By George Vander Voort

Metallographic Techniques

Figure 1: Vagn Buchwald with a section from the 20 ton Agpalilik fragment of the Cape York meteorite, found on 31 July 1963 in Greenland, at the Geological Museum in Copenhagen in 2005 (Courtesy of Vagn Buchwald).

Techniques that have been developed for iron and steel specimens are directly applicable to meteorites. Sections must be removed from the parent mass with as little damage as possible. This may be difficult to achieve when very large masses must be sectioned (see Figure 1 and also Part I of this article). The gross macrostructure of meteorites can be very dramatic, as shown in Figure 2. Octahedrites obtained their name from this very striking macrostructural growth pattern of the kamacite (ferrite) phase, which is visible to the unaided eye. Octahedrites generally contain from ~5 to 10 weight percent Ni. The as-solidified microstructure is FCC taenite (austenite). With subsequent cooling, kamacite nucleates on the prior-taenite grain boundaries producing the Widmanstatten pattern shown in Figure 2. Determination of the prior-taenite grain size has been rarely donwe due to the need for exceptionally large specimens. Vagn Buchwald, in his study of the 20,140 kg Agpalilik Cape York meteorite (Figure 1), measured a prior-taenite grain size of ~2 X 1.5 X 1.5 m – yes, meters, not millimeters!

Cutting does deform the microstructure adjacent to the cut surfaces.  Heat is also generated during cutting and this can alter the microstructure.  Thus, a cutting procedure that produces as little damage as possible must be chosen and properly implemented. After sectioning, small pieces may be embedded within a polymer compound for convenience in holding, either by hand or within a fixture. Automated grinding and polishing equipment has become very common in metallographic laboratories and some machines can accommodate unmounted bulk specimens, even with rather large sizes.

Figure 2: Macrostructure of Spearman (Texas), a medium octahedrite, showing a pattern of elongated kamacite grains and some terrestrial corrosion at cracks.

In most work, a surface is selected for the examination and it is subjected to a series of grinding operations with progressively finer abrasives.  Fixed abrasives, as for SiC grinding paper, is most commonly utilized. In the traditional approach, grinding progresses from a reasonably coarse SiC particle size, e.g., 120-grit, down to a 600 (P1200) grit finish, that is, a SiC particle size of about 15 μm. Modern procedures used by the writer use one or two SiC steps because sectioning was done with a device, such as the Secotom saw, that produces minimal damage. The first SiC step is with an abrasive that will get the specimens in the holder co-planar with minor effort, while removing the sectioning damage. But, the grit size is not so coarse that the grinding introduces more damage than cutting did. Generally, a grit size of 180 is adequate. A second grind with 220- or 240-grit SiC is adequate.

After the various grinding steps are completed, specimens are polished using a series of progressively finer abrasives.  The most common coarse polishing abrasive is diamond.  One or more steps are used, e.g., 9- and 3-µm, and occasionally 1- µm diamond. Generally, diamond is applied to a cloth covering a rotating wheel.  Numerous types of cloths may be utilized.  Generally, for coarse diamond sizes a low-nap or napless cloth is used, such as a hard-woven polyester, silk or nylon.  Fine diamond sizes may be used on a low-nap cloth or a synthetic, napless cloth.  Diamond abrasives finer than 1 µm in size are not commonly used.

Figure 3:  A patch of plessite, a mixture of kamacite and taenite, surrounded by kamacite in Odessa, a coarse octahedrite that fell in Texas, viewed with DIC at 400X in the as-polished condition Magnification bar is 25 μm long).

Final polishing may involve one or more steps, depending upon the material being polished, equipment availability, number of steps with diamond abrasives, personal preference, etc.  Aqueous slurries of graded α- or γ- alumina (0.3- or 0.05-μm diameter, respectively) may be used.  The writer prefers to final polish specimens using colloidal silica, a solution with a pH of about 10 containing very fine (~0.05 μm), spherical amorphous silica particles, or with agglomerate-free 0.05-µm diameter alumina made by the sol-gel process; a neutral pH (as with OP-AN) helps minimize staining problems.  Final polishing can be done using a rotating wheel.

After polishing, specimens should be examined with the reflected light metallurgical microscope to evaluate the preparation and detect any constituents that are “dark” without etching.  Incident bright-field (BF) illumination is most commonly employed. Without etching, very little can be observed of the microstructure, unless differential interference contrast (DIC) illumination is employed. BF examination of the as-polished specimen will detect cracks, terrestrial corrosion and certain second phases (graphite, sulfides, silicates, etc.).  If a small amount of relief is developed during polishing, DIC can reveal the plessite constituent, Figure 3, and phosphides and carbides very well.  This shows plessite in Odessa, a coarse octahedrite that fell in Texas (meteorites are named after the town or post office closest to the fall site). Octahedrites are classified by the width of the elongated kamacite grains. Spearman, shown in Figure 2, is a medium octahedrite while Gibeon, Figures 3 and 4, is a fine octahedrite. Kamcite forms from taenite in meteorites along the octahedral planes. The kamacite growth rate is exceptionally slow – estimates vary from 1 to 250 °C per million years!  So, no graduate student can duplicate these microstructures in the lab! Length-to-width ratios are typically in the range of 10 to 30.

In nearly all cases, etching is needed to properly reveal the microstructure, just as for most man-made metals and alloys.  This is essentially a controlled corrosion process.  Etchants may be acidic or basic and they may contain various chemicals designed to increase the attack rate.  The corrosion rate of metals varies with crystal orientation.  In a polycrystalline metal, the crystal planes within each grain have a specific orientation and this orientation differs from grain to grain.  Because corrosion rates are sensitive to crystal orientation, each grain is dissolved at a different rate which, if controlled, permits the grain structure to be revealed.

Figure 4:  An example of “finger” plessite in Gibeon, a fine octahedrite that fell in Southwest Africa using: a) 4% picral; and, b) 2% nital (200X, magnification bars are 50 μm) Picral reveals only the phase boundary between the taenite and kamacite.

In like manner, the corrosion rate at a grain boundary is different, usually higher, than within the grain due to energy differences.  One type of phase may be attacked at a different rate than another.  The presence of alloy or impurity segregation or variations in residual or applied stresses also influences the local corrosion rate.  All of these factors enable an etchant to reveal microstructural details.  Polarized light has rather limited applicability in the study of meteorites as only a few constituents are optically anisotropic.

Figure 5:  A large patch of taenite, surrounded by kamacite (K), containing a thin outer layer of “clear” (retained) taenite (CT-1), a dark etching “cloudy zone” (CZ), an inner region of “clear” (retained) taenite (CT-2) and martensite (M) in the central region in the Canyon Diablo meteorite, a coarse octahedrite that fell in Arizona (etched with 2% nital, 400X).

The most commonly used chemical etchants for meteorites, which are also the most commonly used etchants for steel, are nital and picral.  Nital is a solution consisting of 1 to 4% nitric acid in ethanol.  Picral is a solution consisting of 4 g picric acid dissolved in ethanol.  Both etchants dissolve the kamacite (ferrite) phase preferentially to other phases.  Nital, however, is quite sensitive to the crystallographic orientation (as with steels) of the kamacite while picral is not.  Hence, picral dissolves the kamacite uniformly while nital does not.  Nital will reveal kamacite grain boundaries but picral will not.  Picral does not reveal the Neumann bands or as-formed martensite.  Nital reveals both well.  Nital is preferred for the overall study of the microstructure of meteorites although picral is good for examining the plessitic structures.  As an example, Figure 4 shows views of the same patch of plessite in Gibeon, a fine octahedrite that fell in Southwest Africa.  Picral, Figure 4a, reveals the very fine taenite (austenite) particles within the patch.  Note that there is a thin band of white taenite around the outer edges of the patch.  Just inside the edge is a dark etching zone of “black” plessite, a very fine mixture of kamacite and ferrite.  Within the plessite patch we observe very fine particles of taenite in a kamacite matrix.  Etching with nital, Figure 4b, reveals grain boundaries within the kamacite matrix of the plessite which obscures the fine taenite particles.  Note that some Neumann bands and some subgrain boundaries can now be observed in the kamacite surrounding the plessite. We should point out that kamacite and taenite were named by Reichenbach in 1861 – 2 years before Sorby first observed  metallographically prepared steel and iron specimens.

Figure 5 shows an example of a large patch of taenite between kamacite (K) grains where the nickel content was low enough in the center for martensite to form.  At the very edge of the taenite wedge there is a thin layer of “clear” (retained) taenite (CT-1) where the nickel content is about 50% – this is an ordered Fe-Ni phase.  Below, this is a dark etching band called the “cloudy zone” which actually contains two phases, ordered Fe-Ni with ~50% Ni and a very fine martensite with ~12% Ni (only resolvable with the transmission electron microscope). The bulk Ni content is ~28-30%.  The cloudy zone (CZ) has a bluish/brownish color when etched with nital. Between the cloudy zone and the martensite central area is a second region of “clear” (retained) taenite (CT-2) where the nickel content continues to decrease.  Martensite is not always found in the wedge-shaped patches of taenite. It is usually seen in large taenite patches, greater than 50 μm in diameter, and in areas where the nickel content is <25%.

Figure 6: Cohenite (C) and schreibersite (S) surrounded by shock-annealed α2 type kamacite etched with 2% nital (a – bright field and b – Nomarski DIC) and with alkaline sodium picrate at 8 V dc for 60 s (c) to color the cohenite.

Carbides can be observed in some Fe-Ni meteorites, but not all. The most commonly observed carbide is called cohenite (no, it was not named after the late Morris Cohen) with the general formula (Fe, Ni, Co)₃C. It is similar to M₃C carbide in steels, called cementite, but Ni and Co are never seen in cementite in steels. The M stands for (Fe, Mn, Cr) in cementite. An alloy carbide called haxonite has been detected in a few meteorites. Figure 6 shows an example of cohenite in Canyon Diablo, a coarse octahedrite that fell in Arizona.

Figure 7: Rhabdites in Canyon Diablo revealed by electrolytic alkaline sodium picrate, 6 V dc, 40 s (magnification bar is 50 μm).	Figure 8: Sphalenite in an as-polished specimen of Canyon Diablo revealed by its natural color (magnification bar is 25 μm).

Phosphides are frequently observed in meteorites. Canyon Diablo contains 0.26% P and Figure 6 shows schreibersite with internal cracks. The phosphides are tetragonal with the general formula (Fe, Ni)₃P. Schreibersite is the name for the more globular shaped phosphides and their size is a function of their P content. The Ni content also varies depending upon the temperature of the meteorite when they nucleated. High Ni content schreibersite particles tend to be smaller and more ductile while low Ni content particles, such as this one, usually have low Ni content and are often cracked. Rhabdites are phosphides that are small and have plate-like or prismatic shapes, as shown in Figure 7. The prismatic-shaped rhabdites are usually crack-free while the plate-like particles often are cracked. Figure 8 shows sphalerite, ZnS with a cubic crystal structure, also detected in Canyon Diablo.

Comments

For the absolute best source of information about meteorites, look for copies of Vagn Buchwald’s three volume masterpiece, Handbook of Iron Meteorites published in 1975 by the Center for Meteorite Studies, Arizona State University and the University of California Press. The next part in my presentation on Fe-Ni meteorites will be devoted to the use of color etching to reveal the microstructure of meteorites.

George Vander Voort has a background in physical, process and mechanical metallurgy and has been performing metallographic studies for 45 years. He is a long-time member of ASTM Committee E-4 on metallography and has published extensively in metallography and failure analysis. He regularly teaches MEI courses for ASM International and is now doing webinars. He is a consultant for Struers Inc. and will be teaching courses soon for them. He can be reached at 1-847-623-7648, EMAIL: georgevandervoort@yahoo.com and through his web site: www.georgevandervoort.com

To View a listing of all George’s articles please click here

Read George Vander Voort’s Biography