The most direct answer to this is that there are some fundamental reasons cryoSTEM and cryoSTEM/EDX for most biological samples are unlikely to work well. However, if there is no other way to address a particular problem, it might be worth attempting (and please contact David Morgan at email@example.com about these sorts of problems).
In order to understand why these techniques are likely to be difficult, a general review of STEM follows.
When imaging a sample using scanning transmission electron microscopy (STEM), the electron beam is focused to a small point (often 1 nm or smaller) on/in the specimen and rastered across the specimen in a regular pattern. Electrons that are scattered at low angles are collected and counted by the bright field STEM detector (BF-STEM imaging) and much more highly scattered electrons are collected and counted by the high angle annular dark field STEM detector (HAADF-STEM imaging). The number of electrons scattered strongly enough to be detected by the HAADF detector is proportional to the atomic number, and HAADF-STEM images are often also called "Z-contrast images." Both the BF and HAADF detectors produce a signal based on the number of electrons that are counted at each position the electron beam touches, and an image is built by associating the number of electrons with each x/y position that the electron beams contacts as it rasters across the specimen.
On the JEOL JEM 3200FS, the beam current in TEM mode varies by about three orders of magnitude, from ~30 pA to ~20 nA) depending on the spot size and condenser aperture (measured using a Faraday cup), and the electrons can be spread across an area that ranges from 10's of nm to 100's of µm.
- Fundamental reason 1: beam too damaging
- Fundamental reason 2: elements to light to image
There are limited ways to reduce the strength of the beam that sites on the sample during any sort of STEM:
- change CLA
- change spot
- drop to low mag STEM
- use very short dwell times