The NASA sample recovery mission called STARDUST is the first ever to sample
material from a comet. At 5 am (New York time) January 15, 2006, the STARDUST landing capsule
parachuted down successfully onto the desert of Southern Utah, containing particles from comet Wild 2,
and also samples of interstellar dust collected in flight on the back side of the collector array.
The capsule containing the samples did not get wet, and is now being examined by the Preliminary Examination
Team at Johnson Space Center and other laboratories. Stay tuned....
The samples of 'star dust' are captured in aerogel, an ultra-low density silica foam. AMNH, Columbia, and CUNY scientists D.S. Ebel, M.K. Weisberg, and H.C. Connolly Jr. are part of the STARDUST Preliminary Examination Teams that will initially characterize the samples. Most of this preliminary characterization will be done on individual cometary 'dust' particles after they have been separated from their host aerogel tiles. However, this first separation step is not trivial. The tomography research described here is a method for non-destructively characterizing the whole tiles, prior to separation of dust particles, which is inherently invasive and sacrifices information contained in, for example, particle track morphologies.
Synchrotron x-ray micro-tomography is a technique that
provides non-destructive 3-dimensional information about rare samples on scales from 1 to 17 microns/pixel.
This technique is particularly well-suited to rare, easily-damaged samples, particularly
those returned by NASA sample recovery missions such as
STARDUST.
Results from the onboard Comet and Interstellar Dust Analyzer (CIDA, 50 cm2 capture area)
and Dust Flux Monitor Instrument (DFMI, ~104 cm2) suggest 2800 (+/-500) particles of
diameter greater than or equal to 15 micrometers impacted the aerogel collectors (Tuzzolino et al. 2004).
Particles encountered aerogel at high relative velocity (~6.1 km/sec flyby speed), and interacted
with aerogel as they slowed down (Tsou et al. 1993, 2003). Intense swarms of particles arrived in bursts,
consistent with particle fragmentation (Tuzzolino et al. 2004). Fragmentation, sintering and other
events will have been recorded in the textures, fragment locations, and track geometries
observable by tomography in the aerogel.
On this page is a demonstration of our recently tested ability to image track geometries,
see aerogel sintering and melting, and locate fragments of what was once a
single particle. These results were obtained on samples that are our best available
analogs of what is expected to return in January 2006. Aerogel tiles were 'shot' with
dust from the Allende (CV3) chondrite meteorite by F. Horz, using the gas gun at NASA.
We performed tomographic analysis using the synchrotron at the Advanced Photon Source (APS)
at Argonne National Laboratory (DOE/U. Chicago). The resolutions we worked at for this
preliminary work were 3.7 and 14.4 microns/pixel. We have learned how to do 'local'
or 'lambda' tomography (so named for the mathematical transform involved) on aerogel tiles,
suitable for their analysis at resolutions of less than 1.5 micron/pixel. However, such a high
resolution is not likely to be necessary. Even at 3.7 microns/pixel, the aerogel appears almost
the same as the air surrounding it and in tracks, making the latter difficult to distinguish. At
8 to 14 microns/pixel, we should get optimal results: see tracks and fragmentation, but also capture 'dense'
spots occupied by comet dust.
Actually, 'pixel' here refers to the edge lengths of cubic volume elements, or 'voxels'. Each voxel in a
volume (x-y-z) has an associated x-ray attenuation, which is a function of the density of
material in that voxel. The tomography experiment results in a 3-dimensional (3D) data set
describing the density in every voxel of the volume around a rotation center in the sample.
The 'lambda' tomography method allows the sample to be very much larger than the field of view
(fov) of the x-ray beam and associated CCD image plate.
The images below, and downloadable stacks and movies, are 8-bit tiff frames constructed from the volume data, cropped to show a single track and particles. The native data are in 16-bit integers, so conversion to 8-bit images requires degrading the resolution from approximately 64,000 grayscales to 256. For this reason, images are thresholded to enhance contrast in the optimal range.
Noise is present in these images. The tomography technique requires reconstruction of the
density structure from multiple still images
(see tutorial).
This results in rings about the rotation center (parallel to z), which manifest as
'bullets' near the center in x and y slices. They are visible as fuzziness around the
centers of image stacks.
Imperfections or contaminants (metal? silicon?) in the aerogel appear to be present as
very bright points with no aerogel tracks near them. The source of these is under investigation.
At the top surface of the aerogel ('up' in all images), there is a scattering of bright
particles, which must be debris from the gas gun used to shoot Allende meteorite dust
into the aerogel tile.
Fig. 1: Bulk aerogel tile mounted in a soft 'vise' prior to tomographic work.
Aerogel tile, shot with Allende meteorite dust, imaged at 14.5 micron/pixel.
These are sequential frames from a single volume 650 x 650 x 515 pixels in size.
The volume is located to the left end of the aerogel block shown in Fig. 1.
The volume was cropped (x:200-329, y:180-219) to focus on a single
track. The grayscale range of the data is truncated to -2500 to +4000,
to enhance contrast between aerogel and air (in tracks).
Particles show up as bright objects (highest density).
The first two image sets below show a small range of consecutive x and y frames.
The following figures show more detail in fewer frames.
Whole sets of x, y, and z 'stacks', and mpg movies containing those stacks,
can be found in the following location:
click here for image stacks in tif format (8-bit).
Image stacks may be most conveniently viewed using the ImageJ (for PC)
or NIH-Image (for Mac) freeware:
click here for NIH-IMAGE.
Fig. 2: Close-up of Y slices 19 to 24. Particle has fragmented into at
least three pieces.
One piece (in oval) is a long sliver, also visible in orthogonal (x) slices.
Arrows point to fragments that caused the track to branch. All these fragments
are clearly part of the same original particle. The x-slices, in the next
figure, show how these two lower particles (arrows) separated by substantial distances
after fragmentation. Each frame in both x and y is 14.5 micron from the previous frame.
click here for full image stacks.
Fig. 3: Close-up of X slices 89 to 108, orthogonal (perpendicular) to y-slices.
The piece in the oval is verified as a long sliver, as seen in the orthogonal (y) slices.
Short and long arrows point to the same fragments as in Fig.2 (y-slices).
These fragments are very separated in the x view, indicating severe horizontal motion.
Perhaps the particle was heated by aerogel friction, and internal explosive forces caused
it to fragment in this way.
Note also that there is some small part of the original particle, that continued to make
a faint track deeper in the aerogel tile than shown in these frames. The track of this
particle, and the particle itself, could easily be recovered by performing tomographic
imaging at a lower level in the aerogel tile.
Fig. 4: Close-up of selected Z slices 90 to 406, orthogonal to y- and x-slices.
Inspection of the stack of z images (stacks) reveals that
the cracks in aerogel, the sharp curved spikes coming off of the circular tracks, are
rifled: They cork-screw into the aerogel with depth.
Fig. 5: Context for Fig. 2, Y slices 16 to 27.
Fig. 6: Context for Fig. 3, X slices 70 to 110.
Fig. 7: Geometry for conventional (A) and local (or 'lambda') tomography (B).
We cannot be held responsible for any circumstances resulting from its use, inavailability, or possible inaccuracy. We reserve the right to suspend or discontinue this service or portions thereof at any time.
We also reserve the right to deny access to any individual or organization that we determine is abusing this service. Be aware that if we are forced to terminate your access, all access from your location may be terminated as a side effect, therefore preventing others at your location from using our service. Examples of abuse include automated transfers resulting in excessive data requests (because it hinders others from accessing the service) and attempting to gain access to documents and host machines not intended for public use.