Hi all, I'm doing a literature review for the secondary ion mass spectrometry (SIMS). I'm still confused about why SIMS could not be fully considered as an type of quantitive analysis tech, are there any specific reason on it, such as ionisation efficiency, etc.
Well, that sounds a bit harsh. It is true that you can't just take relative SIMS peak intensities and use them as absolute concentrations due to ionization cross sections, analyzer transmission functions etc.
However, if you premeasure a sufficiently similar sample with known concentrations and use that as an internal reference, you will end up with numbers in a 10-20% accuracy range depending on how similar your reference actually is.
Hello, It is because the "SIMS matrix effect". It is the dependency of the ionization probability of secondary ions on the local sample composition. This effect complicates the quantification of the SIMS signal (i.e. the conversion of the SI signal detected for element X to the concentration of the element X in the analyzed sample.) This is the main weakness of the SIMS analysis. However, you can use "useful yield (UY)" and "relative sensitivity factors (RSF)" to quantify the SIMS data. These two calculation depend on reference samples.
First of all, there is no theoretical model which predicts the ionization probability for secondary particles. This mean we have no any formula, which let us calculate ionization probabilities for interesting elements sputtered from different targets. If you compare experimental results and prediction of any model from the first principles, for example, the most popular "electron tunneling" model, you will see deviation up to one order of magnitude between them for most materials. As Jürgen and Umutcan mentioned, use of special standards let you recalculates intensities to concentration for dopants and contaminants with concentration less than 1atomic %. This method works well for pure matrixes like semiconductors in the depth profiling regime. If you want analyze main elements, the situation is much complicated - you will need a "calibration curve" for each element, which can be found with several compounds (4-5, as minimum) of known composition, only. The quantitative image analysis by SIMS is still impossible (in general) due to a lot of uncertainty for the surface composition and matrix effect.
- Badges
- Science topic