Interpolated

Global Polynomial Background

Global polynomial backgrounds can be edited by opening the dataset window and choosing the Background function tab. You can add as many polynomial parameters as you wish and refine them one by one or all together.

If you dataset is clean, with an isotropic background, you will be able to refine your data this way. If your dataset is not fully isotropic, you will be able to get a first average adjustment. This first average adjustment will be improved by adding polynomial individual backgrounds.

Individual Polynomial Backgrounds

If your background is not constant for all spectra, you can add individual background parameters. To do so,

• the dataset window,
• choose the Datafiles tab,
• select all spectra,
• click on Add bkg par once for a degree 0 polynomial, click on Add bkg par twice for a degree 1 polynomial, click on Add bkg par three for a degree 2 polynomial...

Once background parameters exist, you will need to perform a refinement. Here is how I proceed:

• open the Analysis -> @@Parameter list menu item,
• click on Fix all parameters,
• click on Free backgrounds,
• in the actual parameter list, fix the global background parameters,
• in the actual parameter list, free the incident intensity,
• refine...

Usually, after a few trials and errors, it works.

Interpolated Background

If your data is dirty, as in radial diffraction data in the diamond anvil cell, background fitting can become fairly tricky. Interpolated background could be your solution. Usually, it is not recommended to do so in Rietveld refinements (you migh smooth out details that contain real data) but, sometimes, there is not other way.

To do so,

• Remove all polynomial background parameters (in the global dataset and for each individual set),
• In the dataset window, choose the Background function tab,
• Select the Interpolate background tab,
• Choose the number of points for interpolation and hit Set interpolation points manually,
• In the plotting window, select the Tools -> Edit interpolated background points to see the locations used for interpolation,
• Right click in the plot to add or remove points.

Once you are done, the background will be interpolated between the points you choose. If your data is fairly dirty, this can become very useful.