Full range scans on WDS spectrometers

[John Donovan]

Sometimes one might be interested in seeing what trace elements might be present in a sample before performing quantitative analysis. Normally one can utilize the EDS spectrometer for a "qualitative analysis" of which elements are present. But if the elements of interest are trace elements and one has little or no idea of what might be present, we may need to resort to some time consuming scans on our WDS spectrometers.

Performing a qualitative analysis on our WDS spectrometers has another advantage, and that is: accurate identification of the trace elements.  Sometimes due to the relative lack of sensitivity for trace elements by EDS, we may not see secondary emission lines that can aid in our qualitative analysis. Obviously, if we have misidentified our elements, then our quantitative analysis is going to be "worse then wrong"!

So for a WDS qualitative analysis we will want to assign a single element to each of our WDS spectrometers using a variety of Bragg crystals in order to obtain a full range of emission lines. In the following example we have specified one element for each WDS spectrometer using a different Bragg crystal, and then peaked on each element in order that our spectrometers are properly calibrated as seen here in the Elements/Cations dialog:



Next using the Peak/Scan Options dialog we can set the scan range on each spectrometer to it's full mechanical extents using the button outlined here in red:



Once that is complete we can see our full range scan setup here:



Next we run our scan with enough dwell time per point to obtain our required sensitivity. One can utilize either a "ROM" (continuous) scan or a step and count scan. For long dwell times we will normally want to utilize a step and count wavelength scan for a couple of reasons. First, with a ROM or continuous scan, the spectrometer may not be capable of scanning at very slow scan speeds reliably, and second, because the instrument only returns the count intensity for each scan point, and not the associated spectrometer position, we need to estimate the spectrometer position. This estimate may or may not provide the spectrometer positional accuracy that we require for a confident x-ray emission line identification.

The following scan was a step and count scan with 4000 points which took 20 hours:



In the above screenshot we've selected the Analyzed Elements option for display of KLM markers. The KLM markers displayed are the elements which we selected for peaking on each spectrometer. These elements could have been any element in that spectrometer/crystal range, but generally we would select an element somewhere in the middle of the spectrometer range for best accuracy.

Other KLM options include the All Elements option which for a full range spectrometer scan can take 5 or 10 seconds to loads all the KLM markers for all elements in the periodic table within that range! A more useful option might be the Specified Element option, which allows the user to click through the periodic table one element at a time for identification of an unknown emission line.

Another option is the User Selected Lines option, where the user can load the NIST x-ray data and select elements by a number of criteria, including individually selected, x-ray range, element type and also from a periodic table.

However all these KLM display methods are only temporary and are not saved to the probe run for subsequent use. For a more permanent selection of the element KLM marker display, one can specify which elements to display by simply adding those elements to the wavescan sample as "specified" elements. That is in addition to the analyzed elements, we can also display "unanalyzed" elements for the KLM markers.

This screenshot shows the Elements/Cations dialog after adding in a number of specified elements (that is without an x-ray line assigned) to the last wavescan sample in the wavescan sample set:



Once we've done this, we plot our wavescan sample again using the Plot! window and now we seen this display for the Analyze d Elements KLM option:



For best results please update to the latest version of Probe for EPMA using the Help | Update Probe for EPMA menu to obtain version 12.5.7 or later.

[Probeman]

I wanted to bump this topic forward because some people might not be aware that when one clicks the "Analyze Elements" KLM option in the Plot! window, that includes elements that are not actually analyzed but also merely "specified".

That is elements without an X-ray emission line as seen here in the Elements/Cations dialog:



Once these unanalyzed elements are specified, they will show up as KLM markers in your wavescan:



This is particularly helpful when analyzing trace elements in matrices where the major elements are not actually analyzed and instead only specified.