The selection "Analysis" in the Menu Bar provides access to a variety of quantitation procedures for thin biological specimens and materials specimens, and to manual peak stripping. Also, present are the commands for fitting an entire DTSA file, for peak identification in all spectra in the file, and for retiring the currently active quantitation procedure.
Contents of Chapter.
The DTSA ZAF quantitative analysis procedure consists of the ZAF matrix correction taken from FRAME with several significant modifications. The atomic number correction has been changed so that the backscatter loss R is now computed by a full integration. The backscatter coefficient is that of Love and Scott; the stopping power S is that of Pouchou; the cross-section is that of Fabre; and the energy distribution of backscattered electrons is that of Czyzewski. The absorption correction uses Heinrich's mass absorption coefficients and also his expression for f(chi). The characteristic fluorescence correction is Reed's. There is no continuum fluorescence correction.
The ZAF quantitation procedure is run on a WORK spectrum file in conjunction with either the MLLSQ or Simplex fitting procedure. The ZAF quantitation is run in two parts: first is the calculation of the standards and the creation of files of standards data, and second is the calculation of the compositions of unknowns. These two operations are executed by selection from the DTSA ZAF sub-menu. We will first describe the selections related to standards.
ZAF Standards files
The first two menu items are for creating, appending, and
displaying files of standards used in the "quantitative"
procedures that convert peak intensities obtained from the curve
fitting procedures into chemical compositions. ZAF Stds Flowchart
gives detailed information on how to make a ZAF Standard. More
information is provided in the section on ZAF quantitation.
The first step in the ZAF matrix correction procedure is to set up the standards that will be needed for the analysis. This is the laborious part of the analysis procedure. However, in many cases ZAF standards may be reused for other ZAF analyses provided the analysis conditions are the same (e.g., the beam voltage, the x-ray line for the element being analyzed, and the fitting procedure used to obtain the peak areas). The spectrum acquisition time (Live Time) and the beam current (Faraday Current) must be entered in the Spectrum Header since the data are normalized using these parameters. If no Faraday Current is entered, the program assumes that it is 1 nA. The steps required to make a ZAF standard are as follows:
1. Read the spectrum from a standard material into the WORK SPECTRUM.
2. Under the "Headers Menu", select "Experiment Header". Make sure that the beam voltage and the detector parameters are correct.
3. Then select "Spectrum Header". Enter the Begin current, End current, and Acq. Time (live time). Also enter a label for the ZAF standard as the first information in the spectrum Comments. This label should clearly differentiate the new ZAF standard from others you might have for the same element.
4. Select Composition Information in the Spectrum Header Window. Enter the elements present in the standard together with their compositions; include elements that do not have x-ray lines in the spectrum. Then click the radio button for Valences and enter the valences for each element. These will be used to compute oxygen composition if you select to compute oxygen by stoichiometry. Also see the chapter on Headers.
5. Click the "Specimen is a Standard" box before leaving this window.
6. Leave the Spectrum Header Window by attaching the header to the WORK SPECTRUM ("OK" button). (You may make any changes permanent by editing the DTSA file - see chapter on Headers).
7. Now Set-up either the Simplex or the Least Squares fitting routine and Do a Fit on the standard spectrum. ***NOTE***: Standards and data must be fit by the same procedure.
8. Pull down the "Analysis Menu" and select "Make ZAF Standards" under "DTSA ZAF". A dialog box will appear on the screen.
9. Select "Run ZAF on Work Fit-Results". If everything was properly set up, a message box will appear displaying the analysis conditions. If these are satisfactory, select the "Its OK" button and the computations will be done with the results displayed in a window. If there are missing or incorrect data detected by the program, a diagnostic message will appear describing what to do.
10. If everything is satisfactory and this is your first standard of the current session, you need to either open an old standards file or start a new standards file. Then select "Add Standard to the file". For additional standards, it is not necessary to open or start a new file if the standard is to be added to the same file. Simply click "Add Standard to the file" after the "Run ZAF on Work Fit-Results" has been done.
Standards may be added to any existing standards file or a new standards file may be started. It is suggested that all of the standards in any file have the same beam voltage, be measured with the same detector, and all be fitted with either the simplex or the least squares routine. This is not a requirement; however, this procedure is suggested to help the user keep track of their standards. All counting data used in the ZAF is computed as counts/second/nanoampere so spectra collected for different live times, for example, can still be used together. The Faraday current is the average of the begin and end Faraday Currents entered in the Spectrum Header. If the Faraday Currents are not specified (zero), the program will use 1.0 nanoampere.
This concludes the standards section of the ZAF procedure. While the steps seem tedious, it should be remembered that standards may be reused for many analyses provided conditions are the same.
**NOTE** There is one situation that should be mentioned. If the spectrum has been fitted and then additional data must be entered in either the experiment header or the spectrum header, the fit must be repeated since these data are only updated into the ZAF procedure after a fit has been done.
Show ZAF Stds in the menu allows the user to examine the contents of any ZAF standards file. NOTE DTSA Standards and CITZAF Standards are identical so these files may be used for either method.
The set-up for processing unknowns is much simpler than that for standards. The individual steps will be described below however, all that is required is to set up the fitting routine, select the appropriate standards for each element, and specify the calculation procedure for an unanalyzed element if one is present. You must set up the fitting procedure first and then the ZAF procedure. Nothing is actually computed until the fitting procedure is activated. Single spectra may be processed using "Do a Fit" or an entire DTSA file of spectra may be processed with the "Fit ALL Work File Spectra" button under the Analysis Menu.
The following steps are required to Set-up a ZAF quantitative analysis:
1. Read the spectrum of the unknown into the WORK SPECTRUM.
2. Be sure that the entries in the Experiment Header and the Spectrum Header are satisfactory. If there are any elements of known composition that you do not wish to analyze, enter the Atomic Number and Composition in the Compositions Window of the Spectrum Header.
3. Set-up either a Simplex or a Least Squares fitting routine according to the instructions for these routines. Do not do the fit yet.
4. Select "Setup ZAF Analysis" from "DTSA ZAF" under the Analysis Menu to begin setting up the ZAF procedure.
5. If a Set-up File has previously been created for this analysis, select the "Use Set-up File" button and open the appropriate Set-up file.
6. Otherwise, select the "Set-up an Analysis" button. This opens a dialog box for the selection of standards for the analysis.
7. Select the "Open Stds File" button. A list of ZAF Standards Files appears in an open file dialog box. Select a file.
8. A scrolling list of Standards included in the file appears in the dialog box of step 6. Select a Standard from the list.
9. A list of Elements and X-ray Lines in the selected Standard appears in a new dialog box with a checkbox for each Element. Select elements needed from this standard and select "OK".
10. Repeat steps 7 and/or 8 and 9 until standards have been selected for all analyzed elements.
11. Select "OK". A dialog box for selecting an "Unanalyzed Element" now appears. If all elements are analyzed, simply select "OK" from this dialog.
12. Otherwise, select the appropriate buttons for the unanalyzed element. Any element may be done by difference however, only oxygen may be done by stoichiometry. For a stoichiometry calculation, the valences for each element must have been entered in the Spectrum Header (preferably in the standards file).
13. If you have any elements of known composition, the compositions of those elements should have been entered in the Spectrum Header. Select "Yes" for using elements of FIXED composition.
14. This completes the Set-up. You may Save the Set-up for use again before exiting this dialog.
15. The ZAF analysis will now be executed at the end of each fit to a spectrum.
16. The compositions will be included in the Analysis Results report and in any spreadsheet that has been requested. In addition a ZAF Output report may also be written. This is a text document containing all of the conditions and parameters of the ZAF computations. Select the "Full ZAF Report File?" button on the main ZAF window to bring up the open file dialog for this ASCII file.
The spectrum in Work is normally processed, however, if a file of fit results already exists, these results may be run through ZAF by selecting "Old Fit File" as the last step before exiting this dialog.
The Cliff-Lorimer quantitation procedure is intended for thin material specimens in which absorption is negligible. In Version 2.0 all elements in a specimen are assumed to be analyzed by the procedure. Cliff-Lorimer quantitation requires K factors which are computed from elemental concentrations and peak intensities in spectra from standard materials. A DTSA K factor is the ratio of the concentration divided by peak intensity of an element in the standard spectrum to the concentration divided by peak intensity of a selected element, the base element, in the same standard spectrum. The base element is commonly Fe or Si but may be any element common in your analytical problems. A set of DTSA K factors are used to calculate the concentrations of the elements in a specimen from the ratios of the element peak intensity to the peak intensity of the base element in the specimen spectrum.
To ensure internal consistency, DTSA K factors are calculated only from data obtained from Simplex or MLLSQ fitting results files for standards. DTSA K factors from spectra acquired with the same detector and beam voltage and fit with the same procedure can be saved in a K factor file. These files can be appended at any time, and K factors can be deleted from them. Saved K factors can be used for any appropriate analysis. The program checks that the experimental conditions and fit procedure of K factors selected for an analysis are correct for the spectrum file being analyzed. It also verifies that the lines used to calculate the chosen K factors are the same as those chosen for the analysis of the spectra using MLLSQ fitting, or are among the lines for Simplex fitting. If you are appending a K factor file and the standard spectrum fit results file has no data for the base element but has an element with a saved K factor, K factors are calculated from ratios to that element and the saved K factor. If a K factor is already in the file the user may choose to replace it with the new value or with an average of the two values. If the standard fitting results file was made using 'Fit ALL Work File Spectra' DTSA will calculate the average value and standard deviation of the K factors for each selected element.
You may use the spectrum generation feature of DTSA to obtain a K factor for an element not available in a standard. If you can generate a spectrum that matches an acquired standard spectrum that has a family of lines for an element nearby in the periodic table, use the same generation parameters to generate a spectrum for the missing element. Use the fit to this spectrum to make the K factor.
Selection of this sub-menu item displays the dialog for making DTSA K factors.
1. Select 'Start New File' to display a file naming dialog; or select 'Append Old File' to display a file selection dialog.
2. Enter a name for the new file, or choose the K factor file to append.
3. On return to the K factor dialog, the button "Fit results file to use" is highlighted; click it. A file selection dialog displays only files of Simplex and MLLSQ fitting results. Select the standard fitting results file with data to make the K factors.
4. A scrollable list of line information from the fitting file is displayed.
The list includes only results for the first spectrum. If you are appending an old K factor file, the base element information will be marked with"¥¥¥". You cannot change this base. If you are starting a new file you must click on the data for the base element first. This data will be marked with " ¥".
5. Now click on those data you wish to use for K factors. These will be marked with " ¥". You may 'turn off' a selection by clicking it once.
6. If the fitting results is a batch fit file - an entire file of spectra was fit using the "Fit ALL WORK File Spectra" selection - the data for subsequent spectra are not displayed. However, K factors are calculated from the data for each spectrum in the file. The average K factor values are saved to the selected K factor file. If the K factor file has a value for any set of lines this value will be averaged with the batch value.
If the fitting results file was made using "Do a Fit" and "Add a Fit" on a number of spectra, you may click on "Next result" to add fit results from the next spectrum to the scrollable list. Click "Avg K's" to average all K factor values obtained from the fitting results or with any values already in the K factor file. If you do not check "Avg K's", DTSA will ask if you want to average each new K factor with one already in the file, or to replace the one in the file. You do not have to select the same data from each Result.
7. Click "OK" to write K factors to the file. Click "Cancel" to exit doing nothing.
Selection of this sub-menu item allows you to examine a K factor file, and to delete K factors from the file. It also allows you to make an ASCII file from the K factor file. This file has the k factor filename appended with '_ASCII'.
1. Click "Show K Factors" to display a file selection dialog. Click the file you want to examine, or click "Cancel".
2. Information about the file and scrollable list of K factor data is displayed.
3. Click once on an item to mark it with for deletion from the file; click once to deselect a marked item.
4. Check "Make ASCII File" to write the data in the file to a formatted ASCII file.
5. Click "OK" to retire the dialog, to delete marked items, and to write the ASCII file. Click "Cancel" to retire the dialog doing nothing.
Selection of this sub-menu item displays the dialog to set up and execute a quantitative analysis base on the Cliff_Lorimer procedure. The calculation assumes that all elements in the specimen are being analyzed. Also, the thickness, to the nearest 5 nanometers, at which absorption becomes greater than 3% relative is determined for each element.
1. Click on "Old Fit File" to do a Cliff-Lorimer type procedure on a saved file of fitting results. Click on "Work File" to do a Cliff-Lorimer on all the spectra in the Work spectrum file. You must have a Simplex or MLLSQ fitting procedure set up to quant a Work file.
2. If you have done a Cliff-Lorimer quant on a similar set of lines and saved the setup, you may click "Use Setup File" to activate the K factors. Otherwise, click "Get K Factors". In either case a file selection dialog is displayed; select the file to use. The program verifies that the K factor base element is among the fitting results lines, and that the beam voltage, etc. of the fitting results and K factor file are the same.
3. Click "Save Setup" to display a file naming dialog; enter a name for the setup file.
4. Click "Output Options" to display a file naming dialog; enter a name for the binary file of quant results. If you omit this step the binary file has the default name 'CL_DefaultResults'. Any data in this file is overwritten by the new quantitation results.
5. Click 'OK' to retire the dialog. The setup data is written to the setup file;
the Cliff-Lorimer quant is done on a fitting results file. The quant is activated for a Work spectrum file when you select "Do a Fit" in the bottom window or "Fit all WORK spectra" in the Analysis menu. This Cliff-Lorimer quant procedure will remain active until you deactivate it by selecting "Deactivate any Quant Analysis" from the Analysis menu, or by activating some other quant procedure. The calculated concentrations and thickness' are written to the C-L results binary file; the calculated concentrations are also written to the fitting results binary file.
Selection of this sub-menu item allows you to examine the data in a Cliff_Lorimer quant results file. The procedure writes the data to an ASCII file named with the C-L quant results filename appended with '_ASCII', and displays the contents of this file on the screen.
1. Click "Show Quant Results" to display a file selection dialog.
2. Click on a filename to make the ASCII file and display the contents; or click "Cancel" to exit this procedure. The ASCII file is saved.
Help
Selection of this sub-menu item displays brief information about the Cliff_Lorimer menu selections.
The "Standardless" Analysis was designed to calculate a quantitative analysis based on x-ray spectra computed by the DTSA Program rather than from measured standards. This method is not as accurate as the method using standards and should not be used if standards are available. "Standardless" analysis is subject to substantial errors and the new user should read the caveat to be found in Appendix II.
To run a "Standardless" Analysis, simply select the "Standardless" ZAF button from the Analysis Menu after setting up a peak fitting routine. The open file dialog appears allowing the user to save a ZAF Output file. The Unanalyzed Element window appears allowing a selection of an unanalyzed element as in the analysis of unknowns by standards. Since this procedure automatically normalizes the results to 100%, the only choice allowed is for oxygen by stoichiometry. All other information is obtained from the Set-up of the fitting procedure.
The Hall-type analysis is a procedure for calculating concentrations in thin biological or polymeric specimens. Some region of the background in the spectra can be selected to give a good measure of the mass thickness of the analytical volume. The average atomic number of the matrix in this volume is assumed to be known. This means that the element concentrations to be determined from the spectra do not contribute significantly to the average atomic number of the analytical volume. This situation is typically encountered for biological or polymeric materials for which the matrix is mostly carbon, oxygen, hydrogen and/or nitrogen and where the analytes of interest are dilute concentrations of calcium, sulfur, etc. The ratios of peak areas to chosen background are used in the calculation of concentrations. These ratios are obtained from either the MLLSQ or Simplex fitting procedure. The Hall analysis can be applied to a file of results produced by running the Linear Least Squares or Simplex fitting procedure on an experiment file of spectra, or it can be applied to an experiment file of spectra after setting up a fitting procedure to obtain the required ratios. Standards are required for the Hall procedure: the factors that convert peak-to-background ratios to concentrations are calculated from the fit results obtained from spectra from standards. The specimen spectra must be acquired with the same experimental conditions and analyzed with the same fitting procedure as the standards. Fitting results from standards must be available for each element to be analyzed, but they need not be in the same file. Once obtained, a fit result for any element in a standard may be used in any appropriate Hall analysis. The chosen background value may be corrected for contributions due to the support film and a bulk source such as the support grid. If there is a bulk contribution to the spectra, the background should be selected in a peak free region of the spectrum but close to a characteristic line of the bulk element. A binary file of quantitative results is always saved. It has the default name 'Hall_Results' unless the output option to name it is selected. A short text version is displayed when the analysis is complete. Spread sheet files and a full text file with the fit results and Hall information may be made from the binary file. The Set-up for a Hall analysis may be saved and invoked to analyze either a fit result file or the Work Spectrum file. Only the background correction information in the Set-up may be changed.
1. To run the Hall analysis on a WORK Spectrum, set up a fitting procedure that includes all the elements with characteristic lines in the spectrum, but do not contribute to the estimated average atomic number of the matrix. The chosen background (Quant ROI) should be selected in the energy region least affected by bulk contributions and free of characteristic lines.
a) If a correction for a bulk contribution to the background is required, include a line from the bulk source in the fitting. Choose a line near the Quant ROI. Either there must be a spectrum from the bulk source in the WORK file, or a fit result file available for a spectrum from the bulk source. Check that the spectrum class for the bulk spectrum is 'bulk'.
b) If a correction for a support film is required, a spectrum from the film must be in the WORK file or in some other spectrum file. Check that the spectrum class for the film spectrum is 'film'.
2. If running the Hall analysis on a file of fitting results, only the Hall analysis itself needs to be set up.
a) To do a correction for a bulk contribution to the background, the bulk line in each spectrum must have a fit result in the file to Hall analyze. Also, there must be a fit result for a spectrum from the bulk source in either the fit file to analyze or in some other fit file. Check that the class is 'bulk'.
b) If a correction for a support film is required, a fit result from the film must be in the fit results file to Hall analyze or there must be a film spectrum file. Check that the class is 'film'.
3. Be sure there is a fit result from a standard for each element to be analyzed. Verify that the concentrations are for the standards are correct in the fit results and that the spectra were saved as standards.
Select Hall Bio Analysis under Analysis in the Main Menu and select Quant to display the Hall Set-up Dialog shown below.
1. Select either Old Fit File, and indicate the file to use, or Work Spectrum.
2. Check Film Correction to activate the dialog. Indicate the source for the film correction data. If requested, select the file. Click OK to retire the dialog.
3. Check Bulk Correction to activate the dialog. Indicate the source for the bulk correction data. If requested, select the file. Input the characteristic line to use. Click OK to retire the dialog.
4. Click Get Standards to select the fitting results file to be searched for standards data. The elements for which standards are not found are reported and the user is requested to supply another file to search.
5. Supply the average atomic number of the matrix.
6. Select Save Set-up at any time to name the current Hall Set-up data. The Set-up file is written when OK is selected to retire the Hall dialog.
7. Select Output Options to name the binary file of Hall results.
8. Select OK to activate the new Hall Set-up data, or Cancel to maintain the current values.
9. If analyzing a file of fit results, the Hall procedure is immediately applied and the results are displayed. If analyzing the Work Spectrum file, the Hall procedure is applied after the fitting procedure is activated.
10. Select Show a Hall File from the Hall dialog to select the file of results to display.
11. To make ASCII files from a Hall binary file, activate the Hall dialog and select only Output Options. The output options dialog will be displayed.
After setting up a fitting procedure (either Simplex or Least Squares), an entire DTSA file of spectra may be processed by selecting this option. If a quantitative analysis is also set up, this will be done as well. Beginning with spectrum 1 in the file, each spectrum will be fit, any quant routine will be run, and the results will be added to any files that were set up for results. Please note that this function will process ALL of the spectra in the file.
Selecting this will produce a qualitative analysis for each spectrum in the file beginning with the first. The parameters used for this calculation are set in the "Auto Peak Options" part under the "Peak ID" button of the Main Control Window. For each spectrum in the file, this function will locate the peaks and try to identify each peak. The results are written to a spreadsheet file in ASCII. There are two options for the spreadsheet: 1 - all of the x-ray lines found are reported (limit of 100 spectra or 200 x-ray lines - see next item) or 2 - only one or two lines are reported for an element in a spreadsheet that has specific columns for each line (not all elements are included but there is no limit as in option 1). For more information, see "Peak ID" under the Main Control Window.
This is the same as above except that the Qual starts with the spectrum currently in the display. Use this to continue a Qual on a file that has exceeded the limits imposed by option 1 above.
Information on this item is provided in the Appendix. This allows for the running of John Armstrong's CITZAF program in a manner similar to the DTSA ZAF procedure. Standards are set up the same way as for DTSA ZAF, in fact the standards files are exactly the same and may be used for either method.
The Manual Peak Stripping feature is a qualitative tool useful for stripping peaks manually, by eye, from a spectrum to see if other peaks from another element might be hiding underneath.
It is generally a good idea to use a Bi-Polar display and have the spectrum expanded horizontally so that only the peak bundle you are working with is visible on the screen. This will considerably improve the response time.
After preparing the spectrum, push Load and the following dialog will appear:
There are three choices controlled by the radio buttons. If the reference peak option is chosen, it is required that there be a valid file of references available to choose. This file of references is the same as used in MLLSQ. After OK, the first dialog reappears and the process of stripping can begin. Use the horizontal bar at the bottom left of the display to control the amount to be stripped. Use the horizontal arrow icons to control the width of the stripped peak(s). Use the buttons at the very bottom left of the display to control the energy of the stripped peak(s).
If the horizontal (energy) position of the stripping peak(s) is different from the peak(s) being stripped then the result will look like the following:
If the width of the stripping peak(s) are different from the peak(s) being stripped, then the result will look like:
If the amount (height) of the stripping peak(s) are different from the peak(s) being stripped, then the result will look like the following:
Finally, if the width, height and position of the stripping peak(s) are the same as the peak(s) being stripped then the result will look like:
This last case will result if there are no hidden peaks beneath the peak in question. If no combination of height, width or position will remove a peak shaped object, then it very well may be another peak. If the structure is small and is on the left side of a peak, it may also be the contribution of incomplete charge collection. If this is the case, the structure will not look quite like a peak because of its strong asymmetry.
In this version of the program, it is not possible to change the width or position of a reference peak. In a later release of DTSA, we will add in controlled amounts of the first and second derivative which will achieve this purpose over a limited range
This is designed for tracking your detector parameters over long periods of time. For this, a spectrum should be collected from an easily available and stable material using your normal operating conditions. For the best results, the material should have x-ray peaks at both high and low energies and there should not be seriously overlapped peaks. Collect the spectrum for enough time to insure good counting statistics.
Load the spectrum into WORK and select the QC Setup from the Analysis Menu as shown in the following figure.
The QC Setup dialog will then appear.
Select either the EPMA/SEM button or the AEM button. Enter the beam voltage, beam current and the acquisition live time in their appropriate boxes. Clicking the "Set up Calibration" button will bring up the Calibration Dialog. Set up the energy calibration as described. Click the "Select Peaks" button. This brings up the setup for a Simplex fit. Select the peaks you want to fit in your spectrum (these may include the peaks used for calibration). When everything is ready, click the "Save Set-up" button.
You may now run the QC procedure that you have saved by selecting the "Run QC" button from the File Menu. The following dialog appears.
Click the "Select the Setup Procedure File" and open the Setup procedure you have created in the previous steps. You may have as many of these files as you have detectors that you wish to monitor. Then select either "Open an Old Spreadsheet File" or "Create a New Spreadsheet File" depending on whether you are starting a new file for the results or adding the results to a previously created file. When everything is complete, click the OK button to run the procedure on the spectrum in work. The results will be entered on one line of the spreadsheet file. The EPMA/SEM results are the same as the AEM results except that the fit for beam voltage is not done for the AEM. The results consist of the fitted beam voltage (for the EPMA/SEM), the slope and intercept for the energy calibration, and the full-width at half-maximum (FWHM), energy and area for each fitted peak.
Continue by collecting a spectrum of the same material at a later date (weekly/monthly) and running the same QC procedure on it. The results will be appended to the results spreadsheet. After a period of time, you will be able to plot any of the different parameters in the spreadsheet. In this manner, you will be able to track your detector performance and possibly locate problems before they become serious.