A Fluorescent Aerogel for Capture and Identification of Interplanetary and Interstellar Dust

Abstract

Aerogels are extremely low-density solids whose superiority as capturing media for hypervelocity (v > 0.5 km/s) grains has been well established. A prominent example is the use of silica aerogel as the collecting medium for cometary and interstellar grains on NASA s Stardust mission. Aerogel collectors have been deployed in low-earth orbit, but severe background from anthropogenic orbital debris has so far prevented the identification of more than a handful of interplanetary particles. No interstellar particles have been identified so far. Since they are on hyperbolic orbits, extraterrestrial particles are faster than orbital debris, so could in principle be identified on that basis, but existing aerogels give little information on impact velocity. With this in mind, we have developed a novel calorimetric aerogel which passively records the kinetic energy of captured hypervelocity particles.

Full text

arXiv:astro-ph/0303612v1 27 Mar 2003
A Fluorescent Aerogel for Capture and Identification of
Interplanetary and Interstellar Dust
Gerardo Dom´ınguez1and Andrew J. Westphal
Space Sciences Laboratory, University of California, Berkeley, CA 94720
Mark L.F. Phillips
Pleasanton Ridge Research Corporation, Hayward, CA 94542
and
Steven M. Jones
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA 91109
ABSTRACT
Contemporary interstellar dust has never been analyzed in the laboratory,
despite its obvious astronomical importance and its potential as a probe of stellar
nucleosynthesis and galactic chemical evolution. Here we report the discovery of
a novel fluorescent aerogel which is capable of capturing hypervelocity dust grains
and passively recording their kinetic energies. An array of these “calorimetric”
aerogel collectors in low earth orbit would lead to the capture and identification
of large numbers of interstellar dust grains.
Subject headings: astrochemistry — instrumentation: detectors — interplane-
tary medium — dust, extinction — meteors, meteoroids — techniques: image
processing
1. Introduction
Interstellar dust is an important component of the interstellar medium. Dust dominates
the opacity from the far ultraviolet through the far infrared and hence controls the spectral
appearance of most interstellar objects. Because of dust shielding against dissociating FUV
1Department of Physics, University of California, Berkeley

– 2 –
radiation, molecules can form in dense clouds which allows cooling to low temperatures and
thus, eventually, gravity to overwhelm pressure support and the formation of new stars.
Small dust grains also dominate the heating of the interstellar gas through the photoelectric
effect and hence controls the structure of the interstellar medium. Despite some 50+ years
of active research, the composition of interstellar dust is still largely guessed at. In essence,
our ignorance reflects the difficulty to infer dust composition from remote astronomical
observations. Here we propose a novel collection agent which allows the discriminatory
collection of interstellar grains and separation from solar system debris. This promises to
open up a new window on the solid component of the interstellar medium.
Although it is known that IS dust penetrates into the inner solar system (Gr¨un et al.
2000; Taylor et al. 1996), to date not even a single contemporary IS grain has been captured
and analyzed in the laboratory. Using sophisticated chemical separation techniques, certain
types of refractory ancient IS particles (so-called “pre-solar grains”) have been isolated from
chondritic meteorites (e.g. (Amari & Zinner 1997)). Isotopic abundance patterns within
these individual grains often differ wildly from solar-system values, and point to the formation
of these grains in specific astrophysical environments such as supernova ejecta and the winds
of Asymptotic Giant Branch (AGB) stars. But because only the most chemically robust
particles (e.g., graphite, SiC, Al2O3) survive the harsh chemical separation, this sample is
extremely biased, and it is unlikely that these particles are typical of those found in the
interstellar medium. A sample of IS dust collected by spacecraft in the inner solar system
would be less biased, and could lead to the first laboratory characterization of the “typical” IS
dust particle. Furthermore, such a sample would allow us to detect isotopic, elemental, and
mineralogical differences between dust in the protosolar cloud and dust currently residing in
the local ISM, and to probe galactic chemical evolution over the 4.6 Gy since the formation
of the solar system. Pre-solar grains are already proving to be a valuable probe of galactic
chemical evolution and stellar nucleosynthesis (Amari et al. 2001a; Nittler et al. 1997).
The vast majority (∼84%) of large (>1µm) ancient pre-solar grains appear to be
of one type, so-called “mainstream” SiC grains. These grains are enriched in 22Ne, show
s-process signatures in Kr and Xe, and probably originate in the outflows of AGB stars.
Grains from other astrophysical sites have been identified but are relatively rare (e.g., Type-
A,B SiC, tentatively identified with J-type carbon stars (Amari et al. 2001b), 3-4%; Type-X
SiC from supernovae (Amari & Zinner 1997), 1%; and alumina from high-metallicity red
giants, <0.5%.) A few pre-solar grains show isotopic patterns that are unique among the
thousands that have been studied so far (Nittler et al. 1997). If dust in the local ISM shows
a similar pattern of diversity, with a dominant common type and relatively rare populations
of exotic grains, a large-statistics collection technique will be required to capture, identify
and study contemporary IS dust grains from a wide variety of astrophysical sources.

– 3 –
Aerogels are extremely low-density solids whose superiority as capturing media for hy-
pervelocity (v > 0.5km/s) grains has been well established (Barrett et al. 1992; H¨orz et al.
2000; Kitazawa et al. 1999). A prominent example is the use of silica aerogel as the collecting
medium for cometary and interstellar grains on NASA’s Stardust mission (Brownlee et al.
1997). Aerogel collectors have been deployed in low-earth orbit, but severe background from
anthropogenic orbital debris has so far prevented the identification of more than a handful
of interplanetary particles (H¨orz et al. 2000). No interstellar particles have been identified
so far. Since they are on hyperbolic orbits, extraterrestrial particles are faster than orbital
debris, so could in principle be identified on that basis, but existing aerogels give little in-
formation on impact velocity. With this in mind, we have developed a novel calorimetric
aerogel which passively records the kinetic energy of captured hypervelocity particles.
The capture of a hypervelocity dust particle in aerogel produces a shock wave that
deforms, heats, and vaporizes the aerogel material in the vicinity of the projectile’s trajectory,
resulting in the formation of a permanent track. The correlation between captured projectile
velocity and track characteristics (e.g., track length, track volume, etc.) is poor (Kitazawa
et al. 1999). This behavior is expected theoretically (Anderson & Ahrens 1994; Westphal
et al. 1998)(G. Dom´ınguez in preparation). The amount of local heating, however, is nearly
linearly proportional to the projectile kinetic energy (Anderson & Ahrens 1994). If this local
heating alters some property of the aerogel in the vicinity of the track, then this property
could be used as a calorimeter. We chose to focus on inducing a fluorescence signal.
2. Observation of Fluorescence from Capture Events
We have observed fluorescence resulting from the thermal alteration of aerogels pre-
viously in various doped aerogel systems, which fluoresce weakly in their amorphous state
and strongly when baked at high temperatures (≃1000◦C) for extended periods of time
(∼1hr). A simple example of such a system is alumina aerogel doped with chromium (III).
The amorphous, unheated phase is only very weakly fluorescent under UV illumination (254
nm or 365 nm). Heating the aerogel to 1450◦C causes it to crystallize to the well-known
luminescent phase α-Al2O3:Cr, known in Nature as ruby, which glows red (λmax ≃700nm)
under UV illumination. More complex systems include alumina gels co-doped with Gd and
Tb. Gd acts as a sensitizer by absorbing UV light at certain wavelengths and nonradiatively
transferring energy to Tb, which emits at several wavelengths, principally in the green.
Local heating that results from the capture of hypervelocity projectiles is rapid and
confined to small regions in the aerogel. However, the inducement of a fluorescent state
as a result of rapid (t<200µs), local heating (within <100µm of the particle track) in

– 4 –
an aerogel has previously not been reported. To test whether local heating in an aerogel
could induce an irreversible phase transformation into a fluorescent phase, the effects of
hypervelocity projectile capture were first simulated by exposing samples of Cr-doped and
(Gd,Tb)-doped alumina aerogels (ρ∼170 mg/cc) with a pulsed CO2laser (300 Hz, 50µm
spot size, pulse width=50µs, power=0.25-0.50 W). The energy per pulse is approximately
the energetic equivalent of a glass sphere 10 microns in diameter impacting at 10 km/s.
Some of these aerogels displayed brilliant green fluorescence in the regions of local heating.
This was encouraging evidence that the capture of hypervelocity dust particles could induce
a fluorescent phase in alumina aerogels. These alumina aerogel samples were selected for
shots with hypervelocity projectiles (a mix of powdered meteorite and glass beads) at the
Advanced Vertical Gun Range at NASA Ames Research Center. Two of these samples
showed intense green fluorescence in the heated material surrounding the particle tracks,
thus establishing that the phase transformation occurs in alumina aerogels. Quantitative
measurements with these shots were precluded because of the large spread in particle sizes
and the unknown effect of particle ablation. These shots were followed more recently, again
at Ames, with projectiles consisting of a mixture of monodisperse glass spheres. This allowed
us to do quantitative measurements of the fluorescence yield as a function of particle size
and velocity.
3. Analysis of Fluorescence Observations
We measured the fluorescence yield using a standard fluorescence microscope with a
cooled color CCD video camera. The fluorescence was excited at 365 nm using a standard
bandpass filter cube at the excitation side and imaged using a long pass filter (λ≥395nm).
The samples were imaged within two hours of each other to minimize the effects of UV lamp
intensity variations. High resolution images of the aerogel surface where tracks entered were
taken and the background fluorescence (weak and mostly blue) was subtracted as follows.
A local blank region of aerogel was sampled, and the average ratio of green to blue, fgb was
determined; for each pixel we defined the net fluorescence in the green as:
Inet
green =Igreen −fgb Iblue (1)
where Iblue is the blue pixel value. We chose this background subtraction method because a
linear increase in both the green and blue channels would be expected, even in the absence
of a phase transformation, due to the increased density of aerogel in the vicinity of the track
mouth. We define the fluorescence yield as the sum of Inet
green for Inet
green >2.5σabove the
pixel noise in the region surrounding the track mouth. The yield increases dramatically
with increasing velocity within each particle population (Fig. 1). In Fig. 2, we show the

– 5 –
fluorescence yield as a function of kinetic energy. Over the range from 2µm to 20µm (three
orders of magnitude in mass), the fluorescence yield appears to be consistent with being a
single-valued function of the particle kinetic energy, Ig∝E0.69
k. We found that the exponent
is insensitive to the choice of fluorescence signal-to-noise threshold.
A reasonable model for the energetics of grain capture can be used to explain, at least
qualitatively, the calorimetric aspects of the aerogel. In this model, we treat the aerogel as
a fluid. In the limit of large Reynolds number, the energy deposited per unit path length by
a grain of radius r, density ρg, and kinetic energy Eis
dE
dx ∼3
2
1
r
ρa
ρg
E=E
λ,(2)
where ρais the aerogel density, and
λ=2
3rρg
ρa
.(3)
This stopping length scale agrees to within 10% of the value obtained following the more
detailed treatment by Anderson and Ahrens (Anderson & Ahrens 1994). The range of the
particle in its supersonic slowing phase Ris
Rsuper ∼2λln v
vsonic ,(4)
where vsonic is the speed of sound in the aerogel. The logarithmic dependence of the su-
personic range on velocity is consistent with the weak dependence observed experimentally
(Kitazawa et al. 1999). If some fraction of the energy loss contributes to the local heating of
the aerogel, we should expect the amount of aerogel crystallized to increase as the projectile
kinetic energy increases. Assuming that the luminescence we observe is dominated by one
fluorescent phase, the mass per unit track length that is converted into this fluorescent phase
is expected to be
dmfl
dx ∝v2r2.(5)
The dependence of Igon the amount of crystallized aerogel is not necessarily straight-
forward, as it depends on the optical properties (ultraviolet and visible) of the aerogel as well
as the track length. For events with large track lengths, such as those due to 20 µm diameter
grains, the fluorescence yield may be dominated by the fluorescence at or near the surface of
the aerogel. If so, then Igshould be proportional to the amount of aerogel crystallized near
the track entrance. For constant density, therefore
Ig∝m2
3v2
0.(6)

– 6 –
where v0is the initial impact velocity. Conversely, for shallow events, the fluorescence yield
is expected to be more sensitive to the total amount of aerogel crystallized, hence we would
expect Ig∝E0. Interestingly, a fit of Igversus m2
3v2
0appears to be a better fit to the
data than Igversus E. Regardless of which functional form (E0or E0m−1
3) turns out to be
more accurate when additional studies are done, the main implication of the results reported
here is the same. Together with an independent measure of the mass, for example using in
situ optical imaging or x-ray fluorescence, the velocity of the embedded projectile can be
determined (Flynn 1996, 2000).
The expectation that the amount of crystallized aerogel scale with the amount of heat
deposited is not obvious. The dependence of emission intensity on grain kinetic energy is
complicated by the fact that several phase transitions can occur as alumina aerogels are
heated (amorphous →γ-Al2O3→θ-Al2O3→α-Al2O3(corundum)) (Mizushima & Hori
1995). Fluorescence efficiencies of dopant ions typically depend strongly on their local crystal
field, and thus it is likely that even if higher grain kinetic energy does not crystallize a larger
mass of aerogel, the higher temperatures produced within the track will yield phases of
different luminescence. For example, the heating of Al2O3:Gd,Tb doped aerogels to even
moderate temperatures (∼1100◦C) can precipitate phases such as the perovskite phase
GdAlO3:Tb and the meta-stable garnet phase Gd3Al5O12:Tb, which yield brilliant green
luminescence under UV illumination.
The kinetics of crystallization in doped alumina aerogels are not known but a more
detailed understanding should be helpful in maximizing their usefulness. The density of
the aerogels used in this study are higher than optimal for capture of small hypervelocity
particles. Developing larger samples with lower density is a major focus of our current work.
This will allow us to characterize the optical properties of our calorimetric aerogel, which
will in turn help us understand the effects that oblique impacts and particle fragmentation
may have on integrated fluorescence. Studies of the crystallographic phase and fluorescence
dependence on temperature could be used as an in situ temperature gauge along tracks
and could improve our understanding of the kinetics of grain capture in aerogels. Detailed,
systematic studies of fluorescence along a track may provide information on the instantaneous
energy loss that a captured dust particle experiences (see Fig. 3).
4. Discussion
The Stardust spacecraft, whose primary mission is to return samples of cometary dust
to Earth for laboratory study, has exposed aerogel collectors to the interstellar dust stream
during two periods of its cruise phase. The Stardust collectors will be returned in 2006.

– 7 –
Models of the IS dust flux in the inner solar system indicate that the Stardust collectors
will capture ∼10 1-µm particles, and perhaps one 2-µm particle. An array of calorimetric
aerogel, with collecting area of 3 m2deployed in low earth orbit for two years, would have
enough collecting power to collect several hundred 1-µm IS particles (Landgraf et al. 2000).
A collector deployed on the wake side of a spacecraft in low-earth orbit could collect IS dust
at moderate velocites (<10 km/s) during periods of the year when the earth’s motion is
most parallel to that of the IS dust stream (Gr¨un et al. 2000). Furthermore, the largest
particle expected to be captured by such an array would be ∼30 times more massive than
the largest particle expected to be collected by Stardust (Landgraf et al. 2000) (see Figure
4). These particles would be large enough to apply multiple chemical, mineralogical, and
isotopic analysis techniques to each particle (Zolensky et al. 2000).
The main point of this paper is that we have discovered a method of distinguishing
between copious anthropogenic debris and relatively rare extraterrestrial particles captured
in a collector in low earth orbit. Although we have chosen to emphasize the importance of
capturing large numbers of contemporary interstellar grains for the first time, it is inevitable
that interplanetary dust particles would also be collected. Both of these populations would be
of scientific interest, and separating these two population is a complex problem and beyond
the scope of this paper. A large-statistics collection of interplanetary dust collected in space
would be a valuable resource for the meteoritics and planetary science community. So-called
Interplanetary Dust Particles (IDPs) have been collected in the stratosphere for many years
(Brownlee et al. 2002). Micrometeorites have also been collected terrestrially in Antarctica
— these are the so-called Antarctic Micrometeorites (AMMs) (Engrand & Maurette 1998).
Each of these collection techniques has its own biases. Both are biased towards particles
that can survive atmospheric entry. The effects of atmospheric contamination are poorly
understood (Flynn, Bajt, Sutton, & Klock 1995). Terrestrially-collected micrometeorites
are selected towards particles that survive weathering and that are readily recognized as
extraterrestrial.
A few chondritic particles have been extracted from ordinary silica aerogel collectors
flown in space and analyzed (H¨orz et al. 2000). These chondritic grains were selected from
a large background of anthropogenic particles. The relationship between IDPs, AMMs,
micrometeorites and ordinary chondrites is not clear (Brownlee et al. 2002; Flynn 2002), but
it appears that AMMs constitute a different population than IDPs, and may have a different
origin. A single collector that is large enough to capture, in space, several 100 µm particles —
characteristic of AMMs — along with IDPs could clarify the relationship between them. The
origin of AMMs in particular is important since they constitute the greatest contemporary
mass input to the earth (Maurette 2000), and could have contributed a significant amount
of water and organics to the early earth.

– 8 –
This work was supported by NASA Grant NAG5-10411. We thank David King, Christo-
pher J. Snead, Peter Schultz, and the crew of the AVGR for helping make these tests possible.
We thank John Bradley and Xander Tielens for useful discussions. G. Dominguez would like
to thank the NPSC Graduate Fellowship Program for their support.
Correspondence and requests for materials should be addressed to G.D.
REFERENCES
Amari, S., & Zinner, E. 1997, Nucl. Phys. A, A621, 99c
Amari, S., Nittler, L. R., Zinner, E., Gallino, R., Lugaro, M., Lewis, & Roy, S. 2001a, ApJ,
546, 248
Amari, S., Nittler, L. R. , Zinner, E., Lodders, K, & Lewis, R. S. 2001b, ApJ, 559, 463
Anderson, W. W., & Ahrens, T. J. 1994, J. Geophys. Res., , 99, 2063
Barrett, R. A., Zolensky, M. E., H¨orz, F. , Lindstrom, D. J., & Gibson, E. K. 1992, Lunar
Planet. Sci. , XXII, 203
Brownlee, D. E. et al. 1997, Meteoritics and Planetary Science , 32, A22
Brownlee, D., Joswiak, D. J., Kress, M. E., Matrajt, G., Taylor, S., Bradley, J. 2002, Mete-
oritics and Planetary Science, 37 (Supplement), A26
Engrand, C., & Maurette, M. 1998, Meteoritics and Planetary Science, 33, 565
Flynn, G. J., Bajt, S., Sutton, S. R., & Klock, W. 1995, Meteoritics, 30, 505
Flynn, G. J., H¨orz, F., Bajt, S., & Sutton, S. R. 1996, Lunar Planet. Sci., XXVII, 369
Flynn, G. J., Sutton, S. R., & H¨orz, F. 2000, Lunar Planet. Sci., XXXI, 1457 (2000)
Flynn, G. J. 2002, Meteoritics and Planetary Science, 37 (Supplement), A47
Gr¨un, E., Landgraf, M., Hor´anyi, M.,Kissel, J., Kr¨uger, H., Srama, R., Svedhem, H., With-
nell, P. 2000a, J. Geophys. Res., 105, 10403
H¨orz, F., Zolensky, M. E., Bernhard, R. P., See, T. H., & Warren, J. L. 2000, Icarus, 147,
559

– 9 –
Kitazawa, Y., Fujiwara, A., Kadono, T., Imagawa, K., Okada, Y., & Uematsu, K. 1999,
J. Geophys. Res., 104, 22035
Landgraf, M., Baggaley, W. J, Gr¨un, E., Kr¨uger, H., & Linkert, G. 2000, J. Geophys. Res.,
105, 10343
Maurette, M. 2000, Planet. Space Sci., 48, 1117
Mizushima, Y., & Hori, M. 1995, J. of Mat. Res., 10, 1424
Nittler, L. R., Alexander, C.M. O’d. , Gao, X., Walker, R. M., & Zinner, E. 1997,
Nucl. Phys. A, A621, 113c
Taylor, A. D., Baggaley, W. J., & Steel, D. I. 1996, Nature, 380, 323
Westphal, A. J., Phillips, M., & Keller, C. 1998, New Astr. Rev., 42, 237
Zolensky, M. E., Pieters, C., Clark, B., & Papike, J. J. 2000, Meteoritics and Planetary
Science, 35, 9
This preprint was prepared with the AAS L
A
T
E
X macros v5.0.

– 10 –
Fig. 1.— Fluorescence images of the surfaces of a Gd,Tb-doped alumina aerogel shot with
monodisperse glass (diameters=2, 5, and 20 µm) beads at various velocities at NASA Ames.
Images were taken with a Zeiss Axiophot fluorescence microscope with an attached Optron-
ics DEI-750 3 chip analog camera. Image contrast was enhanced to improve visibility of
fluorescence.

– 11 –
100101102103104
101
102
103
104
105
106
107
Ekinetic (ergs)
Integrated Fluorescence (Green)
Integrated Fluorescence Intensity vs. Kinetic Energy
Ig ∝ E0.69
20 µ m
5 µ m
2 µ m
Fig. 2.— Inet
gvs. Kinetic Energy of Hypervelocity Projectiles. The error bars are statistical
only.

– 12 –
Fig. 3.— A. Fluorescence image of a cleaved track produced by a particle ≃30µm in
diameter and initial velocity equal to 4.72 km/s. B. The ratio of green to blue, indicating
the degree to which the sampled track region has been transformed into the fluorescent
phase. C. Fluorescence signal in the green channel sampled along the track. Notice that the
fluorescence drops off significantly at the end of the track. Image contrast was not enhanced.

– 13 –
10−1 100101
100
101
102
103
104
105
106
Particles Collected > size
particle size (µm)
Fig. 4.— Cummulative number of interstellar dust grains vs. size expected to be collector
in low-earth orbit (3) compared to the Stardust mission (2). The collecting area of the
low-earth orbit collector is assumed to be =3 m2, exposed for 3 years (30 % duty cycle). The
extrapolated portion of the calculation assumes an IS flux that falls off as m−1.1.