Wavelength Dispersive Spectroscopy - A Geological Example

Quantitative compositional analysis through SEM-WDS

Compositional analysis of a material using electron microscopy is largely based around energy dispersive spectroscopy (EDS) and wavelength dispersive spectroscopy (WDS). Both use the characteristic X-rays released by the sample during analysis but utilise different aspects of them; EDS measures the energy of the characteristic X-rays whilst WDS measures their wavelength.

Although PEMC runs standardised EDS (which is an improvement upon qualitative EDS often used in electron microscopy), WDS is inherently more accurate and is recommended for our users who require improved precision and the quantification of minor (low abundance) elements.

A Geological Example

This case study features a complex piece of jadeite - part of the pyroxene family of minerals with a composition of NaAlSi₂O₆. However, minerals in nature are very rarely an exact composition and show variations in major and minor elements between different localities, different crystals in the same rock, or even within the same crystal. Understanding the distribution of elements in minerals - and therefore geological systems - is a key part of geochemistry and petrology.

This layered element map was collected using EDS and shows just how many different mineral phases are present in this sample. The different colours here reflect variations in calcium, sodium, and aluminium. This image is 1mm across.

Whilst standardised EDS is usually accurate in identifying elements, there are cases when the energies of two or more elements overlap at which point the user must define the correct elements. This is a common issue in geological samples and we often rely on our knowledge of mineralogy in combination with electron microscopy to mitigate it. WDS measures much narrower element peaks and can help remove uncertainties.

During analysis of the jadeite, the EDS often mislabeled samarium (Sm) with sodium (Na) spectra as is shown in the background of this image. Scanning the WDS between the energies of interest reveals the peaks that are present in more detail, and here we find that the Sm peak is not a real feature. By checking element overlaps like this, we can make sure that we are only measuring elements that are present and improve the efficiency of the experiment.

To compare the different standardised EDS and WDS techniques available on our JEOL 7001F FE-SEM, example data was collected at the same locations in three different sodium-poor minerals as shown here using EDS Point ID, EDS Analyzer, and WDS Quant techniques in Oxford Instruments' AZtec software.

In order to make the most accurate comparison, the EDS was optimised at the chosen operating conditions. This lets the software calculate the compositions measured by the detector with reference to the accelerating voltage, probe current, and selected processing time. Without doing this, the un-normalised values for each element would dramatically over- or under-shoot the total of 100%. Ensuring the EDS is optimised is especially important for users investigating hydrated samples or materials with components that cannot be measured such as lithium (Li).

EDS Point ID

Point ID is the easiest way to collect EDS data and gives the composition at a point selected by the user within the instrument's current field of view. It can be readily combined with element mapping and multiple points can be queued up at once. However, the selection of areas to analyse can often be imprecise and can accidentally incorporate small features that may not be visible to the user at lower magnifications.

The data collected in the jadeite shows that Mineral 1 has a distinct composition (also inferred by it's yellow colour in the element maps), and that Mineral 2 and Mineral 3 are quite similar to eachother. In Mineral 3, the Ca content is low enough that we would recommend further investigation to establish the viability of a Ca peak in the spectrum.

EDS Analyzer

Analyzer uses EDS in a more similar way to WDS, wherein the spectra of the entire field of view is measured. This means that the user must increase the magnification of their image until the entire screen is filled by the area of interest. As such, ensuring proper alignment and focus is paramount.

Here we see some variations between Point ID and Analyzer, particularly in the oxygen (O) content of Mineral 2 which is likely the result of the sample charging during analysis. We also note that the Ca in Mineral 3 does not appear in the data produced using Analyzer.

Analyzer can be automated to reduce active time. A series of points can be queued up with the option to capture an image at lower magnification for context.

WDS

WDS gives more accurate data and can also separate out element overlaps. In this case, the WDS was able to quantify strontium (Sr) which was hidden by the silica (Si) peak in the EDS as well as identifying traces of elements not detected by EDS. Overall, the WDS is in agreement with the EDS and offers further reassurance to the user.

For reference, this suite of WDS elements took approximately 10 minutes per mineral whereas the EDS took less than a minute per point. Both WDS and Analyzer can be automated to reduce active time. A series of points can be queued up with the option to capture an image at lower magnification for context.

Drawbacks of SEM-WDS

WDS is typically used on an electron microprobe (EPMA), where four or five WDS detectors can run concurrently. This isn't an option with SEM-WDS. Instead, we have a single detector which measures one element at a time and, as such, the analysis can take a significant length of time depending on the number of elements required to be measured. This also means that ensuring the sample is correctly prepared and has a thorough conductive coating is vital, as any drift will ruin the dataset, and that mitigations will need to be established for beam sensitive materials.

Additionally, SEM-WDS does not have the element mapping capabilities of an EPMA and we would conduct EDS element mapping instead. Both EDS and WDS at PEMC are run through the same software, and it is easy to switch between the two modes without changing the operating conditions.

As WDS relies on standards to accurately quantify the material being analysed, it is beneficial to inform your contact at PEMC of the elements you are interested in prior to analysis so that we are able to verify if we already have suitable standards or if more will need to be purchased from a reputable supplier. If you are unsure of what your sample is made from, we would typically suggest some preliminary EDS analysis prior to WDS, saving time in the long run. The periodic table below shows the elements that we currently have standards for, as of April 2023.

If you would like more information about the standardised EDS and WDS capabilities available at PEMC, please contact the lab emc@plymouth.ac.uk and we will be happy to answer your questions.