Why do you need EDS? What are you looking for? Specimen preparation may depend on it.

If you want to measure a leaf, a leaf taken from the field is cut as small as possible, say, 1mm x 1mm or 5 mm x 5mm, and put on the metallic sample holder by electric-conductive double-sided adhesive carbon tape (the size of the tape should be wider than the specimen), and you can meadure the secondary electron image, and then you observe the X-ray characteristic lines.

If the SEM image is white, it is charged-up, and you must make weaker the electron beam current. Or you must make the magnification smaller. Or you must move the beam posision as near as possible to the edge of the leaf. Then the charge will escape to the electric conductive tape. You do not need any kind of carbon coating nor gold coating. If not enough avoiding the charge-up, you can drop diluted ionic liquid in order to avoide charging.

Other plant specimen is similar to the above procedure. That is to say, the answer to your question is, any sample preparations are needed; you can mesure as it is.

Please check the paper below how to use the ionic liquid.

Article Effect of electrical charging on scanning electron microscop...

Article Possibility of Scanning Electron Microscope Observation and ...

Article SEM Observation at High Magnification and EDX Analysis of In...

Article Effect of Charging on EDX Spectra in SEM-EDX Analysis

Heartiest Thanks Madam

Can You Provide me the references as well??

Just a warning: protocol suggested by Janhvi Pandey is not suitable for EDX, because osmium tetroxide is used. Do not use OsO4 if you wish to do analysis.

Just an erratum to my answer.

Error: That is to say, the answer to your question is, any sample preparations are needed; you can mesure as it is.

Corrected: That is to say, the answer to your question is, any sample preparations are NOT needed; you can mesure as it is.

Jun Kawai

And what if EDS analysis should be correlated with morphology of a specimen? Would you still recommend to "measure as it is?"

Your advice is partial: good for some cases, bad for others. It is good, say, for study of contamination of a plant surface, bad for cell structure.

Granted, we still do not know what the question we discuss was about.

Thanks for Kind support and guidance

Vladimir Dusevich, Jun Kawai Janhvi Pandey

The comment of Vladimir Dusevich is important. That is to say, what and why we want to measure or observe is important for preparing a specimen; without purpose or aim, we cannot prepare or even the sampling.

However the SEM-EDX is not very sensitive for trace elements, not sensitive for high energy (>15 keV) nor low energy (

Dear Jun Kawai , energies above 15 kV are rather good for EDX, energies below 0.5 are not too bad, L and M lines are used widely, contamination is a serious problem, and water in specimen is unwanted substance.

I thank Vladimir Dusevich for comments, "energies above 15 kV are rather good for EDX, energies below 0.5 are not too bad, L and M lines are used widely, contamination is a serious problem, and water in specimen is unwanted substance." and " It is good, say, for study of contamination of a plant surface, bad for cell structure."

However these two comments are originated from misunderstanding that the SEM-EDX is sensitive to surface contamination, as well as bulk (near surface).

SEM-EDX is not sensitive to the surface contamination compared with XPS; SEM-EDX is not sensitive to the bulk elements compared with optical spectroscopy e.g. LIBS.

However, again, the comments of Vladimir Dusevich is good for making deeper the discussion here.

The following two papers can be read at ResearchGate,

A paper published in Spectrochimica Acta:

Article Risk assessment of TiO2 photocatalyst by individual micromet...

A Japanese language journal

Article SEM-EDX－SR-XRF-XANES

where an SEM-EDX instrument was installed on a beamline at synchrotron facility, and then the synchrotron X-ray beam was introduced into the SEM vacuum chamber, and the same specimen was measured by the same SDD detector (an EDX spectrometer) without moving the sample nor the detector. Photo and schematic illustration of this experimental setup are shown in the above papers.

The spectra shown above papers indicate that Mg, Al, Si is more sensitive for SEM-EDX than for SR-XRF. This indicates that the comments of Vladimir Dusevich is mostly valid. But on the other hand, Mn, Fe, Ni, Cu are more sensitive for SR-XRF than for SEM-EDX. Usually SEM-EDX is not sensitive for later-half transition metal elements compared with X-ray tube XRF.

The detection of X-ray energy lower than 0.5 keV are usually not easy because of the electric noise, as well as the thick detector window. We can avoid these factors, but skilled experimental technique and expensive detector are needed.

The Bremsstrahlung (the X-rays produced by the deceleration of incident electron) is not avoidable for SEM-EDX and thus the Bremsstrahlung makes the background of SEM-EDX higher, making worse the detection limit compared with e.g. PIXE.

L lines are usually one order of magnitude weaker than K lines; M lines are also two orders of magnitude weaker than the K lines.

Wet specimen (of which vapor pressure is about the same as a few degree Celsius temperature water) is usually measureable for "low vacuum SEM", which are now popular SEM.

Dear Jun Kawai , you site a paper with poorly executed experiment, it is not something other people can learn from. In reality EDS method is quite different from your understanding. And low vacuum SEM by itself is not suited for wet specimens, because a cooling stage is required.

The comment of Vladimir Dusevich that my group papers are "poorly executed experiment", is because the chief interest of my group is on X-ray spectra and on elemental analysis, while the main interest of Vladimir Dusevich is probably on the secondary electron imaging.

The specimens measured by my group are not complicated and thus a rough imaging of a single particle is enough. However in case of biological specimens, imaging is most important, and thus the osmium treatment is sometimes very important as Janhvi Pandey has informed, even though Os is toxic, and Os sometimes interfares the X-ray spectra.

As I pointed out before, the detection limit of SEM-EDX is not enough for most of the biological specimens, imaging is the main aim of using SEM for most of the bio-users.

The comment of Vladimir Dusevich , "low vacuum SEM by itself is not suited for wet specimens, because a cooling stage is required. " is not an appropriate comment, because cooling down to +1 oC is enough; liquid nitrogen temperature cooling stage is not required. Rather the spatial resolution of low vacuum mode is sometimes not enough for biological specimens.

Finally I would like to cite two more papers from my group.

Volcanic Particles:

Article SEM-EDX Analysis of Insulator Specimens by Diluted Ionic Liq...

Baculogypsina:

Article SEM Observation at High Magnification and EDX Analysis of In...