Qualitative EDS X-ray microanalysis using SEM and TEM
What elements are present in the sample?
Qualitative microanalysis means that the elements present in the sample are identified from their Characteristic X-ray peaks, but their abundances are not determined. While commercial peak-identification software is improving all the time, it is not yet 100% accurate. Elements that are present in the sample may be missed, and elements that are not present may be falsely identified.
For qualitative microanalysis of an unknown sample, consideration must be given to the operating parameters of the microscope and also the properties of the sample to be analysed.
Microscope operating parameters
In the SEM, if the sample is stable under high-vacuum in the electron microscope and is not susceptible to damage by the electron beam, then an accelerating voltage of 15-30 kV is recommended for SEM analysis. This is sufficient to efficiently generate at least one family of X-ray lines for all elements. In the TEM, the chief requirement is to collect enough X-ray counts in the spectrum from a thin specimen for good analysis. This is generally achieved at the highest accelerating voltage.
If the sample is likely to be damaged by a high-energy primary electron beam in either the SEM or TEM, then it may be necessary to use a lower accelerating voltage. If this is the case in the SEM, the higher energy X-ray lines may not be efficiently generated and low-energy X-rays will have to be used for element identification, e.g., the L or M lines for elements with Z > 20. An alternative approach to reducing sample damage is to lower the electron dose by using a broader (defocused) beam.
The electron beam current (probe current or spot size) will control the X-ray count rate or intensity of the generated X-rays. The beam current should be adjusted to minimize damage to the sample but generate sufficient X-rays to allow reliable identification of peaks. At the same time, the beam current needs to be adjusted to minimize spectral artifacts and achieve system dead times of 20-50%.
The electron beam – sample – detector geometry should be optimized. Set the sample stage to the microscope's manufacturer's recommended working distance. The X-ray EDS detector will have been installed with an optimal working distance also, and this will vary from microscope to microscope.The X-ray detector should be as close as possible to the sample to maximize collection of the generated X-rays, but this may be limited by the presence of other detectors and safe operation of the microscope. There should be a clear path between the sample and the X-ray detector.
No special sample preparation beyond that required for electron imaging is required for qualitative analysis in either the SEM or TEM. In the SEM the elements present in rough or unpolished samples can be determined by qualitative EDS microanalysis but some care needs to be exercised when collecting X-rays from these samples. If the electron beam is focussed in a hole in the sample then X-rays may not be able to escape from the hole and will not be detected. Similarly, if the X-rays have to travel through the sample to get to the detector they will be absorbed and not detected in the ED spectrum.
Conductive samples do not need to be coated, but insulating samples may need to be coated to make them conductive prior to qualitative X-ray microanalysis. Carbon, Au, Pt or any other conductive coating material may be used, but X-rays will be produced from the coating material and will be present in the ED spectrum. Characteristic X-rays from the coating material may overlap with X-rays produced from the elements in the sample, so carbon is the preferred coating material as it minimizes spectral interferences.
X-ray peak identification
The energies of the Characteristic X-rays emitted by the elements in a sample allow them to be identified. However, the presence of overlapping peaks from different elements means that correct identification depends on being able to recognize the peaks in the different X-ray families. The relative weights of the Characteristic X-ray lines in each family are consistent and this, along with their energies, allows related peaks in the X-ray spectrum to be recognized. By measuring the energies of the major X-ray peaks in each family, the corresponding element can be identified. Commercial peak-identification software can do this, but the results need to be verified.
At energies above 4 keV, the peaks of the K and L families are resolved
The K family of X-rays has two peaks, Kα and Kβ, with intensity ratios of approximately 10:1. Note: There are actually two X-ray lines in the Kα X-ray peak: Kα1 and Kα2, but they cannot be resolved by ED spectrometers so the combined peak is referred to as Kα.
The L family of X-rays is more complicated with six lines commonly visible in the spectrum. The Ll (ell-ell) line results from electron transitions between the M I and L III subshells, and forms a small peak on the low-energy side of the family. The Lα, Lβ1 and Lβ2 lines form a series of three peaks with descending magnitudes and relative intensities of ~10:7:2. The Lγ1 and Lγ3 lines form small peaks on the high-energy side of the family.
M family peaks are not present at energies above 4 keV
The Mα and Mβ peaks may not be fully resolved, but the Mζ (M zeta) peak on the low-energy side of the family and the Mγ peak on the high-energy side are usually present and allow M family peaks to be identified.
At low energy
At energies below ~2.5 keV, the main peaks of the K, L and M families are not resolved and only one main peak is present. The minor peaks, Ll, Mζ and Mγ, can be used to help identify the family.
- For SEM analysis, use an accelerating voltage of 30 kV to ensure efficient X-ray generation. The higher accelerating voltages used in TEM analysis means that higher energy K and L lines can be used for identification.
- Run the automatic peak-identification software that comes with your system. Verify the results as follows.
- Identify and label major peaks at 4 keV and above:
- There are no M family peaks
- There are at least two resolved peaks for both K and L families (see above) so these can be identified along with related minor peaks
- If there are L family peaks there should be M family peaks at low energy, and these can be identified, e.g. Ta L and M families
- If there are K family peaks there should be L family peaks at low energy, and these can be identified, e.g. Zn K and L families
- Any unidentified peaks below 4 keV should be K family peaks, and can be identified.
- Consider unidentified minor peaks in the spectrum: are they Characteristic peaks of elements present in minor/trace amounts (only the most intense peak of the family may be visible above the continuum), or are they spectral artifacts?
- Consider the possibility of overlapping peaks, particularly at low energies (below ~2.5 keV) where each family has only one main peak, e.g. Si, Rb and W or S, Mo and Pb.
Clearly, the more you know about your sample and how it was prepared before undertaking chemical analysis the fewer surprises you will have. But be suspicious: don’t just look for the elements that you think will be in the sample. Make sure you identify all the elements that are present.