Chemical Composition of Meteorites

I suggest to people who think that they have found a meteorite to get a chemical analysis by Actlabs or some other lab that can provide good compositional data. There are several different ways to determine whether or not a rock is a meteorite. A chemical analysis is a good one because it is cheaper to do than most of the other tests and it is usually unambiguous (meaning, with a chemical analysis, I am not likely to say, “I still don’t know” or “maybe.”)

So that you can check your data yourself, I show plots here of concentrations or ratios of concentrations of several chemical elements in meteorites compared to rocks people have had analyzed by Actlabs or some other lab. The horizontal axis of all the plots is “Fe2O3(T) + MgO.” Actlabs and most labs that analyze rocks reports total iron as Fe2O3 because in Earth rocks much or most of the iron occurs as Fe(III), that is, ferric iron. There is little or no Fe(III) in freshly fallen meteorites; it is all Fe(II), ferrous iron, and Fe(zero), iron metal. For convenience, however, I use Fe2O3 in the plots. If you had an analysis done, just add the Fe2O3 and MgO values together for comparison.

Many terrestrial rocks have higher concentrations of SiO2 (silica) than any meteorite because they contain quartz and meteorites do not contain any significant amount quartz or other silica minerals. If SiO2 is greater than 60%, the rock is not a meteorite. The only possible exception would be a lunar granite, which is a volumetrically insignificant component of the Apollo collection, and no granitic lunar meteorites have been found yet. The low-SiO2 meteorites are all carbonaceous chondrites. Carbonaceous chondrites have a substantial proportion of volatile elements that, in terms of this plot, act as diluants that drag the compositions toward the origin. The low-SiO2 terrestrial rocks are mainly limestones [low Fe2O3(T) + MgO] and iron ores [high Fe2O3(T) + MgO]. Terrestrial rocks with 45-55% SiO2 are usually basalts.

Most meteorites have low concentrations of aluminum (Al2O3) compared to terrestrial rocks. The exception is meteorites from the Moon because most lunar meteorites contain a high abundance of plagioclase feldspar. For an explanation of why lunar meteorites plot along the diagonal trend, see How Do We Know That It’s a Rock From the Moon? Feldspars are common in terrestrial rocks, too.

Lunar meteorites also have high concentrations of calcium (CaO) compared to most terrestrial rocks (except, of course, those rocks rich in calcite, like the limestones that plot around 50% CaO). The terrestrial rocks that plot with the martian meteorites are probably basalts, which have similar mineralogy to the martian basalts (but, see Na2O and K2O).

In most meteorites, virtually all the Al2O3 and most of the CaO is carried by feldspar, so there is little variation in CaO/Al2O3. Terrestrial rocks tend to have lower CaO/Al2O3 than meteorites.

Among meteorites, MgO necessarily increases with Fe2O3(T) + MgO. Some terrestrial rocks lie off the trend because they have MgO/Fe2O3(T) ratios outside the range for meteorites (below). High-MgO terrestrial rocks are mainly ultramafic rocks like dunites and serpentinites.

The same is largely true for Fe2O3(T). High-Fe2O3 terrestrial rocks are iron ores, usually hematite concretions.

The ratio of magnesium to iron does not vary greatly among most kinds of stony meteorites. Rocks with MgO/Fe2O3(T) <0.2 or >6 are not probably meteorites. Only rare lunar granites have low MgO/Fe2O3(T), and none of these has been found as a meteorite.

The ratio of manganese to iron does not vary much within different groups of meteorites. The MnO/Fe2O3(T) ratio (or Fe/Mn or FeO/MnO) is not as reliable a test for meteorites as some sources imply, although ratios of <60 are inconsistent with lunaites and >50 are inconsistent with eucrites, for example.

Most terrestrial rocks are richer in Na2O than any meteorite. Rocks with >2% Na2O are probably not meteorites.

Sodium and potassium are both alkali elements, and all alkali elements are in low concentrations in meteorites compared to most terrestrial rocks. Rocks with greater than 0.6% K2O are probably not meteorites.

One of the best elements for distinguishing meteorites from Earth rocks is Cr. Nearly all stony meteorites have high concentrations of Cr compared to most Earth rocks. The most Cr-poor meteorites are the feldspathic lunar meteorites, which have high concentrations of Al2O3 and CaO. Cr-rich terrestrial rocks tend to be ultramafic rocks like dunites, peridotites, pyroxenites, and serpentinites. Actlabs has a detection limit for Cr of 20 ppm. I have plotted all Actlabs data reported as “<20 ppm” at 10 ppm.

Notes, Caveats, and References

1) Terrestrial – Meteorwrong.  All the “meteorwrongs” in the plots (large white circles) represent rocks that people have had analyzed by Actlabs or some other lab and for which people sent me the data. If you obtain a chemical analysis of your rock, please send me the numbers!

2) Terrestrial – Geostandard.  Many countries have agencies that pulverize large quantities of rock for use as interlaboratory standards. Several hundred geostandards are available that represent all common, and many uncommon, rock types of the earth. For most of these, there are many analytical data available. I have selected from the compilation of Govindaraju (1994) and Korotev (1996) data for 156 such rocks. I have avoided data for soils and unconsolidated sediments, monominerallic rocks (except chert, sandstone, limestone, hornblendite, magnesite), ores (except for some iron ores, because these are sometimes mistaken for meteorites), and geostandards that do not have data for the elements that I plot here. In total there are data for 7 andesites, 5 anorthosites, 17 basalts, 2 carbonatites, 1 chert, 6 diabases, 2 diorite or diorite gneiss, 2 dolomites, 4 dunites, 15 gabbros, 21 granites and related rocks, 5 granodiorites, 1 hornblendite, 6 iron ore or iron formation rocks, 2 kimberlites, 1 quartz latite, 10 limestones, 2 lujavrites, 1 magnesite, 1 monzonite, 1 norite, 3 peridotites, 1 pyroxenite, 5 rhyolites including 1 obsidian, 1 sandstone, 5 schists, 3 serpentinites, 12 shales, 2 slates, 6 syenites, 2 tonalites, 3 trachytes, and 2 “ultrabasic rocks.” Data for some trace elements are missing for some of the GRS’s.

3) Terrestrial – Tektite. I have plotted data of Koeberl (1986) for various types of tektites. Note that tektites have compositions like terrestrial rocks (because they are terrestrial rocks!), not like meteorites.

4) All the white points represent terrestrial rocks. All the black points and colored points are for meteorites.

5) All the meteorites plotted in the plots (all square symbols) are stony meteorites, not stony-irons or irons.

6) Most (~95%) stony meteorites are chondrites, and most chondrites are ordinary chondrites. If you have actually found a meteorite, it is probably some kind of chondrite. That’s why I made the points for chondrites black and the ordinary chondrites BIG and black. Chondrites are most dissimilar to Earth rocks. Each black point represents the average composition of one of the chondrite groups: H, L, LL, EH, EL, CI, CM, CV, CO, CR, CO, R, Ac, & K. Data from Wasson & Kallemeyn (1988).

7) The lunar meteorite data are from my own database. Each point represents a different meteorite.

8) For the martian meteorites, eucrites, howardites, diogenites, and “other rare achondrites,” each point represents a meteorite or analysis. Data mostly from Jarosewich (1990), Lodders & Fegley (1998), and Mittlefehldt et al. (1998).

9) The plots presented here reasonably represent >99% of all meteorites.

10) It’s like lottery numbers – you don’t win unless the composition is consistent with ALL the chemical-composition parameters shown here, not just some of them!

11) I should show some plots here for chalcophile (sulfur-loving) elements – Cu, Zn, As, In, Sn, and Sb. The problem is that the concentrations of these elements are so low in achondrites (= meteorites that are not chondrites) that there are few data to plot. Chondrites have higher concentrations of chalcophile elements than achondrites:

Chalcophile-element concentrations in chondrites (range of group means of Wasson & Kallemeyn, 1988) values in ppm
Cu Zn As In Sn Sb
80–185 17–312 1–4 2–80 0.7–1.8 0.06–0.2

So, for example, if you have a rock with >5 ppm As (arsenic), then the rock is not a meteorite. Many terrestrial sedimentary rocks, as well as metamorphic rocks that formed from sedimentary rocks, have concentrations of chalcophile elements much higher than those in the table above.


Govindaraju K. (1994) 1994 compilation of working values and sample description for 383 geostandardsGeostandards Newsletter 18, 1–158.

Jarosewich E. (1990) Chemical analysis of meteorites: A compilation of stony and iron meteorite analysesMeteoritics 25, 323-327.

Koeberl C. (2006) Geochemistry of tektites and impact glasses. Annual Review of Earth and Planetary Sciences 1986 14, 323-350.

Korotev R. L. (1996) A self-consistent compilation of elemental concentration data for 93 geochemical reference samplesGeostandards Newsletter 20, 217–245.

Lodders K. and Fegley B. Jr. (1998) The Planetary Scientist’s Companion, Oxford University Press, New York, 371 pp.

Mittlefehldt D. W., McCoy T. J., Goodrich C. A., and Kracher A. (1998) Chapter 4. Non-chondritic meteorites from asteroidal bodies. In Reviews in Mineralogy, Vol. 36, Planetary Materials (ed. J. J. Papike), pp. 4-1–4-195, Mineralogical Society of America, Washington.

Wasson J. T. and Kallemeyn G. W. (1988) Compositions of chondritesPhilosophical Transactions of the Royal Society of London, Series A 325, 535-544.