Comets are generally assumed to have formed in the cold outer regions of the early solar nebula and therefore represent assemblages of essentially unaltered interstellar material. The current view of solar system cosmogony is of comets and planetary bodies forming together within the nebular disk encircling an already existing Sun. Within the disk, grains are thought to have accumulated first into kilometre sized cometary bodies and then into larger planetary ones (Bailey 1994). In this scheme the outer planets, at least in part, represent cometary aggregates, whilst comets themselves represent the primitive building blocks of the present-day solar system. The chemistry and mineralogy of comets observable at the present epoch ought therefore have much to tell us about conditions prevalent in the early solar nebula as they should represent repositories of the unused leftover materials from which the rest of the solar system was built.
Originally, Oort cloud comets were thought to have formed in the region between Jupiter and
Neptune, approximately 5-30 AU from the Sun, and subsequently to have been scattered to the Oort
cloud by gravitational interactions (Oort 1950). Later Safronov (1969, 1972) proposed that
Uranus and Neptune, instead of Jupiter and Saturn, were responsible for dynamic scattering to the
Oort cloud. This basic idea was confirmed later still by the detailed calculations of
Fernandez & Ip (1981, 1983) and Ip and Fernandez (1988). Since it is now generally accepted that Jupiter
would have scattered planetesimals largely to gravitationally unbound orbits, the majority of the
present day Oort cloud comets are likely to have formed either well outside the orbit of Jupiter
(e.g. Delsemme 1999), or formed at the Oort cloud distance itself (Hills 1982;
Hills & Sandford 1983; Bailey 1987). Edgeworth-Kuiper belt comets on the other hand are assumed
to have formed in situ and are therefore likely to have sampled even more distant, and hence even
cooler, regions of the early solar nebula (out to 
45 AU, Weissman 1995).
The grains of interstellar molecular cloud dust constituting the initial material present in
the pre-solar nebula have long been known to contain a significant proportion of Mg and Fe rich
silicate grains typically with olivine and pyroxene compositions (Tielens & Allamandola 1990).
These grains reveal themselves via characteristic infrared resonances at 10 
m and
20 m and previously much work has been done on the optical characterisation of
analogue amorphous silicates in the laboratory (e.g. Nuth & Donn 1982; Hecht et al. 1986;
Dorschner et al. 1988; Stephens et al. 1995). The bands at 10 m and 20 m
for amorphous silicates are well known to be characteristically broad and devoid of any fine
structure that would otherwise be indicative of the silicate grain material possessing some form
of structural symmetry. Given that comets formed in the cold outer regions of the early nebula
where conditions would have been very close to those of the interstellar medium
(Napier & Clube 1997), amorphous silicate grains can be expected to be well represented in comet
mineralogical compositions. The detection therefore of fine structured features, apparently
characteristic of crystalline dust, in the spectra of certain comets is thus both surprising and
problematic. This is because such grains remain undetected in the interstellar medium and are
generally regarded as being the product of annealing processes more characteristic of the hotter
inner-nebular regions rather than the cooler outer zones. Depending on size and composition,
grains from the interstellar medium are likely to survive the change from the initial cold
pre-solar molecular cloud to the hotter solar nebula as far in as 1 to 3 AU from the Sun
(Chick & Cassen 1997). The distance at which grains would become hot enough to crystallise is only
weakly constrained as it requires detailed knowledge of the physical properties of the grain
material itself.
The first piece of spectroscopic evidence suggesting the existence of crystalline silicates
came in the form of a narrow sub-feature at 11.2 m embedded in an otherwise amorphous
10 m silicate band. This sub-feature has been observed in several comets, including Comet
Halley and Comet Hale-Bopp (e.g. Bregman et al. 1987; Campins & Ryan 1989; Crovisier et al. 1996;
Russell & Lynch 1996; Hayward & Hanner 1997). Furthermore, the presence of an 11.2 m feature in at least
one dynamically young comet (Hanner et al. 1994a), along with certain laboratory evidence
constraining the annealing conditions required for amorphous grains to undergo crystallisation,
strongly suggests that crystallisation of the dust grains must have occurred prior to their
inclusion in the comet body, as the temperatures experienced by comets during their short
passage near the Sun would not be enough to promote grain crystallisation (Hallenbeck et al. 1998).
Early laboratory work on crystalline silicates (e.g. Koike et al. 1981, 1993) led to the
11.2 m feature being widely identified by various authors with the sharp feature present in
the spectrum of Mg-rich olivine. Colangeli et al. (1995, 1996) have produced fits to the
10 m bands of several comets with 
m features using laboratory spectra
measured for different classes of crystalline silicate minerals. Observations of the Comet
Hale-Bopp coma using ISO, covering the region 6-45 m (Crovisier et al. 1997), revealed a
rich variety of strong emission features attributed mainly to the presence of forsterite olivine
grains (Brucato et al. 1999a). Ground based observations by Wooden et al. (1999) made when
Hale-Bopp was close to perihelion have also revealed the presence of a sub-feature at 9.3 m which, based on the laboratory measurements of Koike et al. (1993), was attributed by them
to Mg-rich pyroxene (enstatite) grains. Observations of certain Kuiper belt comets
(Crovisier et al. 1999) have also shown the presence of an m feature.
A link between comet dust and the dust of the debris disk surrounding the star 
Pictoris
was provided by Knacke et al. (1993) who compared
Pictoris observations with Comet
Halley, while Greenberg & Li (1996) showed that
Pictoris dust can be modelled using
porous cometary-like particles, typically 10 m in diameter, built from aggregates of smaller
sub-micron sized components. However the band-to-continuum ratio for
Pictoris was found
to require the existence of a significant population of grains smaller than a few microns
(Sitko et al. 1999). As these would be easily swept away by radiation pressure from the central
star, they would need to be continuously replenished by the erosion of larger objects such as the
Falling Evaporative Bodies identified in Herbig Ae/Be systems as in-falling star-grazing comets
undergoing rapid loss of both gas and dust close to the central star (Ferlet et al. 1987;
Hobbs 1986). It would appear therefore that comets are an intrinsic feature of the formation
and evolution not only of our own solar system, but also of the other planetary and
proto-planetary systems now being discovered. The presence of crystalline silicate dust may also
reflect the early evolution of such systems. Sitko et al. (2000) have proposed that the mid-IR
spectra of Herbig Ae/Be stars evolve as a function of stellar age and form a spectral sequence
starting with features easily characterised by amorphous astronomical silicate, through to
features consistent with processed crystalline olivine grains. Nuth et al. (2000) have suggested that
the ageing of nebular dust grains, reflected in the formation of crystalline structure, could
form the basis of a method for dating comets themselves since those that formed later on in the
nebular history are likely to contain a higher proportion of grains with crystalline features.
In unravelling the relationship between nebular dust processing, comet formation and the evolution of the solar nebula itself it is necessary to understand how silicate materials behave when annealed in order to place constraints on the thermal histories of grains both before and after comet formation and hence on the possible conditions prevalent in the solar nebula. It has therefore become necessary to study in the laboratory the structural transformations occurring within thermally processed silicates. In particular we need to study the changes that occur as the silicate undergoes crystallisation as previous comparisons between laboratory and cometary spectra have shown cometary silicates to be best represented by laboratory spectra that fall somewhere between amorphous and fully crystalline (Hallenbeck et al. 1998). This being the case, it also raises an additional question as to what, in an astrophysical context, the term crystalline silicate grain actually describes.
Being directly comparable to the astronomical data, infrared spectroscopy is a suitable technique to use in the laboratory. However it is limited in the amount and type of structural information it can return about the material under investigation and for this reason we have combined this astronomically appropriate method with the high resolution structure probing capabilities of synchrotron X-ray powder diffraction (XRD). In doing so, we have obtained information on the structural evolution taking place within an annealed silicate that suggests possible constraints on the interpretation of its spectral evolution.
Copyright ESO 2002