Case Study 3: Satellite Subtraction

Spectra acquired using achromatic X-ray anodes include features in the photoelectric spectrum corresponding to minor resonance lines in the X-ray spectrum from the X-ray gun anode. Figure 1 shows a portion of a survey spectrum in which the smaller satellite feature to lower binding energy of the C 1s line is a result of a minor Mg X-ray resonance. The energy separation and relative intensity of satellite lines with respect to the primary X-ray resonance are well known and so, to a first approximation, these features can be removed from the electron spectrum. The method used to remove these satellite lines is to iteratively generate the satellite subtracted spectra using the convolution of a response function with the current iteration to partition the data into the primary spectrum and the satellite spectrum.

The satellite subtraction procedure requires the identification of each photoelectric line using quantification regions (the contribution to Auger lines are not separated in energy from the primary Auger spectrum and therefore should be excluded from the satellite procedure). These regions define which lines are to be included in the satellite subtraction algorithm and also define the background to the peaks. The background choice determines, in part, the shape of the primary line and a poorly defined background will impair the outcome of the satellite subtraction.

Figure 1: Data Editor property page on the Quantification Parameters dialog window.

Satellite subtraction does not fit well with the processing regime in CasaXPS. The reason being that the procedure uses the current definition of quantification regions in order to perform the subtraction. The set of regions used in satellite subtraction is typically different from the set of regions used to quantify a spectrum and so it is difficult to track the procedure, in a processing-history sense. As a result, satellite subtraction is performed on the Quantification Parameters dialog window using buttons found on the Data Editor property page. Here, spectra can be processed individually or collectively, where those spectra overlaid in the active tile are the set of spectra processed by the Satellite Subtract All button indicated in Figure 1. Spectra modified via the Data Editor property page may either be permanent (i.e. the raw data is edited) or temporary in the sense that resetting the processed data using the Processing History property page will undo the satellite subtraction. If the raw data is edited then the process cannot be reversed and an entry is written into the VAMAS comment for that spectrum indicating the data is satellite subtracted. If the raw data is modified it is recommended that the data is saved to a new filename rather than overwriting the original data file.

Figure 2: Same spectrum shown in Figure 1 after satellite subtraction.

Figure 2 shows the data in Figure 1 following satellite subtraction. Here the Satellite Subtract All button was used, where the resulting dialog box (Figure 3) indicates how many spectra will be processed and asks whether or not the raw data should be edited. Select the No button unless it is desired to permanently modify the data.

Summary

• Create quantification regions on each photoelectric line in a spectrum
• Display all spectra for which satellite subtraction is required into the active tile.
• On the Data Editor Property Page select Satellite Subtract All button.
• Answer the dialog question in Figure 3.
• If raw data was modified, save the spectra to a new file name.

Figure 3