A wide variety of petrographic techniques are enabled by the signals generated by a beam of accelerated electrons.  This subtutorial attempts to convey at the undergraduate-level the most basic items that one needs for a beginning appreciation of the images made by these techniques. In examples given below click the thumbnails to open a larger view of the image. Image windows can be re-sized or scrolled, depending on your preferences and the capabilities of your computer monitor.

The great thing about electrons is that, unlike photons, they are charged.  This gives us some control over them that we just don't get with light.  We can push electrons around (with magnets), tell them to 'go here!' 'go there!' 'hit the crystal!' 'now that one!'.  And, unlike photons, we can directly intervene to speed them up or slow them down, according to our needs, according to how we wish to use them to 'probe' crystals.

The basic instrument for controlling electrons for the purpose of generating petrographic images is the scanning electron microscope (SEM).  The SEM has 4 main parts (plus a variety of electronic gadgets for displaying and recording the data):

  1. An electron gun to generate the beam of electrons
  2. An electron column with magnets that are used to adjust the properties of the electron beam.  This beam can have a lot of electrons, or fewer, and the electrons can be highly energized (accelerated) or less so, the choices depending on the information we wish to collect and the nature of the sample being probed.  The key point is that we can (because the electrons have charge!) push/pull this electron beam around.  It can be a fixed beam, set to strike at some precise position after it exits the electron column.  AND, we also have the possibility to set the beam into continuous motion, dragging it in some set pattern across the surface of a sample.  The beam can go back and forth, back and forth, in a 'line scan'.  Or, it can follow a succession of line scans progressively down the field of view, in which case we say that the beam is 'scanning' or 'rastering'.  As the field of view (the area being scanned) gets smaller the magnification increases.
  3. There is a place for the sample to sit, usually called a 'stage', beneath the electron column.  In fact, this stage sits in a vacuum chamber because we want to control the electrons very precisely and air molecules would, frankly, just get in the way. Also, certain components get really hot (like up in the electron gun) and we don't want them reacting with oxygen molecules (burning up).
  4. Finally, we come to the part we really care about.  It would be no fun to simply sit and blast rocks with our neatly controlled and energized electron beam.  What we care about is what happens when those electrons interact with the crystals as the beam goes scanning along.  Happily for petrographers, a lot of things happen.  These 'happenings' include the generation of a wide range of radiations ("signals"), both charged (electrons) and uncharged (visible light, X-rays), each carrying information that can be interpreted to reveal properties of the crystals that are being probed.  Positioned around the vacuum chamber are detectors, each one specially designed to report to us something about the intensity of an incoming signal.  Most SEMs have 2-3 detectors.  Because the radiations are being generated in spots that we can locate very precisely by our control of the electron beam (a beam is, typically, on the order of a micron across), we can create maps that correlate the variations in intensities of the signals with positions on the specimen.  These maps are, for petrologists, petrographic images.

Click the thumbnails below to open a larger view of the image. Open image windows can be re-sized and scrolled.

Fluorescence Microscopy

Fluorescence microscopy (also called epifluorescence) is performed by bombarding the specimen with UV radiation and observing the resulting longer-wavelength emissions.  There are three main uses: 1.  The distribution of organic matter in rocks, for example petroleum inclusions, may be characterized in this manner.   2.  Fluorescent epoxy can be used for impregnation of microporosity which becomes prominently visible under UV exposure.  This technique typically reveals rocks to be more  porous than is apparent in transmitted light.  3.  Similar to cathodoluminescence, some activator elements in minerals produce UV fluorescence, allowing chemical zonation crystals to be observed. Compare images A. Plane-polarized light, B. Cross-polarized light, and C. UV epifluorescence.

(A) Grainstone in plane-polarized light (B) Grainstone in cross-polarized light (C) Grainstone in UV epifluorescence

Cold-cathode cathodoluminescence

cold-cathode example

Interactions of the beam electrons with the outer-shell electrons in the crystals can induce emission of visible light, a phenomenon called cathodoluminescence (CL). The physics of CL is complex and the intensity and color of CL can be related to the intrinsic properties of the crystal lattice, to trace the elements that have substituted for the 'normal' elements in the crystal, to crystal defects, and to temperature. In the adjacent CL photomicrograph the variations in CL color and intensity relate to variations in in Fe and Mn in the calcite crystals. The zoned calcite cement is nucleated on a grain that is centered in the lower right corner.

The cold-cathode CL method is older, perhaps more 'primitive', but still, very useful. "Cold cathode" CL observations are performed on a conventional polarizing microscope with a special vacuum chamber attached.  "Cold cathode" refers to the way in which the electrons are generated.  This simple electron gun spews a broad swath of accelerated electrons onto the sample surface (a polished thin section in this case) and the CL that is generated can be viewed directly (yes, by your very own, natural, paired photon detectors) and photographed in the usual way.  The "cold" here refers to the electron gun, not the sample.  So be careful-- the epoxy can come out looking like burned toast.

Scanned cathodoluminescence

scanned cathodoluminescence example, image by Rob Reed

This method has been explained partially above: a scanned beam of accelerated electrons excites the light emission (the detector in this case being a photomultiplier tube).

The CL signal in the SEM can be collected in ways that are either quantitative (actual spectra of intensity versus wavelength) or qualitative (images). The most basic way to handle the CL signal is to simply collect the intensity of the light to create a map of relatively dark and bright regions of the crystal. It turns out that many relatively subtle chemical differences, completely invisible in transmitted light microscopy, are glaringly obvious in CL. Thus, CL is useful for looking at chemical zoning or for discriminating rock components (say, cements and grains) made of the same mineral. In the adjacent image the CL allows us to see the prominent chemical zonation in dolomite crystals, including the ones that fill the fractures.

Secondary Electron Images

secondary electron example

Loosely bound electrons near the surface of materials are 'shook loose' (no, not really, better to say they are 'energized to the point of escape'). These are called secondary electrons and their abundance at any one spot is related strongly to the tilt of the surface.  Thus, we can use the SE signal to create topographic images.  Samples for this method can be little rock chips or thin sections or loose sediment. In this image, Fe-oxide coated microbial filaments partially fill the pores in a sandstone.


Back Scattered Electron Images

back scattered electron image

The highly accelerated electrons from the primary beam may penetrate into the sample where they are ultimately absorbed (thereby generating a variety of stimulated emissions, or, cause heating), or, they may undergo collisions with atoms that deflect their path, in some cases leading to deflections large enough to allow their escape from the surface (whereupon they are called back-scattered electrons or BSE). The probability of escape is related to the atomic weight of the material, heavier atoms, causing more deflections and a greater probability of escape. Hence, the BSE signal is used to create gray-scale images that are a depiction of relative atomic weight (darker=lighter elements; brighter=heavier elements). In the image at right crystals in a dolostone display prominent zonations that are related to variations in Fe and Ca.

We call these 'compositional images'. Samples must be polished thin sections. A key point concerning BSE images is that the signal creating them comes from within a few nm below the specimen surface. Compare that to the 30-micron thickness of a standard thin section. For this reason, BSE images show textural relationships between rock components with startling clarity (startling, at least, to anyone accustomed to getting their petrographic information from transmitted light images). BSE images are particularly useful for very fine-grained rocks.

X-ray Maps

X-ray map, image by Rob Reed

The electron beam also generates X-rays by interacting with the inner shell electrons of the elements present in the material. X-rays generated in this manner are called 'characteristic X-rays' because their wavelength and energies can be related specifically to the elements from which they come (those inner electron shells, and the energies they can generate when whacked by an accelerated beam electron, being pretty specific to the energy level of the shell in which they reside, and hence to the element...). For instance, there are silicon X-rays, calcium X-rays, and so forth, through the whole periodic table. Well, this is something useful indeed! With an X-ray detector staring patiently at our sample, we don't have to ponder, "Well, that sorta hexagonal-looking crystal looks like it could be quartz". Instead, we can stop the scanning, place the beam squarely on the crystal of interest, tell the detector to 'Zap!', and we can find out. Only Si-X-rays? (most detectors cannot detect elements lighter than F, though many of modern ones reach down to C). Well, then, yes, it's quartz! Do we see also, Na-and Al-X-rays? Well, maybe we're dealing with albite or some sort of zeolite. The point is, X-ray detection can very quickly boost us out of the realm of speculation, and into something more like 'science'.

X-rays can contribute a lot to our understanding of minerals in just the sort of qualitative way described, but we can also create images of X-ray intensity for particular elements. This allows us to not only recognize the presence chemical zoning (as with BSE and CL images) but to specify which elements in particular are varying. In the image above we can see zonation in a dolomite crystal that is similar to that seen in BSE (above) but here we learn that, indeed the zonation relates to Fe (shown in red). Fe-poor parts of the crystal are green.

Very importantly, X-rays also can be used in a quantitative way, moving beyond petrography to actual elemental analysis. To sum up simply, the intensity of particular X-rays can be compared to the intensities of the same X-rays in materials of known composition (standards) and an actual analysis of considerable accuracy and precision can be obtained. Special SEMs (electron microprobes), fitted with special X-ray detectors, are devoted to this type of analysis. Not surprisingly, a wide array of different imaging detectors (SE, BSE, CL) are routine components of microprobes and provide vital guidance to the operator concerning which points on the specimen are desirable for analysis.