Except for the STEM detector, all of the other detectors in the SEM analyze how the sample "bounces" electrons off the top of the sample or "glows" photons from the top of the sample.  The STEM detector works in the opposite way because it detects the electrons that shoot all the way through the sample without interacting with the sample.  It is a detector that sees the "shadow" made by electrons that do not pass through the sample.  (2.4 Mb original diagram image)

In biological specimens, almost everything is carbon, oxygen, and hydrogen, which are tiny elements with few electrons, so they do not interact much with beam electrons.  If we put a very thin slice of cell tissue in the SEM and beamed electrons at it, most beam electrons would go right through, just like light through a glass window.  That would not be interesting.  Instead, electron microscopists stain cell organelles and viruses using an element that has lots of electrons and protons, and is therefore opaque to the electron beam.  The most common elements for staining are uranium, osmium, tungsten, or the lanthanide rare earth elements.  At Kutztown University, we can stain with lanthanides, but we do not own an ultramicrotome (the special knife for cutting ultra-thin slices of cells).

Transmitted electron image of uranium-stained mouse brain cells taken by Devin Peterson

Although we do not have uranium staining facilities at Kutztown University (because the hazardous waste is really expensive to dispose of), two students in my SEM course were keen to use the STEM detector.  They ambitiously reached out to research universities that use big Transmitted Electron Microscopes (TEM) for analyzing tissues.  The University of Pennsylvania very graciously donated some old mouse brain samples.  I love when people like Devin and Stephanie show initiative and act sua sponte.

In the SEM lab, I keep a sign on the wall that states, "Dr. Friehauf is not a biologist" because I don't know the answers to questions about the different blobs in images like this.  In spite of my ignorance, they are still aesthetically interesting and question-inspiring, which I suppose makes them like art.  (3 Mb original file)

Uranium-stained mouse brain tissue imaged by Devon Peterson using the STEM detector

Transmitted electron imaging is usually done using a special electron microscope called a Transmitted Electron Microscope (TEM) which is designed solely for this kind of work.  TEMs work on the same principle as the STEM detector, except TEMs use much higher energy electrons.  Our SEM has a maximum energy of 30 keV, but most TEMs use 120 keV or 300 keV beams.  The stronger beams mean more electrons pierce through the sample.  Just as importantly, higher energy electron beams can see smaller things for a really interesting reason.  I've been illustrating electrons as behaving like particles - little comets that crash into the sample.  Subatomic stuff, however, is weird.  Electrons also behave like energy waves that oscillate and interact like rhythmic ocean waves on the shore.  The distance between crests of a wave is called the wavelength of the wave.  Higher energy electrons have shorter wavelengths, which means they can image smaller things.  (I equate this in class to using a fine-tipped paintbrush for making more finely-detailed paintings compared to using a broad house-painting brush.)  Kutztown University does not have a TEM because few of our users need one.
(11.8 Mb original image)

Uranium-stained mouse brain tissue imaged by Devon Peterson using the STEM detector