Energy Dispersive X-ray Spectroscopy in Materials Characterization

The next blog in the characterization mini-series is from my colleague Andrew Deal.  In his blog, Andrew will speak about Energy dispersive X-ray Spectroscopy (EDS) which is a very powerful technique for providing information about the composition/chemistry of a sample that is being analyzed by Scanning Electron Microscopy (SEM).

Hello All,

Deal_Andrew-150x150My name is Andrew Deal, and I’ve been at GE for almost seven years.  I’m originally from Maryland, and I was all over the east coast before ending up in New York State.  I obtained my Bachelor’s degree from the University of Virginia in Engineering Science.  I received my Masters and Ph.D. from the University of Florida, where I studied metallurgy.  My thesis work focused on quantifying the morphological stability of solid/liquid interfaces in faceted materials solidification, specifically the Ge-Sb system.  I also spent a year and a half as a post-doc at Lehigh University in Pennsylvania, where I became proficient in electron microscopy and the characterization of materials.  One of my specialties here at GE is Electron Backscatter Diffraction (EBSD), which is an analytical tool used on polycrystalline materials to understand the arrangement of atoms in a bulk sense, e.g. providing texture, grain size, plastic strain, boundary, and phase data.  However, I use a variety of characterization techniques to understand microstructure-property relationships in all kinds of materials.   One of the most common technique is Energy Dispersive X-ray Spectroscopy (EDS, or sometimes EDX), which is the subject of today’s conversation.

1. What is EDS?

Energy Dispersive X-ray Spectroscopy, EDS (or sometimes EDX), is an analytical technique employed in electron microscopes.  Using EDS, a material’s local surface chemistry is obtainable in point or map format, with a resolution up to a few hundred nanometers.  It is fast, reliable, and widely accepted.  In general, most elements of the periodic table can be identified, with the exception of those lighter than boron.  EDS data is usually semi-quantitative in nature, giving a sense of which element are more prevalent.  However, an accurate quantitative analysis providing the exact at% or wt% of each element present is only possible in ideal circumstances.  Unfortunately, modern EDS equipment readily reports quantitative estimates and element identification without requiring verification.  This very frequently leads to incorrect reporting of the material chemistry in the hands of inexperienced operators.

2. How does it work?

Example of an EDS spectrum (# X-rays vs. Energy) from a standard glass, with characteristic X-ray energy peaks (element 'fingerprints') identified
Example of an EDS spectrum (# X-rays vs. Energy) from a standard glass, with characteristic X-ray energy peaks (element ‘fingerprints’) identified

The electron microscope supplies a beam of electrons, of a particular energy, that is focused onto a sample of material.  When the electrons hit the sample they lose energy in a number of ways, one of which is through an X-ray emission process.  Specifically, a beam electron may eject an electron from an atom of the material through an inelastic collision.  That atom will then have an electron vacancy, making it an ion, and it will reconfigure its electron shells to minimize its energy.  In doing so, an X-ray is often released having an energy that is characteristic of the atom that generated it.  These ‘fingerprint’ X-rays are what an EDS detector gathers to identify the elements in the sample material.  Most elements have multiple, characteristic X-ray fingerprints of unique energy.

The process of acquiring and EDS spectrum
The process of acquiring and EDS spectrum

It is important to note, however, that an EDS detector has a finite energy resolution.  This means certain characteristic X-rays are not distinguishable.  For example, it is often impossible to detect nitrogen if titanium is also present.  This is because the nitrogen fingerprint is too similar to one of the titanium ones.  Such an overlap can lead to the misidentification of elements by the instrument, requiring operator verification.  Additionally, this example highlights a case where accurate quantitation is difficult, if not impossible in a practical sense.

3. Why do we use the instrument?

The main advantages of EDS are its spatial resolution, relatively high signal-to-noise level for major elements, analysis speed, and ease of use in obtaining chemical information from a material.  Due to the high throughput of modern detectors, all elements can essentially be identified at simultaneously.  EDS can be utilized on any flat, conductive sample that can handle a moderate vacuum, e.g. anything that can go into a scanning electron microscope (SEM).  It is also employed in the transmission electron microscope (TEM).   Both instruments can provide morphological information in the form of images to accompany the EDS data, making it easy to identify specific locations on a sample to analyze.  Analysis times can range from a few seconds (taking a single spectra) to hours (mapping an area with trace elements).

Example of EDS data:

The EDS data below are an example of what information can be obtained using the technique.  The material is alloy 718, a common material used in GE’s aircraft engines, which contains several phases.  A single 512×512 raster scan of EDS data was collected in an SEM, and this was broken down into chemical maps of various characteristic X-ray fingerprints.  Nine elements were mapped: Ti, Cr, Fe, Ni, Nb, C, Al, N, and Mo.  As you can see, the chemical segregation of the elements is apparent, but everything is not as it seems!  Note that the N fingerprint matches that of Ti, and this is the exact example discussed before.  By looking at these maps, you can’t definitively say that the N is present.  It turns out in this case it is present, as confirmed by other techniques (and historical knowledge of the alloy), but the maps would look the same if only Ti were present.   Additionally, you’ll notice that the Nb and Mo maps are similar.  This is another case where the characteristic X-rays are very close.  So, to say the large oblong phase has about the same amount of Nb and Mo present based on these images is incorrect.  There is much more going in this example than what I’ve been able to discuss here, so feel free to ask me any questions below.

Screen-Shot-2012-10-12-at-11.15.25-AM-300x234

– Andrew


3 Comments

  1. Sebastian R M

    Dear Sir,
    I have been working on the characterisation of different components of municipal solid waste. Can the data from SEM/EDX be relied upon for determining the elemental composition of say, paper present in waste?

  2. siuli Dutta

    Dear sir,
    I siuli, during EDS analysis found MO-L, and Sulphur(S-K)) overlap in same energy level. Can u tell me, how i distinguish
    between them?

  3. Bob Jones

    Why would it be difficult to detect nitrogen on stainless steel using EDS?