Hi John,

In a transmission electron microscope (TEM) the electrons are flying through the specimen. The image is a result of the electron density's distribution in the specimen. A typical thickness for the TEM samples is about 50 nm. 

There is a simple intuitive explanation: if you have an envelope with an enclosed picture, you can see this picture by transmission with bright light, but if there are dozen pictures in the envelope, you can't see them.       

Good luck!

For very small details, and I think the mitochondria membrane is such a small detail, sections of 50 nm thickness are better. Just like Denis allready said. For most TEM purposes the sections are placed on a grid with a formvar film. If the details are not well visible, the formvar film can be omitted, leaving a very fragile section.

Good luck.

I would not recommend to use grids with any film, formvar or carbon. They just decrease contrast and have no benefits in most cases. Only if you have some special task you may need coated grid, but not for usual sections of embedded tissue.

With that, Vladimir, I do not agree completely. For many regular purposes it is fine to have grids with a formvar coating to back the section.  You are right that the contrast is a little lower. But it makes the section so much stronger. If you need to look at details that are bigger than ~2 nm the formvar coating will not hamper you. If I have to work w/o formvar coating, I do cover the section with carbon coat. The benefit is that the heat from the electron beam is better conducted and the section isn’t ‘burned’ away.

So, it depends on what you want to study.

For many regular EM-work a formvar coated grid is OK, it saves you from a lot of extra work, so it is just more economical and efficient.

When the experiment requires it, work without a coating.

I do a lot of EM on skin tissue. The upper stratum (s. corneum) is quite electron dense and breaks very easy in the microscope. So obviously I usually coat my grids. With other tissues this might be less of a problem.

Less resolution results with thicker specimens. Very fine membrane work utilizes 50nm epoxy embedded specimens; but doing cryo-sectioning without a plastic embedding medium will not allow this thickness. The specimen has to be much thicker or it will fall apart. Therefore, section as thin as you can and realize that your resolution (e.g. the ability to discern 2 spots) will be less. Unfortunately this is the nature of cryo-sectioning. The same is true with light level cryo-sectioning.

I should apologize for my misleading post (above).

I missed part about CRYO microscopy.

Sure, support film in this case is very helpful.