Of the 15,000-60,000 tons of extraterrestrial material that bombards the Earth per annum, 95% is in the sub-micron size range, e.g. [1]. This cosmic dust may be a more representative sample of small body material in the solar system than are meteorites. The possible origins of the dust may include comets, asteroids and interstellar dust etc., e.g. [2]. Many of the laboratory investigations looking at dust particles have been based on material collected from a number of terrestrial locations: i.) deep-sea spherules collected from the sea-floor [3], ii.) micrometeorites collected from the Antarctic Ice Sheets [4] and iii.) interplanetary dust particles (IDPs) collected from the stratosphere [5], These collections have been extensively investigated over the past 30 years using a wide range of analytical techniques [6]. Yet, whilst substantial information has been obtained in terms of mineralogical and chemical compositions and processes within the particles, the transit through the Earth's atmosphere does make them susceptible to selection and modification process, e.g. [7], Therefore, it is almost impossible to unambiguously state the parent body origin of the particles. However, in this first decade of the twenty first century, we should see samples returned from missions that have been purpose-built to collect particles from examples of the two most likely sources of dust: i.) NASA STARDUST mission to Comet Wild 2 [8] and ii.) The Japanese MUSES-C mission to an asteroid [9]. This paper will discuss techniques that might be applicable to the particles that we hope will be returned in 2006 by the STARDUST mission.

Capturing intact particles from a comet will not be an easy task as the impact encounter velocities will be >5 km/s. Studies of particles sampled in low Earth orbit (LEO) at similar velocities using a variety of dedicated collector surfaces, e.g. [10] and non-dedicated surfaces, e.g. [11] have shown that the particles undergo extreme shock and melting and are generally highly altered, if not destroyed completely. To overcome these problems and to collect near-intact and pristine particles, STARDUST is using silica aerogel [12], a low-density material. Previous experimental work, e.g. [13] and [14] has shown the suitability of the aerogel as a capture cell and the technology has already been used to collect particles in low Earth orbit, e.g. [14] and [15].

Previous studies, e.g. [15], have focused on the interpretations of the physical characteristics of the impacted particle and the associated features in the aerogel generated by the hypervelocity impact event, e.g. track length and morphology [14]. In this paper we present chemical analyses and proxy crystallographic information from olivine particles captured in aerogel in the laboratory.


The laboratory experiments were carried out using a two stage light-gas-gun at the University of Kent. The aerogel targets for the study had a density of 96 kg m3. The olivine projectiles (125-250(im diameter grains from the coarse Admire pallasite, approximate homogeneous composition:- forsterite 88 : fayalite 12) were selected on the basis that mafic silicates are one of the major mineral components of cosmic dust particles. The projectiles were sabot-mounted and accelerated in the aerogel targets at approximately 5.1km/s using a buck-shot technique.

The raman spectroscopy was carried out on an Instruments SA model HR640 spectrograph with a liquid nitrogen cooled CCD detector, linked to an Instruments SA Raman scattering module based on an Olympus BX40 microscope. The analytical conditions for the work were: a spot size of approximately 25(im, and a He-Ne laser (wavelength 632.8 nm) was used at 20 mW power.

The electron microscopy was carried out using a Jeol 840 scanning electron microscope at 2nA beam current and at 20kV accelerating voltage. The samples were carbon-coated to reduce the effects of electrical charging during SEM investigations. The X-ray elemental maps and energy dispersive X-ray spectra (EDS) were acquired using an Oxford Instruments eXL energy-dispersive spectrometer microanalyser (EDS) with a Pentafet detector fitted with an ultra-thin window (this allows the detection of X-rays from light elements - e.g. carbon). A full description of the analytical protocol for this type of work is given in [16].


When the STARDUST samples are returned to Earth, one of the first fundamental tasks will be to identify the bulk chemical composition of the particles. This can be carried out using a number of analytical techniques - see [6] for a comprehensive review. The choice of which technique to apply is clearly important, for example whilst analytical SEM and electron microprobe analysis are the conventional techniques in many meteoritic investigations, the samples require a conductive coating (typically carbon or gold). Coating the samples with carbon would potentially contaminate them to the extent that highly specialised analytical techniques such as carbon stable isotopic mass spectrometry [17] can not be carried out reliably. The chemical information derived from X-ray microanalysis may give strong clues as to the likely composition of the particle, but is unlikely to unambiguously reveal the full mineralogical identity, as this also requires crystallographic parameters to be determined. X-ray microanalysis is also unable to differentiate materials of similar composition but widely differing states of shock deformation. Therefore, it is desirable to use a technique that is nondestructive, does not contaminate the sample, can provide some crystallographic information by proxy and has a small spatial sampling area, for example: raman spectroscopy. This technique obtains a vibrational spectrum of the crystal lattice of the sample by illuminating it with a laser and collecting the spectrum of inelastically scattered photons. The detected spectrum contains a suite of peaks whose identity and morphology are characteristic of the chemical structure of the sample. In other words raman spectroscopy is a chemical analysis technical rather than an elemental one, it asks "what is it?" rather than "what is it made of?". As each component of the sample that is illuminated by the laser may contribute to the spectrum seen, it is fortunate that the aerogel does not significantly absorb or scatter the laser light. Consequently, only the selected impacted particle material is analysed.

The olivine grains trapped in the aerogel following a light-gas-gun shot were subjected to this technique and dispersive raman spectra were obtained (figure 1). The raman spectrum obtained for the impacted olivine showed a wider peak width than that obtained from the analysis of a pristine grain. This may indicate that the particle has undergone a degree of alteration at the crystallographic scale. A previous investigation [18] studied the thermal and mechanical alteration that particles impacted into aerogel targets may have undergone.

Whilst raman spectroscopy does appear to be a very good technique to carry out in-situ mineralogical analysis of the trapped particles, a separate study [19] has suggested that during analysis the there can be small-scale localised thermal alterations to the examined particle. This observation clearly requires further work to determine the extent of damage.

wavenurrfcer / crrr'

Figure 1. The raman spectra obtained for an olivine grain captured in aerogel and a unaltered grain.

wavenurrfcer / crrr'

Figure 1. The raman spectra obtained for an olivine grain captured in aerogel and a unaltered grain.

Figure 2. Secondary electron image of 3 tracks generated in a block of aerogel by olivine projectiles.

After the raman spectroscopic analysis, trapped grains were examined in a SEM, using a combination of secondary electron imaging (SEI), back-scattered electron imaging (BEI) and X-ray elemental microanalysis. The key goal of the SEM study was to locate the position of the particles within tracks, and if possible examine their surface morphology. To enable the detailed examination of the trapped grains the aerogel block was cut so that the penetrating track could be seen at, or very close to, the surface (figure 2). It is possible to see that there were at least three major tracks generated in this small block. It appears that in one of the tracks the impacting particle had exploded on penetration and fine debris was dispersed along the length of the track - obviously this particle was not 'intact' following capture.

As the aerogel substrate has a very simple bulk composition, i.e. Si K-alpha dominates EDS spectra of pure aerogel, the best technique to locate extraneous impactor debris is X-ray elemental mapping for Mg and Fe K-line emission. This can rapidly define the position of the olivine grains. In our study several individual grains were located by this method (figure 3) and, interestingly, it was observed that the grains typically had a layer of aerogel adhering to their surface. EDS spot analysis of the individual grains produced a spectrum typical of an olivine (figure 4). From this very preliminary investigation it is not possible to determine whether the olivine grains have undergone any shock or crystallographic alteration as a result of the impact event. It is planned that the analytical work will be extended to include the use of transmission electron microscopy, in which it will be possible to examine the crystallographic structure of grains extracted from the aerogel.

Figure 3. Secondary electron image of an individual olivine grain captured in the aerogel.

Was this article helpful?

0 0

Post a comment