George F. VanderMetal Physics Research, Carpenter Technology Corporation, Reading, PA 19612
Meteorites are particularly fascinating subjects for metallographic examination, not merelybecause of their extraterrestial origin, but also because of the rich variety of microstructuralconstituents encountered. While their structure can be assessed use of the familiarnital and picral etchants used extensively for steels, application of selective etchants,including those that produce selective coloration, greatly increases the amount of information obtained by light microscopy. The article describes efforts to preferentially darkenor color microstructural constituents in iron meteorites: kamacite (ferrite), Neumannbands (mechanical twins), taenite (austenite), cohenite (Fe-Ni carbide similar to cementite), martensite and schreibersite/rhabdite (Fe-Ni phosphides). The influence of concentration gradients, deformation, and reheating on the structure can be clearly observedby the use of tint etchants.
iron portion of meteorites, as inthe iron meteorites, is opaque to andrequires reflected-light studies.
The classification of iron meteorites is acomplex Classification is based onmacrostructural and microstructural characteristics or chemical ,-,,..,.,.,,.. -e-,,..,.,,i.-ir'n
Ga, Ge, Ir contents). The two "UQYIPn,,,
tunately. are consistent. In terms,there are three categories of iron meteo-rites-hexahedrites, andataxites. Those that do not thesethree groups are cate-gory-anomalous.
The hexahedritescrystals of kamaciteboundaries aresubgrain boundaries are common. Theirstructure may be alteredcosmic whichthe kamacitegrains. Most contain Ni in therange of about 5.2-5.8%.
Octahedrites received name based
INTRODUCTION
Meteorites have fascinated engineers andscientists for many years. Research on meteorites began with mineralogists and geologists and the major phases and constituents in iron meteorites were identifiedand named prior to Henry Clifton Sorby'sfirst observation in 1863 of the microstructure of polished and etched manmade metals. Iron meteorites are but one of severalgeneral categories of meteorites.Fe-Ni alloys with nickel contents of4.3-34%, although few have levels >20%;and most are in the range of 5-10% Ni.Small amounts of cobalt 0.4-1.0%), sulfur (up to about phos-phorus (up to about 0.3%), and carbonto about 0.2%) are usually present, whiletrace levels of many other elements are detectable. There are also numerous types ofstony and stony-iron meteorites. The stonymatter can be studied using reflectedor transmitted light on thin sections. The
on their macrostructural characteristic, visible to the unaided eye afteretching, caused the of kamaciteon the octahedral of the n1"'ln1"_.t-",~'n_
ite (i.e., The octahedritescontain of Ni than the hex-ahedrites, in the range ofalthough a have Ni contents.The kamacite in octahedrites con-tains more Ni in hexahedrites, 5.5-7.5% versus 5.0-5.5%. There are six basictypes of octahedrites; five are based on thewidth of the kamacite grains. As the Nicontent increases, the kamacite width decreases. Thus, octahedrites are classed infive steps from coarsest (lowest Ni) to finest (highest Ni) and then plessitic, wherethe kamacite width is <0.2 rnm, and theyexist in discontinuous short fibers or spikeswith an octahedral orientation. The mi-
(a)
(c)
G. F. Vander Voort
crostructure may also be altered cosmicshock and reheating events, as well asterrestial corrosion.
Ataxites contain higher Ni contents thantypically about 15-18%, but
do not exhibit gross macrostructuraln-""r'TH~·h patterns. Instead, their macrostructure is basically featureless. Relativelyfew ataxites have been found. The kamacite are equiaxed and small in size,typically <30/-Lm in diameter.
MICROSTRUCTURAL FEATURES
In 1861, Reichenbach named the two majorphases in iron meteorites, kamacite andtaenite, and their mixture, plessite. Kamacite is the body-centered cubic ex phaseknown to metallurgists as ferrite, while
(b)
FIG. 1. Examples of plessite morphologies: (a) combplessite in Odessa (4% picral); (b) pearlitic plessite inOdessa (unetched, DIC); (c) comb and spheroidizedplessite in Odessa (unetched, DIC); (d) comb, pearliticand spheroidized plessite in Odessa (2% nital): (e)pearlitic plessite in Arispe (4% picral): (f) net (right)and cellular plessite in Gibeon (2% nital): (g) blackplessite in Gibeon (4% picral): (h) acicular plessite inCanyon Diablo (4% picral); (i) altered plessite in BoxHole (4% picral).
Metallography of Fe Meteorites
taenite is the face-centered cubic phaseknown to metallurgists as austenite. Plessite is a mixture of kamacite and taenite(ferrite and austenite) that occurs in meteorites but not in manmade, iron-based
(d)
(h)
225
alloys. Plessite 1) occurs in numerousconfigurations two that resembleferrite-cementite mixtures in steels, thatis, pearlitic and spheroidized plessite. Note that meteorites are named after
(e)
(g)
FIG. 1. (Continued)
226 G. F. Vander Voort
grains as films or as triangular patches("wedges") at kamacite junctions. The Nicontent within taenite is nonuniform [2,despite the extraordinarily slowrate of meteorites (typically from 1 to 250°Cper 106 years between 700 and 400°CDetermination of the prior-taenitesize has rarely been done, mainlyone must have an exceptionally largeimen. Buchwald [1], in his study the20,140kg Agpalilik Cape Yorkfound a prior-taenite grain sizex 1.5 x 1.5m (yes, meters!).
The structure of the taenitegenerally quite complex and verying (Fig. Starting at the kamacite-facnite interface and moving inward, observe that etching with nitaI reveals a seriesof zones. First, we see a very narrow zone,1-2/-Lm wide, called clear taenite 1which contains 51.4-45.6 ± 1.3% Ni andis ordered Fe-Ni [5]. Next is abrownish appearing zone, up to about25/-Lm wide, called the "cloudy zone"
the place where they fell. Table 1 lists details of the meteorites whose microstructures are shown in this article.
Neumann bands, that is, mechanicaltwins, are observed in kamacite, be it inhexahedrites, octahedrites, or ataxites (aslong as the kamacite is more than about20/-Lm wide [1]). Because hexahedrites aresingle crystals, Neumann bands can bequite long; up to several centimeters is notunusual. The vast majority are formed byextraterrestrial collisions, not by their impact with the earth. The band width is usually narrow, typically 1-10/-Lm (Fig.
Plessite is observed in octahedriteswhere it becomes more common as the Nicontent increases. The morphology of theplessite mixture also changes with Ni content, and a variety of forms has been observed and classified (creating jargon torival anything developed by metallographers!). Also in octahedrites, taenite can beobserved between adjacent kamacite
Coahuila Mexico HexahedriteGibeon Southwest Africa Fine
OctahedriteHenbury Northern Territory, Medium
Australia OctahedriteMundrabilla Western Australia Medium
OctahedriteNorth Chile Tocopilla, Chile HexahedriteOdessa Texas, USA Coarse
OctahedriteToluca Mexico Coarse
OctahedriteWashington Colorado, USA Ataxite
a Meteorites are named after the place where the fall occurred.
(a)
Metallography of Fe Meteorites
K
,NtQOfLm-
(b)
(b)
(c)
2. Examples of Neumann bands (mechanicaltwins) in meteorites: (a) Neumann bands (N),subgrain boundaries (sb), piessite (P), taenite (T) andkamacite (K) in Gibeon (2% nital); (b) Neumann bands
Odessa (2% nital); (c) Neumann bands intersectingprismatic and plate rhabdites in North Chile (2% nital,DIC).
FIG. 3. Examples of taenite wedges and martensitemeteorites: (a) taenite wedge in Odessa showingthin CT-l zone at the a--y interface, the dark CZ,terior retained taenite (CT-2) and martensite (verylight because no tempering has occurred) (2% nital);(b) taenite wedge in Canyon Diablo showing the thinCT-l, the dark CZ, retained taenite (CT-2) and darketching martensite in the interior (2% nital); and (c)acicular piessite in Arispe containing martensite (4%picral).
228 G. F. Vander Voort
rites. The most common is called cohenite which is orthorhombic and
identical to cementitecohenite contains Ni
and cobalt accord-to Scott (quoted by Buchwald [1]), and
has the formula (Fe,Ni,CohC. It appearswhite when being quitehard 1100 it will stand in relief.
an alloy carbide with a cubicstructure and the formula
(Fe,Ni,Coh3C6 , has been observed in onlya few meteorites. It is smaller in size than
slightly less hard, and also whitewhen When one considers the
slow cooling rates that meteorites are subject to, rates as dose to equilibrium conditions as one can ever hope toget, it seems surprising that carbideswould form rather than graphite. Graphite
where spinodal uecomposmonvery fine,ordered Fe-Nia surroundingNi) of martensiteserve a zone of aboutthe same width called the dear taenite 2(CT-2) zone 25.8-28.1 %which may be ordered [5]. The struc-tural details in these three zones can onlybe observed by transmission electron mi-croscopy. In the interior where thenickel content is martensite plate-lets are observed within a retained taenitematrix. These platelets may be quite coarseand can be observed at low magnificationsby light microscopy. the taenitewedge must be at least wide to pro-duce martensite platelets.
Carbides can be observed in iron meteo-
(a) (b)
(c) (d)
FIG. 4. Examples of cohenite in meteorites-(a-c): cohenite (C) and schreibersite (5) in Canyon Diablo (2% nital):(d) cohenite and schreibersite in Arispe (4% picral).
Metallography of Fe Meteorites
is observed in meteorites, but cohenite ismore common.
Phosphorus is present in meteorites inlevels much greater than in steels. Hence,it is not surprising that phosphides are acommonly observed meteoritic constituent. Although the phosphides are all tetragonal with the formula (Fe.Nij-P, twotypes are defined, schreibersite and rhabdite. Both were identified and namedmineralogists in the mid-19th Century.
Schreibersite (Fig. 5) is yellow to brownish when polished and more brittle thancohenite, although slightly less hard. The
(a)
(b)
FIG. 5. Examples of schreibersite in meteorites: (a)brittle (note cracks) schreibersite in Odessa(unetched): (b) small schreibersite with
Ni content between two taenite grains in Toluca (4%
229
Ni content is rather variable. When thephosphorus content is above about 0.06%,very small schreibersite precipitates are observed, frequently along subboundariesand a-a or a-"{ phase boundaries. As thephosphorus level increases, the particlesget larger. Because of its brittleness, schreibersite usually exhibits extensive cracking.These massive particles contain 10-15%Ni, and they precipitate from the taenite.Particles that precipitate later, generally ata-a or a-"{ phase boundaries, contain 2050% Ni, are smaller in size, more ductile,and free of internal cracks.
Rhabdites (Fig. 6) exhibit prismatic orplatelike shapes. are commonly observed in hexahedrites and many octahed-
(a)
(b)
FIG. 6. Plate (P) and prismatic-shaped rhabdites innorth Chile: (a) note the difference in appearance ofthe Neumann bands (N) compared to the plate rhabdites (2% nital): (b) note that the twins have fracturedthe intersected prismatic rhabdites (arrow) (2% nital,DIC).
230
FIG. 7. Daubreelite inclusion with a chromite rim inBox Hole (unetched).
rites, particularly those with rather coarsekamacite grains. The particlesgenerally exhibit numerous transversecracks, while the prismatic particles areusually crack free. Rhabdites will becracked by mechanical twins in the kamacite that intersect the rhabdites, as illustrated in Fig.
A number of other mineral phases canbe observed in such as FeS(troilite), CrN (carlsbergite), graphite, diamond, FeCr2S4 (daubreelite, Fig. 7),FeCr20 4(chromite, Fig. silicates (Fig. 8),ZnS (sphalerite, Fig. and so forth. Theseare less commonly observed, but they canbe readily identified.
FIG. 8. Silicates next to black plessite in Gibeon (4%picral).
G. F. Vander Voort
FIG. 9. Sphalerite in Canyon Diablo (unetched).
METAllOGRAPHIC TECHNIQUES
specimen preparation ofiron meteorites is basically identical to thatfor steels. The major problems areelimination of polishing scratches, controlof preparation-induced deformation in thekamacite, and control of polishing relief associated with the hard cohenite or schreibersite. Rhabdite platelets and schreibersite are rather brittle and exhibit extensiveinternal cracking. Regardless of the caretaken in specimen preparation, thesecracks will always be present.
Nital is, far, the most common etch-ant used to reveal the microstructure ofmeteorites. Concentrations from 1 to 5%have been utilized. Perry [6], however,preferred a 5% picral solution (4% is morecommonly used) over a 5% nital solutionbecause its action was slower and morecontrollable. However, a 5% nital solutionis rather strong, and most nital users prefer1 or 2% solutions because they are sloweracting. Perry also reported use of equalparts of nitric and acetic acids for etchingNi-rich ataxites. Many of the etchants usedfor maraging steels would also be satisfactory for the ataxites.
Perry [6] claimed that the best etchantfor identifying schreibersite (and all P-richareas) is a neutral sodium picrate solutionused boiling. The solution consists of 2gpicric acid dissolved in an aqueous solutionof 25g sodium hydroxide dissolved inlOOmL of water. This is the familiar alka-
Metallography of Fe Meteorites 231
EXAMPLES
Small specimens of a of meteorites(see Table 1) were mounted and polishedusing techniques identical to those forirons and steels Specimens wereground 600-grit SiC paper, usingwater as the They were then pol-ished using a medium nap synthetic suedecloth and 3- and diamond abrasivefollowed by an aqueous slurry of 0.05/l-mde agglomerated alumina. This was fol-lowed by polishing with 0.035/l-mcolloidal silica The general structureswere assessed etching with either 4%
or 2% nital. Picral is excellent for revealing the structure of plessite and for thestudy of rhabdites. 10 comparesplessite as revealed with 4% pi-cral or 2% nital. Picral has revealed thetaenite globules and with great clar-ity. Nital out the taenite but also a-a grain and boundaries and theNeumann bands. Picral does not bring outthe Neumann but can be ob-served differential interference con-trast either as polishedor after DIC is useful for ex-
gation, and are of considerable value to themetallographer andKreye used tint etchants in their studyof the Coahuila and Gibeon meteorites.Color contrast is observed with bright field
in some cases, contrast can befurther enhanced use of polarized light(plane polarized or by slightly off-crossed polarized Coloration is pro-duced by the of an interference filmover either matrix phase, the secondphases, or all of the structure, that is, an-
cathodic, or reagents. Thefilms consist of an sulfide, complexmolybdate, elemental or chro-mate. The purpose of this article is to demonstrate the value of certain tint etchantsas compared to the use of the very com-
used nital or reagents. Resultswith a few other selective etchants will alsobe presented.
line sodium picrate solution. statesthat the solution can be neutralized butdoes not say what to use.states that neutral sodium picrate is madeby adding a saturated aqueous acidsolution to a concentrated caustic soda solution. A yellow precipitate of sodium picrate forms, which is collectedand dried. (Dry picrate is an '..J'"f../J..'-'u....
it must be handled with A 1%aqueous solution is prepared and neutralized by adding aqueous picric acid.[6] says that this solution can be madewithout going through theprocess, but his details are mC'OII1D!letedoes say that if the solution is it is"ineffective;" if it is it will colorboth carbide and phosphide, while if it isneutral, it only colors the
To darken cohenite, Perry used boil-ing alkaline sodium picrate (4g, rather thanthe usual 2g, picric acid is added to a100mL aqueous solution containing 25gNaOH). The solution must be freshlymade. It can be used electrolytically atroom temperature at 6V dc. Berglundstates that alkaline sodium willdarken phosphides as well as cementite.
Perry [6] reported limited use of Murakami's reagent. Numerous versions ofMurakami's reagent have been usedbut employed the standard mixture:109 potassium ferricyanide andsium hydroxide in 100mL of water. Hestates that cohenite is colored after 5-10simmersion in a boiling solution, while thephosphides are unaffected. Perrythat the surfaces etched this wayexhibited scratch patterns, regardless ofthe care he took in polishing.
Dorfler and Hiesbock utilized vapor-deposited ZnSe, the interfer-ence layer method var-ious constituents in meteorites. Little difference is observed between kamaciteand taenite (pink), but the second phasesand minerals exhibit more distinct colordifferences.
Tint etchants produce color variations asa function of ori-
and segre-
232
(a) (b)
G. F. Vander Voort
(c) (d)
FIG. 10. Comparison of plessite revealed using 4% picral (a and c) or 2% nital (b and d) in Gibeon. Note thatnital brings out the a-a grain boundaries within the plessite and the Neumann bands, which picral does notreveal.
amination of picral etched specimens asadditional information may be obtained;Fig. 11 shows such an example. In somemeteorites, if nital is used, there are somany Neumann bands visible that it is difficult to assess the rhabdite content. Consequently, seeing less of the overall structure can be beneficial in certain instances.
Both picral and nital bring out the CT-l,CZ, and CT-2 zones within taenite wedges,and in some taenite strips. When martensite is present within the wedge, nital usually reveals its structure better unless themeteorite experienced some degree ofcosmic reheating after the martensiteformed, that is, some tempering has occurred. After the general structure is assessed, the etch is removed, and vibratorypolishing is repeated. Sometimes it takes a
bit of effort to polish the kamacite free ofpreparation-induced deformation andscratches. Vibratory polishing will eventually remove these problems, so long asthe prior steps were properly executed.This generally does introduce substantialrelief at cohenite, rhabdite, and schreibersite particles, and generally producesenough relief to view plessite with DIC inthe as-polished condition, as shown in Fig.
and c).The writer has tried a number of tint
etchants and selective etchants on meteorite specimens. This study has not beencompleted, and only a limited number ofmeteorites have been examined, so the results must be viewed as preliminary.Nevertheless, the results have been quite
and more work of this type should
Metallography of Fe Meteorites
Received June 1992; accepted June 1992.
lective etching, in AppliedMetallography, Nostrand Reinhold Co., Inc., New York (1986), pp.1-19.
14. E. Hornbogen and H. The ml,rrr,,,,h"l1r+l1"O
of two iron meteorites (Coahuila andZeitschrift fur Metallkunde 61:914-923 (1970).
15. G. Hallerman and M. L. Picklesimer,ence of the of metal specimensprecisioniron Probe Microanulueis
Marton (eds.), Acauemicpp. 197-225.
8. G. F. Vander Voort, Metallography: Principles andPractice, McGraw-Hill Book Co., New York (1984).
9. G. Dorfler and H. G. Hiesbock, Investigations ona quantitative mineralogical characterization ofmeteorites by modal analysis, Meteorite Research,Springer-Verlag, New York (1969), pp. 669-682.
10. H. E. Buhler and H. P. Hougardy, Atlas of Interference LayerMetallography, Deutsche Gesellschaftfur Metallkunde, Oberursel, Germany (1980).
11. E. Beraha and B. Shpigler, Color Metallography,American Society for Metals, Metals Park, OH(1977).
12. G. F. Vander Voort, Tint etching, in MetalProgress127:31-33, 36-38, 41 (March 1985).
13. G. F. Vander Voort, Phase identification by se-
