UR logo

Comparison of Inclusions in the Springwater and Imilac Pallasite Meteorites Using SEM

Robert McKinley

University of Rochester, Department of Earth and Environmental Sciences


Pallasites are unique among meteorites. They are discrete mixtures of mineral Fe-Ni alloy and olivine ([Mg, Fe]2SiO4). There are two classes of Main Group Pallasites: those with rounded olivine crystals, and those with angular olivines. The Springwater and Imilac Pallasites are two good examples of each of these types.

There are two generally accepted theories of pallasite formation. One is that, since olivine and Fe-Ni are the main constituents of terrestrial planetary mantles and cores respectively, Pallasites come from mixing at the core-mantle boundary of an asteroid. Another theory suggests that the metal and olivine were melted and mixed due to the heat and displacement from an impact between two asteroids rather than by simply mixing with each other in situ.

Scanning electron microscopy is a good way to get an idea of the physical and chemical properties of the boundaries between the olivine and the metal in these two meteorites. The amount of diffusion between the two can give clues about the cooling history of the meteorite and, by extention, through which mechanism it formed.

Samples and Preparation:

Imilac sampleSpringwater sample

Above are images of the smaples used, taken before any modification. On the left is the Imilac Pallasite. Imilac was found in the Atacama Desert in Chile in 1822. Note the lighter, more angular olivines. The sample on the right is the Springwater Pallasite. It was found in Saskatchewan, Canada in 1931. Its olivines are noticably darker and more rounded. Samples from Imilac were taken from the areas labelled K and M, while Springwtaer samples were taken from areas A and H. Samples were provided by the Paleomagnetic Research Group in the Department of Earth and Environmental Sciences at the University of Rochester.

Appropriately small samples were harvested from the samples above by extracting olivine crystals with a brass rod and rock hammer. These were then usually ground into slightly smaller bits (usually a couple of millimeters on their longest axis) with a ceramic mortar and pestle. Effort was taken to sort through all of the resultant fragments under a light microscope for the "dirtiest" olivine crystals that had been in direct contact with metal. After samples were chosen, they were photographed under a light microscope. They were then mounted in plasitc on a SEM sample stub and polished down to a grit of 0.2 µm. Samples that protruded significantly from the plastic were left without polishing. The sample stubs were then coated with 80 angstroms of platinum (Pt) to establish a conductive surface for SEM imaging. Finally, the coated sample was grounded to the stub with a graphite paint.

sputter coater
Above: The sputter coater in the SEM/TEM Prep Lab at the University of Rochester.

Above: The Zeiss Auriga Crossbeam SEM System at the University of Rochester used for imaging.

Imaging and Results:


By far the most important aspect of this project was obtaining maps of the elements present at the boundaries between olivine and metal in each sample. Backscatter electron images do a good job of showing relative atomic weights of materials using a greyscale, but Energy Dispersive X-Ray Spectroscopy (EDAX) is an ideal tool for obtaining specific elemental data.

Imilac samples proved to have a couple of interesting features. First of all, the absence of a significant nickel component in any metal or metal residue was very surprising. Even more surprising was the presence of a large amount of sulfur in conjuction with the iron metal, suggesting a combination of iron sulfide and iron oxide instead of the Fe-Ni alloy described in Pallasite literature.

IM-4M 83X
Above is the image and respective EDAX maps for a low magnification micrograph of two olivine crystals abutting a metal phase in between. Note the large signal on the sulfur map, inconsistent with expectations.


The Imilac sample seen here has veins of darker material noticable in this light microscope image. This part of the crystal was very close to a metal boundary, and SEM imaging seen below revealed what these veins were.

IM-2K 478XIM-2K 478X Maps


Above: The image and respective elemental maps for another Imilac olivine crystal, this time looking at a vein of iron oxide permeating the olivine matrix. In this crystal, no sulfur phase appears at all. Also of note is the iron residue stuck on the outisde of the olivine. Below the EDAX maps is a backscattered electron image clearly showing compositional differences between the vein and the surrounding crystal. See the EDAX spectrum below:

IM-2K Spectrum
IM-2K 1720XIM-2K 1720X Maps

IM-2K vein BSD

Above is an image (and elemental maps) of the same iron oxide vein as above, pictured elsewhere in the crystal and on a polished surface. Diagonal streaks in the image are remnants of polishing. Below this is a backscattered image of the same vein elsewhare on the crystal. The compositional difference is clear.


The Springwater samples provided quite an interesting contrast to the Imilac samples. While Imilac held a considerable amount of sulfur in its metallic areas, Springwater had more diffuse boundaries between phases, with other metals present. Also interesting was the varying composition of the olivine present in each meteorite. While Imilac samples were almost entirely magnesium rich olivine (forsterite), Springwater samples were more towards the middle of the olivine solid solution series, with some iron rich olivine (fayalite) being present. Also of note was Springwater's lack of veins in the olivine as seen in Imilac samples. The iron (and other metallic) reside left on unpolished surfaces was similar in composition and morphology between the two meteorites, but the veins in Imilac were largely unseen in Springwater samples.


Above is a light microscope image and a micrograph of the same Springwtaer sample. Note the residue seen on the micrograph has a platy texture: this area of the crystal was directly abutting a metal area before being removed.

SP-2A protrusionSP-2A maps

This image was taken of a different sample of the same parent crystal as the sample just shown. The metal phase somehow fractured and clung to the olivine. Small patches of magnesium rich olivine can be seen at the base of the protrusion. Platy textured iton oxide residue coats the olivine below the metal protrusion. The metal itself appears to be un-oxidized, with large components of iron, nickel, and cobalt. The presence of cobalt in the Springwater Pallasite has also been seen by Wasson and Choi (2003), although this concentration seems higher than their estimates.

SP-H imgSP-H Maps

Above is a low magnification micrograph and associated compositional maps of a second Springwater sample. This sample was unpolished and retained some surface defects. The dark spots seen in the olivine are the result of surface imperfections being filled with an epoxy to give the sample a smooth look for display. The grooves in the metal are remnants from when the transect was cut with a rock saw. Nevertheless, a similar patter is seen with the metal phase. Iron, nickel, and cobalt are all seen. The olivine is largely magnesium rich, but some bleeding of iron signatures can be seen in the top left of the iron map. A phosphate mineral, something only seen in this sample, composes the bottom of the metal phase. Phosphates have been seen in the Springwtaer pallasite before, and are documented in Buseck (1977) as well as Davis and Olsen (1989).


Since the data gathered only represents limited samples from two pallasites, it is difficult to make broad conclusions regarding pallasite formation. However, the implications of the data at hand are quite clear. The discrete, angular boundaries between metal and olivine seen in Imilac samples as well as the cracks and veins seen, seem to indicate that there was an impact at some point in this meteorite's history. Springwater on the other hand has less discrete, more rounded boundaries between olivine and metal, contains accessory phosphate minerals, and more varied metal alloy composition, indicating a longer, less interrupted thermal history.


I would like to thank Brian McIntyre for all of his lectures and good humor. I would also like to thank Margaret Samuels for her help throughout course lab exercises. Finally, I would like to thank the Paleomagnetic Reserach Group for providing the necessary meteorite samples.

References and Further Reading:

Buseck and Holdsworth. "Phosphate Minerals in Pallasite Meteorites." Mineralogical Magazine. (1977): 41, 91-102.

Davis and Olsen. "The Origin of Phosphate Minerals in Eagle Station and Springwater Pallasites." Lunar and Planetary Science Conference XX, 1989.

A. Desrousseaux Et al. "An Analytical Electron Microscope Investigation of Some Pallasites." Physics of the Earth and Planetary Interiors (1997): 103, 101-115.

Wasson and Choi. "Main-group pallasites: Chemical composition, relationship to IIIAB irons, and origin." Geochimica et Cosmochimica Acta, (2003): 67, 16, 3079 –3096.

Please enter any comments, criticisms, questions, etc. below.

Your name:

Email address: