Meteorite testing

back to “Some Meteorite Realities”

I do not test or classify meteorites

I do not test rocks to determine if they are meteorites. I do not classify meteorites and I do not provide “Certificates of Authenticity.” I am a retired geochemist and I no longer have a laboratory. Do not send me samples.

If, on the basis of the information that you provide me, I think that your rock might be a meteorite, then I can probably put you in contact with someone who does classify meteorites. I will not do that, however, unless I am >95% certain that the rock is, in fact, a meteorite. It will likely cost you several hundred dollars to have it classified because classification requires a lot of time on expensive laboratory instruments. For amateur finders it is more convenient to find a meteorite dealer who will buy it unclassified as the dealer will have contacts who can classify it. Be aware, however, that most meteorite dealers will ignore you because, like me, they are contacted every day by sincere persons with meteorwrongs. 

Alternatively, click here. This is the list from the Meteoritical Bulletin Database of all the meteorite names that have been approved in the past 6 months. If you click on the name of a meteorite, you can find the names and institutions of the persons who classified the meteorite. You can try to contact classifiers directly. You almost certainly will be ignored, however, because you do not make a convincing case that the rock is, in fact, a meteorite. Follow my advice below. All classifiers are besieged with requests from well-meaning but overly optimistic people who have found meteorwrongs. Most classifiers will only accept samples from experienced finders and collectors who have a record of recognizing a meteorite when they see one.

Stony meteorites

If you are particularly certain that your rock is a meteorite and you really want to convince me or any other scientist, then I urge you to obtain a chemical analysis at a commercial rock-testing laboratory. There are many labs around the world that can provide such tests. At a minimum, I need “whole-rock” data for the major rock-forming elements: Na2O, MgO, Al2O3, SiO2, K2O, CaO, TiO2, Cr2O3 or Cr, MnO, and Fe2O3 as well as trace elements Ni and Co.

See also: Geochemical Analysis Reports

I recommend Actlabs, which has branches on several continents. Ask for analysis code Meteorite (ICP/ICPMS). I have no financial interest in Actlabs, I just know that they do a good job. First, however, send me some photos of the rock so that I have the opportunity to say, “If that rock were mine, I would not spend the money to have it analyzed because it does not look like a meteorite.”

Actlabs requests a 5-gram sample (a US nickel coin weighs 5 grams). They can do the analysis on as little as 1 gram, however, if you request “no LOI” – (loss on ignition, i.e., percent weight loss when the sample is heated to a high temperature. LOI is sometimes useful but not critical for determining whether or not a rock is a meteorite. Actlabs and many other commercial labs use a technique called ICP-MS – inductively coupled plasma mass spectrometry. ICP-MS requires dissolving a sample in acid.

Send me a copy of the report that Actlabs sends you (the XLS file, not the PDF file) and I will tell you whether the rock composition is consistent with that of a meteorite. A chemical analysis is sufficient for me to say “yes, it is” or “no, it is not” 99 times out of 100. If I conclude that the composition of your rock is not consistent with any kind of meteorite, then I probably will not be able to tell you just what kind of rock it really is. Rock-type identification requires other kinds of tests. If it is a meteorite, then a meteorite petrologist is required to classify it and obtain an official name. For example, I can say with 99+% certainty that your rock is an ordinary chondrite from the chemical composition but I cannot reliably tell you which type of ordinary chondrite it is (H, L, or LL). It will be easier for you to catch the attention of an overworked meteorite petrologist or meteorite dealer if you have the compositional data. 

Please

Do not bother Actlabs (or any other lab) with questions about meteorite identification. They are chemists who are experts in rock analysis. It is not their job to interpret results. They have no one on staff who is a meteorite expert. They do not classify meteorites. They do not offer “Certificates of Authenticity.” I am the meteorite composition expert, which I do for free. Actlabs sends persons with meteorite questions to me.
A portion of an Actlabs report sent to me be a rock prospector. The first elements I look at are SiO2, Fe2O3(T), Na2O, K2O, Cr, and Ni. This rock is not a meteorite. SiO2 concentrations in stony meteorites are typically 35-57%. Fe2O3(T) [total iron expressed as Fe2O3] is typically 15-40% except that lunar meteorites can be as low as 3%. Except for some martian meteorites, Na2O is almost always <1% and K2O is always <1.5% and typically 0.1% in ordinary chondrites. A characteristic difference between terrestrial rocks and meteorites is that meteorites have high Cr concentrations, typically 1000-8000 ppm, except for a few lunar and martian meteorites, whereas terrestrial rocks typically have <500 ppm. Chondritic meteorites have 10,000-18,000 ppm Ni. This rock is probably a granite or closely related rock like dacite or rhyolite.

Check your own data with Chemical composition of meteorites.

July, 2022: I have received results of analyses of 672 samples from Actlabs and more than 140 samples from other labs. Only 9 of the rocks have been meteorites, 6 ordinary chondrites, 2 iron meteorites, and 1 pallasite. More than half of these rocks were from northern Africa or the Middle East and a couple, I believe, were stones that someone had bought or inherited.

Wavelength dispersive X-ray fluorescence spectrometry

X-ray fluorescence (XRF) spectrometry (also known as XRF spectroscopy) has been used for decades as a means of determining the elemental composition of rocks. Historically it has been a laboratory technique requiring a large instrument that generates an intense beam of X-rays from an X-ray tube. Samples are typically ~1 gram (0.25-5 g) of pressed rock powder or glass made from melting of rock powder. The entire sample is exposed to the X-ray beam. The primary X-rays excite secondary X-rays in the rock sample. Each element diffracts secondary X-rays of characteristic wavelengths in different directions. The detector is moved in a circular path about the sample counting x-rays emitted by each element individually. This technique is known as wavelength dispersive XRF (WD XRF) spectrometry. When used by experts it is very accurate and precise, on the order of 1-2% for elements occurring at concentrations of 1% or greater as well as several trace elements of lower concentration. Other advantages are that a large fraction of the elements in the periodic table can be determined and, because the sample is powdered, the will be more representative of the whole rock than a technique using spot or beam analysis (below). Wavelength dispersive XRF is mainly done in commercial and university labs as well as by mining companies.

Energy dispersive X-ray fluorescence

EDX or, more properly, ED XRF – energy dispersive XRF is a cheaper alternative. Such data are commonly obtained with a scanning electron microscope (SEM). In these instruments a small (usually much less than a millimeter) electron beam is aimed at the sample and all the emitted X-rays are collected with a detector that sorts the X-rays in order of increasing energy to yield an X-ray spectrum. Alternatively, there are instruments that use either small X-ray tubes or radioactive sources that emit gamma-rays as the excitation source. Again, however, the detector “sees” all the emitted X-rays at once and sorts them by energy producing an energy spectrum.

For the purpose of identifying meteorites there are three main problems with ED XRF.

  • The energy resolution is considerably worse for ED XRF than for WD XRF. For example, because iron (Fe) and nickel (Ni) are adjacent to each other in the periodic table, the spectral peaks overlap and are, consequently, more difficult to quantify. Identification of peaks is sometimes ambiguous. In the wavelength spectra obtained by WF XRF, the peaks are much narrower and there are many fewer peak overlaps.
  • The excitation sources used in ED XRD are much weaker than that from the X-ray tubes used in WD XRF, thus longer count times are required to accumulate the same number of counts. Precision improves as the number of observed counts increases. Elements occurring at abundances of <5% are sometimes not determined with enough statistical precision to be useful.
  • In ED XRF, unless the sample has been ground to a fine powder, a spot analysis is obtained, not a bulk (whole-rock) analysis as in WD XRF. We need to know the bulk composition, not the composition of several small spots that may only represent individual mineral grains. If the beam is at least 2 mm in size or the beam is rastered, an ordinary chondrite could likely be distinguished from a terrestrial rock with ED XRF data. It would not be possible, however, to definitively recognize most achondrites as meteorites.

Bottom line: Data obtained by ED XRF are often not sufficiently accurate, precise, or complete to distinguish a meteorite from a terrestrial rock.

Handheld XRF analyzers

Handheld XRF analyzers are useful for recognizing iron meteorites. This photo was sent to me by someone who had what he believed to be a sample of Campo del Cielo, one of the most commercially available iron meteorites. The sample was tested in a coin shop. Note at the top that the instrument is set for “precious metals.” The “official” Ni, Co, and Fe concentrations of CdC are 6.7%, 0.46%, 92.9%. The meteorite might be a Campo, but then it might not. It is definitely an iron meteorite, however, which typically contain 5-25% Ni, much greater than nearly any man-made iron-rich object. (A U.S nickel coin [5-cent piece] is 25% Ni and 75% Cu.) Handheld XRF analyzers tend to overestimate concentrations rare elements like gold (Au). The concentration of gold reported above is 0.083% (830 ppm) yet I am unaware of any iron meteorite with greater than 0.0005% (5 ppm) Au. Similarly, concentrations of copper (Cu) in IAB irons such as Campo del Cielo are typically only 0.01-0.04%, not 0.3%.

Several persons, before contacting me, have brought their rocks to a scrap-yard dealer, mining company, or jeweler to have them tested with a hand-held “XRF gun” or other handheld XRF analyzer. Because of the issues discussed above, most of the results that I have been sent from such analyzers have not been useful for determining if a rock is a meteorite. Also, there are at least three additional problems:

  • Most of the instruments are designed or programed to do analysis of ores and metal for metallic elements – “mining mode” for miners [e,g., Cu (copper), Mo (molybdenum), and W (tungsten)] or “precious metals” [Rh (rhodium) and (Pt) platinum] for jewelers. These modes focus on elements that are useless for stony meteorites because the concentrations of all these elements in meteorites are much lower than the analyzer can actually detect (parts-per-million or parts-per-billion) in rocks. To determine if a rock is a meteorite, we need to know the concentrations of the major rock-forming elements: Si, Al, Fe, Mg, Ca, Ti, Mn, and (ideally) Na, K, and Cr as well as the minor or trace element Ni. In metals mode, for example, there are typically no data for Si (silicon) or Ca (calcium), critical elements for determining if a rock is a stony meteorite.
  • Second, data for Na and Mg are also critical to distinguish earth rocks from meteorites and those elements cannot be determined by X-ray fluorescence in air.
  • Third, it seems that some users do not really know what they are doing. For, example, a fellow contacted me once saying “Big time reputable gold dealer tested it with his X ray gun.” The gold dealer told him that his rock contained “at least 15% Bohrium,” an element that does not occur as a stable element in nature and the only isotope of which has a half-life of ~85 milliseconds. (I suspect that the instrument actually detected boron, not bohrium.) As noted above, others report the presence of platinum-group elements, which cannot be detected in rocks by XRF because the concentrations are much too low – part-per-billion levels in terrestrial rocks and part-per-million levels in chondrites.

This is what happens when a silicate rock is examined in “precious metals” mode. The rock is composed mainly (probably >90%) of O, Si, Al, Fe, Mg, and Ca. Except for Fe (iron), none of these elements are “seen” because the instrument is not set to look for them. Elements identified as Pt (platinum), Rh (rhodium), Cd (cadmium), Ni (nickel), and Pb (lead) are based on spectral peaks of interferences from the major rock-forming elements. All of these elements actually occur in the rock but at parts-per million or parts-per-billion levels. These rare elements cannot actually be detected, however, in the presence of the major rock-forming elements. “Precious metals” mode should only be used on metal samples.

X-ray diffraction (XRD)

People also send me results obtained by X-ray diffraction. XRD identifies the major minerals in a rock, not the chemical composition. In my experience, XRD results are often ambiguous. Some minerals that occur in meteorites largely do not occur at all in terrestrial rocks. So, the identification of, e.g., kamacite and troilite would be strong evidence that a rock is a meteorite. These two minerals are common in iron meteorites. Most of the minerals that are unique to meteorites are nevertheless minor to rare in stony meteorites and may not be detected by XRD. The three most abundant minerals in stony meteorites are olivine, pyroxene, and plagioclase. These three minerals are also among the most common minerals in terrestrial igneous rocks. Some minerals that are common in terrestrial rocks are rare to absent in freshly fallen meteorites. Quartz, calcite, micas, and clay minerals are good examples.

Bottom line: XRD can often prove that a rock is not a meteorite but it rarely provides unambiguous evidence that a rock is a meteorite.

Iron meteorites

If you have found a piece of metal that you think might be an iron meteorite, you need to have it analyzed (at a minimum) for iron (Fe), nickel (Ni), chromium (Cr), and manganese (Mn). A metallurgical lab could do this. Unfortunately, I do not know a lab that does it cheaply. If somebody out there does, please let me know. In proper hands, a handheld XRF analyzer would be useful for iron meteorites.

back to “Some Meteorite Realities”