Contents of Chapter.
Direct Entrance of Electrons into the Detector. The Si(Li) detector is capable of responding to an energetic electron that enters the active region of the detector. A pulse is developed, the height of which is a measure of the energy of the electron. When the beam electrons strike the sample, a significant fraction, about 30% for copper with a beam normally incident, are backscattered with a wide energy range. Many of these backscattered electrons retain a substantial fraction of the incident energy. It is inevitable that some of those electrons will be scattered in the direction of the detector. The beryllium window with a typical thickness of 7.6 µm (0.3 mil) is capable of stopping electrons with an energy below about 25 keV. Above energies of 25 keV, electrons will begin to penetrate the window and activate the detector, although with a loss of energy due to inelastic scattering in the beryllium. When higher-energy (> 25 keV) beams are employed, the electrons that enter the detector can have a substantial effect on the background. An example is shown in the following figure for a 40-keV beam incident on arsenic. The background below 20 keV is greatly distorted by the electron contribution added to the expected x-ray continuum shown by the solid curve. Above 25 keV, the background is mostly due to the normal x-ray continuum; virtually no electrons are able to penetrate the beryllium window and retain energies in this range. Note that the analyst usually examines the region between 0 and 10 keV or 0 and 20 keV. In the example of the following figure, it may not be obvious that the background is anomalous unless the entire spectrum is examined.
Artifacts arising from scattered electrons entering the detector can be eliminated with magnetic shielding in front of the detector snout. In the windowless or thin window variety of Si(Li) spectrometers, such shielding is an absolute necessity. The artifact may become more pronounced for samples with a high atomic number or with surfaces highly tilted relative to the beam. These two conditions will produce the greatest number of high-energy backscattered electrons. Light element targets that are flat produce a relatively minor effect. Operation with an acceleration voltage below 20 kV will also minimize the effect.
Figure 1. EDS spectrum of arsenic excited by a 40-keV electron beam showing an anomalous background below 25 keV due to direct entry of electrons scattered from the specimen into the detector. The x-ray continuum was generated, then scaled using channels from the high-energy end of the spectrum.
The following is an example of a quantitative analysis of a copper gold specimen using pure copper and pure gold as standards. The beam voltage is 20.5 keV. The required information has all been entered into the Experiment Header and the Specimen Header. For standards, the compositions must be entered in the composition window of the Spectrum Header.
Set-up a ROI for the CuKa peak as shown in Figure 1.
Figure 1.
Set-up a Simplex Fitting procedure for CuKa and do the fit. The results are shown in Figure 2.
Figure 2.
Select "Make ZAF Standards" under the Analysis Menu as shown in Figure 3.
Figure 3.
This opens the dialog box shown in Figure 4. Select "Run ZAF on Work Fit-Results".
Figure 4.
The box shown in Figure 5 will appear if the "This is a Standard" checkbox is not checked in the Spectrum Header.
Figure 5.
If the analysis is allowed to continue, the standard results are displayed as in Figure 6.
Figure 6.
If the results are acceptable, they must be stored in a File of Standards. You may either create a new file (see Figure 7) or open an old file by selecting the appropriate button shown in Figure 4. You MUST then Add the Standard to the File (see Figure 4). Once open, it is not necessary to reopen the file of standards to add subsequent standards to the file - simply "Add the Standard to the File".
Figure 7.
Repeat the process for the pure gold standard (see Figure 8). Fill in the headers, Set-up the fitting procedure and fit the AuL peaks.
Figure 8.
Run the ZAF on the spectrum as in Figures 4, 5 and 6. The results are again displayed as shown in Figure 9.
Figure 9.
Then simply "Add standard to the file" (see Figure 10). This will add the standard to the file previously opened.
Figure 10.
This completes the standardization process for this analysis. More standards may be added to this file at any time.
The spectrum of the unknown may then be read into the WORK Spectrum and ROIs selected for the two peaks (see Figure 11). These ROIs should include the same peaks as were measured on the standards, although they do not need to be exactly the same since all calculations are done on a per channel basis. Then set-up the Simplex fitting for both the Cu K and Au L x-ray peaks.
Figure 11.
ZAF is then selected from the Analysis Menu as shown in Figure 12.
Figure 12.
The Main ZAF Quantitation window (see Figure 13) should now be on the screen. You may choose to normalize or not normalize your results. If you have saved a Set-up of a similar analysis, you may use that Set-up file, otherwise, select "Set-up an Analysis". You may also elect to quantitate an already fitted spectrum by selecting "Old Fit File". If you do this, select the "Old Fit File"as the last operation before closing the window.
Fugure 13.
The "Select Standard" window opens (bottom of Figure 14). Select the "Open File" button and select the file of standards you wish to use from the dialog box.
Figure 14.
Selecting the copper standard in the lower part of Figure 15 displays the elements and lines stored for that standard. Since the standard was pure copper, only the CuKa1 line appears in the window. Click the check box for copper followed by "Done" from the upper window.
Figure 15.
Repeat the operation for the gold standard.
When standards for all analyzed lines have been selected, click the "Done" button in the lower window of Figure 15. This will bring up the "Unanalyzed Element" window shown in Figure 16. One unanalyzed element may be selected here if necessary. Since the example has measurements for both elements in the specimen, leave the "No" button clicked and click "OK".
Figure 16.
This completes the ZAF analysis Set-up. Figure 13 should again be on the screen. Before leaving this window, you may save the analysis Set-up if you will be using it again at a later time. Also, a file of the ZAF results together with the correction factors used may be saved by clicking the "Full ZAF Report File?" button. Clicking the "Accept" button signals the program to run the ZAF analysis at the conclusion of each fit of a spectrum.
With an unknown spectrum in WORK and the ROIs set for the peaks to be fit, click "Do a Fit". The spectrum is fitted as shown in Figure 17, and after a slight pause, the analysis is complete.
Figure 17.
The "Add Fit" button adds the fitting data and the concentration to the results file. If a spreadsheet file has also been requested, the compositions are also included in that file. To see the results file, click "See Fit" and the results will be displayed as shown in Figure 18.
Figure 18.
If a file of spectra is to be processed, this will all be done automatically and the results can be viewed at the end of the calculations.
The ZAF procedure may also write an output file containing all the correction factors as well as the results. This file can be examined after exiting the DTSA program (see Figure 19).
Figure 19.