During EDX analysis we normally get elemental composition both in Normalized atomic % concentration and Normalized wt % concentration. For reporting EDX data for chemical composition atomic % is what I think we should look for. But what information we can obtain from wt % data?
Actually, difference is not much bigger than difference between inches and centimeters. Data presentation depends on your task. For many (if not most) applications wt% is more suitable for reporting results. For example, when performing EDS of bone research, I give wt% and never use normalized data (it will be misleading). From other side, for the same bone research, when Ca/P ratio needed, I give it for atomic ratio (as research tradition demands).
Actually, difference is not much bigger than difference between inches and centimeters. Data presentation depends on your task. For many (if not most) applications wt% is more suitable for reporting results. For example, when performing EDS of bone research, I give wt% and never use normalized data (it will be misleading). From other side, for the same bone research, when Ca/P ratio needed, I give it for atomic ratio (as research tradition demands).
Since both are relative quantities unit doesn't matter...but as Vladimir says depends on what you are reporting...and the relevance of the unit to your report. If you are analyzing the bulk composition the wt % becomes more relevant. But actually you are getting the relative concentration of those elements only which are quantifiable from the EDX spectrum (typically > 0.1%)....but the sample may have other elements also which are not detectable...or quantifiable.
Dear Amin,
Would you please explain how electron beam energy influence acquired wt%?
Dear Amin,
I am afraid...you got some vague/partial concept from the instruction manual.
As far as fluorescence phenomena is concerned you are partially correct.
There are multiple reasons why light (or low-Z) elements are not detected efficiently in EDX.
For Low-Z elements there is also a mutually exclusive phenomenon that takes place i.e. Auger electron emission. If the total photo-ionization probability is one, and the fluorescence yield probability is "w" then Auger electron emission probability is "1-w". Auger emission probability is higher in low-Z elements (the reasons you should find from a std Atomic Physics text book).
The fluorescent characteristic X-ray energy of low Z elements being less...they are absorbed in the detector window depending on the material of the window and its thickness. Different X-ray detectors have diff. efficiency also because the detector crystals may be different....there are many details.
By increasing/decreasing electron energy (voltage) you can only change the inner shell-ionization events and not the characteristic w or (1-w) for each element. A voltage of some factor above the electron binding energy (particular Shell: K or L or M) of an element will give you the highest ionization events. You will also probably be able to yield more X-rays, hitting more no of atoms by increasing electron flux. (current)..but cannot change the fluorescence phenomenon...Hence the net atom % (weight %) should not change with increasing electron-voltage/current. The software should automatically correct for your current and voltage.
Electron induced fluorescence is more a surface phenomenon since electrons cannot probe deep into the sample...the information depth is also sample specific. Homogeneity and roughness are sample issues and causes changes in absorption and scattering and hence final detection.
I suppose there is a ZAF correction for atomic %(converted to weight %) relative quantification in SEM-EDX.
You should consult a more standard text book like "Handbook of X-ray Spectrometry" By Van Grieken and Markowicz to understand the issues....and also think about it.
Dear Amin,
As was explained by Dhrubajyoti Gupta, wt% does not depend on accelerating voltage.
Also, element detection (qualitative analysis) does not depend much on surface quality. While some older EDS systems has detectors with beryllium windows and cannot detect elements with numbers lower than 11, it should not be a problem for a relatively new SEM, such as Zeiss Auriga, which almost certainly has ultrathin window and should detect light elements such as oxygen and nitrogen. And there is no EDS systems which are limited to detection of elements with numbers lower than 16! Problem with your EDS may be in serious contamination with oil. Layer of oil on window will absorb x-rays of lighter elements and limit usage of instrument. Your technician should know better and perform window cleaning as needed according to manufacturer instructions.
Dear all, first of all, thank you for this discussion. Second, if I have understood you correctly, if I want to report on the ratios of say Mg and Fe in different olivine (Mg/Fe silicate mineral) types, I should better use Atomic% instead of Weight% ? Correct ?
Both are very important.
https://www.researchgate.net/publication/268389305_NISHAN-E-ZAFAR
I have a pure Iron oxide nano particle sample(hematite) . I got EDS on my sample in different ages. The EDS result shows the height of Oxygen and Fe picks on EDS result changed after 30 month aging.The change on the height is very obvious. How ever it is still a pure iron oxide. What the height change shows ?? Thank you for your help !
- Badges
- Science method