image of George Santayana from Wikipedia

Those who cannot remember the past are condemned to repeat it.

What TEM Magnifications Should I Use?

Under Construction!

One of the first questions that a new user of any sort of microscope asks is usually "What magnifications should I use when collecting images?" This is actually a much more difficult question to answer than one might suppose, and the following lengthy discussion attempts to provide a few things to consider and some guidelines to follow when choosing the magnification for a project.

The most simple answer to this question is that one needs to use a magnification high enough to see what is interesting. However, this answer invokes an additional question of its own: what is interesting for a given project (or perhaps better put, what does one want to learn from the project)? Here are three examples of TEM work where the answers to these questions involve entirely different length scales and result in the use of very different magnifications:

A key aspect in the discussion of these three types of project is that both the sample itself and the goal of the work are important in determining the proper magnification(s) to use. It may also be useful to note that in the discussion above and all that follows, the assumption is made that images will be recorded using some sort of digital camera (i.e., a conventional TEM CCD camera or one of the new, direct electron detectors). If one is, in fact, recording images onto film, then the magnification used will depend not only on what one wants to see/learn, but also the details of the film scanner that will be used to digitize the images. Further discussion of aspects of these issues can be found here.

Additional issues can affect the idea of "magnification high enough to see what is interesting," especially for cryoTEM projects that involve extensive data collection and image processing. The following list highlights several issues to consider when beginning a cryoTEM project, keeping in mind that some of these really relate back to the question of what one wants to see and how one makes certain that experimental details do not interfere with that goal:

One of the goals of many cryoTEM projects is an atomic or near-atomic resolution structure of an isolated biological complex. There have been enough recent success stories with crystalline, helical, icosahedral and non-symmetric single particles to say that while such a goal may be difficult to attain, it should be possible in many cases. However, not all biological complexes are this highly ordered, at least not as they have been prepared for TEM at the present time. For example, there are numerous two-dimensional (2d) crystals that are well-enough ordered to solve their structure at (say) 8 Å, but that have proved recalcitrant to all higher resolution studies.

What factors should I consider?


Pixel size in the image

Typical atomic spacings range from the separation between layers of graphite (3.4 Å) to the separation between atoms of a metal such as gold (2.35 Å). Using our JEOL JEM 3200FS, the graphite spacing can be seen at as low a magnification as 80,000 (or perhaps even 60,000) and the gold spacing can be seen at 100,000 (and even 80,000 in some images). However, these lowest estimates of the magnification really address what could be described as a relatively ideal specimen imaged under the best conditions.

Sampling across the image (Shannon sampling)


optimize information AND field of view