What is the correctness of SEM, TEM and EDX analysis?
What is the percentage of error? How do you calibrate the machine?
I have got some data from the internet source. I request suggestions from the experts.
Imaging and Surface Analysis
Scanning Electron Microscopy (SEM), Transmission Electron Microscopy (TEM) and Energy Dispersive X-ray Analysis (EDX) can be used to image and investigate the properties and compositions of a broad range of sample types.
- High resolution surface inspection and backscatter analysis of samples
- Non-destructive, semi-quantitative elemental analysis of solids using energy dispersive X-ray micro analysis
- X-ray mapping, line profiles and Cameo X-ray imaging
- Surface measurements and roughness characterization using stereo-imaging
- Low vacuum variable pressure for examination of uncoated material
- Hard copy or digital export of images
- Carbon and gold coating of non-conducting samples.
X-Ray Diffraction (XRD)
- Powder X-ray diffraction
- Phase identification
- Search match and semi-quantitative analysis using Siroquant (Rietveld)
- Characterization and quantification of clays and other mineral phases.
You cannot talk about SEM/TEM "correctness", its not something permanent.
As usual (for most research methods) it depends mostly on your task and operator's skills. If, for example, you need to measure grain size, "correctness" will depend on number of viewed fields (should be correctly chosen), on correct software application and on SEM calibration (properly calibrated SEM gives accuracy of linear measurements about 2%-0.5%.
With proper application EDS accuracy can reach 2%, with poor handling its error can be 80%.
SEM and TEM images usually offer more qualitative than quantitative information. These methods are good for determining the morphology of your sample. As far as "correctness" goes, you just need to make sure that your samples are sufficiently conductive (carbon or gold coated) so your images are not distorted due to charging. Because TEM and SEM images are at high magnifications and only show a small area of your sample, it is also important to take many representative images. It is possible to extract measurements (ie. particle size) from images using analysis software, but the accuracy will be limited if your sample is variable.
If you're looking for quantitative sample composition EDX is a better way to go. SEM and TEM EDX values will be different because of the penetration depth. The percentages generated in EDX are from a spectrum collected from your area of interest. Don't trust the automatic curve fitting--make sure that all the peaks make sense for your sample. XPS is another option if you're looking for sample composition but it only gathers information from the surface and also requires curve fitting.
Percent error is difficult to determine, but it is a good idea to measure samples with known compositions or lattice structures (for XRD).
Ultimately SEM and TEM are good for if you need to know what your sample looks like and EDX is good if you need to know what's in it. The accuracy of quantitative measurements depends on the machine you're using, the settings you use, and how your analyze your data.
You cannot talk about SEM/TEM "correctness", its not something permanent.
As usual (for most research methods) it depends mostly on your task and operator's skills. If, for example, you need to measure grain size, "correctness" will depend on number of viewed fields (should be correctly chosen), on correct software application and on SEM calibration (properly calibrated SEM gives accuracy of linear measurements about 2%-0.5%.
With proper application EDS accuracy can reach 2%, with poor handling its error can be 80%.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction