Fluffy OM From Petrologic Type ≤3 Chondrites
The structure and nanostructure of the fluffy OM from the most primitive objects (Orgueil, Tagish Lake, Murchison, and Semarkona) are similar at the TEM scale. This observation indicates that their respective precursors were probably structurally similar and that their different aqueous and thermal histories did not significantly influence their structural evolution. Their polyaromatic layers are small (<1 nm), randomly oriented and mainly unstacked (Figs. 2 and 3) (Derenne et al. 2005; Garvie and Buseck 2006). The structure of the organic matter from the three aqueously altered meteorites (Orgueil, Murchison, Tagish Lake) are expected to be similar as they all escaped significant heating. Interestingly, Semarkona, less altered and probably more heated, also exhibits similar structure and nanostructure for its fluffy OM at the scale of HRTEM imaging. However, the Raman spectra of these chondrites are different (smaller full width at half maximum for its defect band; FWHM-D = 228 ± 20 cm−1 for Semarkona compared to FWHM-D > 300 ± 30 cm−1 in Orgueil and Murchison; Quirico et al. 2003, 2005a). As the bulk composition of the acid residue from Semarkona shows lower H/C and N/C ratios than in Orgueil and Murchison (Alexander et al. 2007), the discrepancy between Raman and TEM results may suggest that the spectral differences mainly originate from the chemistry and the nature of the functional groups. Indeed, the polyaromatic structures, as seen by HRTEM, are similar. This implies that the maximum temperature possibly encountered by Semarkona (260 °C; after Alexander et al. 1989) does not permit the growth of the aromatic layers above the resolution threshold attainable by high resolution TEM. This is usually also the case for terrestrial coals, which evolve chemically, but are almost not structurally modified (i.e., without noticeable layer length change) below 200 °C, as imaged by HRTEM. Furthermore, Murchison and Orgueil may have been subjected to different alteration temperatures, respectively, up to 80 °C (Baker et al. 2002) and up to 150 °C or more (Zolensky et al. 1989). However, their fluffy OMs do not record detectable structural differences. Conclusively, the similar structures and nanostructures of the fluffy OM thus strengthen the assumption that these meteorites accreted precursors similar from the structural point of view (even if their isotopic and molecular compositions may have been different), which were not significantly modified by aqueous or thermal processes in the temperature range 80–260 °C.
Increasing Organization Degree along Metamorphism
From chondrites of petrologic types 3.1 to 3.6, the fluffy OM becomes progressively more organized (Figs. 2–6), probably under the influence of the parent body thermal metamorphism. HRTEM provides an estimation of the evolution of the polyaromatic layer sizes. They grow from less than 1 nm in chondrites of petrologic type ≤3.0 to less than 3 nm in type 3.1 chondrites and are finally in the range 5–10 nm in type 3.6 chondrites, the longest fringes reaching 20 nm. This growth is accompanied by a higher number of stacked layers.
These fluffy OM are characterized by a turbostratic biperiodic order that is clearly different from the triperiodic order of perfect graphite. Hence, this structural evolution should not be called “graphitization” as it mainly involves carbonization. This process, in terrestrial coals for instance, consists of the progressive loss of H, N, and O atoms (Rouzaud and Oberlin 1990). In chondrites, however, only H and N are lost in the course of metamorphism (H/C ratio decrease from 0.8 to 0.1 between type 1 and type 3.6 chondrites), whereas O seems to be retained (Alexander et al. 2007).
This is in contrast with terrestrial systems where O is usually among the first hetero-atom to be lost. Two possible reasons may account for this discrepancy (1) the different nature of the precursors (more frequent stable oxygen bridges in chondritic IOM) as well as (2) the absence of confinement pressure on the parent body (Quirico et al. 2009).
Both Raman spectroscopy and TEM are sensitive to the degree of structural organization of the chondritic carbonaceous matter, but at a different scale. Indeed, the improvement of the degree of organization along increasing petrologic type seen by HRTEM is consistent with previous conclusions obtained by Raman spectroscopy (Bonal et al. 2006, 2007; Busemann et al. 2007). We show here that the whole evolution range described by Raman parameters between petrologic type 1 and type 3.6 corresponds to a noticeable, but limited development of the size of the polyaromatic layers, not exceeding a few nanometers. The increased size of polyaromatic domains in the petrologic type >3 are consistent with the previous observation in the same meteorites of increased intensity of the 1s-σ* exciton (Cody et al. 2008) that is shown to correlate with an increase in the number of delocalized electrons. The “1s-σ* exciton” absorption band was attributed in graphite to the presence of planar domains of highly conjugated graphene sheets (Batson 1993).
In agreement with previous works (Bonal et al. 2006, 2007; Alexander et al. 2007; Cody et al. 2008; Remusat et al. 2008; Quirico et al. 2009), we suggest that this evolution, correlated with petrologic type, is controlled by the extent of parent body thermal metamorphism. In this study, HRTEM data, when compared with terrestrial IOMs and laboratory heating experiments, add a line of evidence to the previous conclusions based on chemistry and spectroscopy data. The observed nanostructures, which are similar in all chondrite groups (with the exception of the enstatite chondrite Sahara 97096, discussed below), can only be produced by heating over long time scales.
The parameters controlling carbonization, subsequent structure improvement, and finally graphitization are the chemical and structural nature of the precursor, the pressure, the heating duration, and the temperature (Deurbergue et al. 1987; Oberlin 1989; Beyssac et al. 2002, 2003; Bonal et al. 2006). In chondritic IOMs, the high reticulation degree of the polyaromatic moieties, maintained by cross-linkers, such as oxygen, is responsible for their random orientation and gives a nongraphitizing character to the precursor. This implies that graphite cannot be obtained without applying confinement pressure (Oberlin 1989; Beyssac et al. 2003). The latter is generally considered negligible within asteroidal parent-bodies, due to their small size and low gravity (Bennett and McSween 1996). Temperature and heating duration are therefore the main parameters controlling the evolution of the IOMs in chondrites, and we discuss below their relative influences.
If it is clear that temperature mainly controls carbonization and graphitization processes, an important kinetic barrier is associated with the sluggish graphitization mechanism, and heating duration should also be considered (Beyssac et al. 2003). Geological time scales may be necessary to overcome some of the reaction steps (loss of heteroatoms, polyaromatic layer growth, defect resorption, stacking development, and, finally, establishment of a three dimensional organization). Indeed, at atmospheric pressure, high temperature laboratory pyrolysis (T > 2600 °C) of nongraphitizing material, such as saccharose-based chars, yield highly disordered carbon if confinement pressure is not applied (Rouzaud and Oberlin 1989; Beyssac et al. 2003; Bernard et al. 2010). Even if these synthetic carbons are not produced by the same processes as chondritic IOM, they are structurally similar enough to be used for comparison of their graphitization behavior.
For instance, the pyrolysis of Orgueil IOM at 1000 °C yields a degree of structural organization much lower than what is observed in Allende, whereas Allende parent body has not been heated to such a temperature on its asteroidal parent body. Therefore, if one assumes that chondrites have structurally similar precursors, long heating duration is probably necessary, in the absence of pressure, to reach the organization degree of the fluffy OM observed in the higher petrologic types. We conclude that the nanostructure observed in petrologic type >3.1 chondrites are not likely to be produced by short-time scale heating events (shock metamorphism, Le Guillou et al. 2010; or nebular transient heating for instance), and that the most plausible thermal event is a long-term heating, compatible with the 26Al decay on the parent body over millions of years (Bennett and McSween 1996).
A given structural organization degree, as quantified by Raman spectroscopy, could potentially be obtained by different combinations of duration/temperature conditions of thermal treatments. However, the direct observation of the carbon nanostructures by HRTEM shows that chondritic IOMs from all chemical groups display similar structures and nanostructures for a given petrologic type, which is a strong indication that chondrites (1) accreted precursors relatively similar in terms of structures, (2) recorded comparable temperature-duration pathways (with the exception of Sahara 97096, see below).
Origins of Heterogeneities within Fluffy OM of Petrologic Types >3.0
Along petrologic types, the fluffy OM evolves from relatively homogeneous compact residues to assemblages of more structurally and nanostructurally heterogeneous particles. We can divide this heterogeneity into two groups. The first one corresponds to the development of a structural heterogeneity within the fluffy OM itself. It is illustrated by images of fluffy OM from petrologic types >3.6, which show a wider distribution of fringe length, as well as a more variable number of stacked fringes from one particle to the other (Fig. 5). The second type of heterogeneity occurs at a larger scale (greater than approximately 50 nm) and relates to the presence of particles with a clearly higher structural organization degree and showing different nanostructures (mesoporous and lamellar) as described in Allende (Fig. 8).
Laboratory experiments, such as pyrolysis under pressure (Beyssac et al. 2003) or iron-catalyzed graphitization at temperature above 1150 °C have shown that structural and nanostructural heterogeneities are systematically produced. Heterogeneities are also observed in terrestrial metamorphism (Deurbergue et al. 1987). This inherent nature of the carbonaceous material behavior along metamorphism could explain some of the final heterogeneity of the fluffy OM. However, we propose here two additional processes, specific to chondrite metamorphism, which could also contribute to this heterogeneity.
First of all, carbon grains in chondritic matrices are characterized by variable sizes from the submicron scale to below the nanometer scale (Brearley 1999; Floss and Stadermann 2009; Le Guillou et al. 2011). We suggest that the initial size distribution, for a given time-temperature metamorphic pathway, influences the final size distribution and structural parameters of the polyaromatic layers. Very small particles scattered through the inorganic materials of the matrix cannot grow as much as the larger one, thus producing layers of heterogeneous sizes. This mechanism would also imply that the initial particle size distribution, at the onset of metamorphism, could be an important parameter influencing the structural evolution of the IOMs: the larger the initial particles, the higher the final averaged graphitized domains potentially attainable. Furthermore, we suggest that the mineral environment may also play a role, by orienting the growth of the aromatic layers along its surfaces (template effect). This surface-controlled growth mechanism is observed experimentally during the carbonization of silicates-associated organic matter (Beguin et al. 1996; Zhai et al. 2008). The long and continuous fringes observed in Tieschitz (Fig. 7) show angles and appear like they grew around a faceted mineral, supporting this hypothesis. We may consequently infer that the mineral environment plays a role, as IOM scattered through chemically and morphologically heterogeneous phases could produce heterogeneous nanostructures.
The second category of heterogeneities (e.g., the presence of graphitized particles), which had already been described in previous work (Smith and Buseck 1981; Brearley 1990; Vis et al. 2002; Remusat et al. 2008) is difficult to reconcile with low temperature parent body metamorphism, and may rather relate to high temperature processes. It is indeed unlikely that metamorphism would have produced such a high graphitization degree on a limited number of particles, especially because we have not observed intermediate graphitization degree between the fluffy OM and those particles. In contrast, graphitized particles have previously been reported in ordinary and carbonaceous chondrites by Raman and TEM, and are associated with metal (Brearley 1990; Mostefaoui et al. 2000, 2005). Some of these metal/graphitized particle associations were sometimes even found in close contact with chondrules, for instance in Bishunpur (Mostefaoui et al. 2000). At this point, even if it is not possible to completely exclude metal-related parent body metamorphism, it is more probable that the particles observed in our residues have been formed by metal-catalyzed reactions at high temperature prior to parent body accretion, a mechanism which is experimentally supported (Audier et al. 1981; Audier and Coulon 1985; Charon et al. 2011). Xenoliths with different thermal history, included within the host meteorite, and carrying graphitized particles could account for the heterogeneity. However, the chondrule forming event would better fit those observations, and especially the carbon metal association found by previous authors.