Micro-m´et´eorites antarctiques

Par opposition aux IDPs qui sont collect´ees en altitude, les microm´et´eorites antarc- tiques (AMMs) sont des poussi`eres extraterrestres (d’une taille comprise entre 20–500 µm) r´ecolt´ees au sol. Ces r´ecoltes ont donn´e lieu `a quatre campagnes de collectes `a la station franco-italienne de Concordia Dˆome C en 2000, 2002, 2006 et 2013. Ces microm´e- t´eorites montrent un certain degr´e de d´esordre structural. La plupart de ces grains ont une structure d´esorganis´ee, nanoporeuse, avec une proportion variable de carbone (entre 22% et 58%). Une petite fraction (≤ 3%) de ces AMMs est extrˆemement riche en carbone (≥50% en volume, Dobrica et al. 2008). Ces AMMs sont appel´ees microm´et´eorites an- tarctiques ultra carbon´ees (UCAMM)s. D’un point de vue structural, les AMMs peuvent ˆetre regroup´ees suivant quatre morphologies distinctes : `a grains fins (Fg), scoriac´ees (Sc), interm´ediaires (Fg-Sc) et ultra-carbon´ees (UCAAMs). La plupart de ces poussi`eres ont montr´e des indices Raman de chauffage rapide, r´esultant de l’entr´ee et de l’oxydation atmosph´erique (Dobrica et al. 2011). Ces compos´es carbon´es montrent ´egalement une cer- taine continuit´e de texture et de composition avec les chondrites carbon´ees (Dobrica et al. 2011). Une certaine proportion d’AMMs (15%) contient des mol´ecules solubles complexes comme des acides amin´es, en proportion plus importants que dans les chondrites carbon´es CM (Matrajt et al. 2004). Plus g´en´eralement, les analyses men´ees sur les microm´et´eorites (voir notamment Dobrica et al. 2009) ont permis de mettre en ´evidence un continuum de composition (pr´esence de mol´ecules organiques complexes comme PAH et des acides amin´ees) et de structure (mati`ere carbon´ee amorphe) entre les grains com´etaires et la ma- ti`ere fine (IDPs, AMMs) issu d’ast´ero¨ıdes, malgr´e des scenarii et des lieux de formations diff´erents de ces types d’objets.

(Fig. 4). It has been suggested that this main trend in UOCs may be due to a parent body process (10). However, it is difficult to conceive that the main trend from the UCAMMs may result from the same process, because (i) the UCAMMs and UOCs strongly differ both chem- ically and mineralogically [in particular, the UOC organic matter concentration (<0.5 wt %) (10)

is much lower than that observed in UCAMMs]; (ii) we do not observe any sign of thermal pro- cessing of the UCAMMs and, by contrast to the UOCs, we observe for C/H > 3 a constant D/H plateau rather than a correlated increase of D/H with C/H (fig. S2) (13); and (iii) the data presented here (Fig. 3) show that the whole range of var- iation of the main trend coexists within a few tens

of square micrometers, whereas the UOC main trend concerns IOM residues from the bulk mete- orites. Therefore, the UCAMM main trend seems compatible with the sampling of a heterogeneous organic matter reservoir.

Above the main trend, the data broadly spread toward extreme D/H ratios with D/H > 2.5 × 10−3 and C/H = 2 to 6 (Fig. 4). The D-rich hot spots in IDPs for which C/H ratios have also been reported span a large range (0 < C/H < 3) (4), whereas the hot spots from IOMs of CR2 prim- itive chondrites have C/H ratios limited to a more restricted zone (1 < C/H < 1.5) (5). The UCAMM extreme D/H component seems to extend the high C/H trend observed in IDPs, including particles collected during the meteor shower associated with comet 26P/Grigg/ Skjellerup (17).

High D excesses observed in interplanetary materials have long been attributed to interstellar chemistry, because large D enrichments (D/H > 0.01) are observed in the gas phase of cold molecular clouds (7). However, there is a strict upper limit on the fraction of crystalline relative to amorphous silicates in the interstellar medium (<0.2% by mass) (18). If the organic matter from the UCAMMs was a direct heritage of interstellar origin, one would expect the associated minerals to be dominated by amorphous silicates, which is not the case. Quite the opposite, the organic matter of the UCAMMs contains crystalline phases typical of silicates processed within the accretion disk (19), such as those observed both in anhydrous IDPs (20) and in the fine-grained fraction of Wild 2 particles (21). Therefore, the UCAMMs cannot be considered as a direct interstellar heritage but most probably sampled material (organic matter and minerals) from the protoplanetary disk itself.

Substantial D excesses have been identified at the molecular level in IOM from the Orgueil and Murchison meteorites, supporting an exchange mechanism between the organic matter and a local gaseous D-rich reservoir within the nascent solar system (9). Numerous astronomical obser- vations demonstrate the occurrence of deuterated molecules in protoplanetary disks, some of them exhibiting large D/H variations (0.01 < D/H < 0.1) for radial distances between 30 and 70 AU (22). The large range of D/H ratios observed in the UCAMMs may be reminiscent of the D/H gradient that once existed at several tens of as- tronomical units from the young Sun.

Other than the bona fide Wild 2 particles returned by the Stardust mission, the assignment of a cometary or asteroidal origin to a given interplanetary dust particle remains speculative. The unmelted nature of the UCAMMs precludes high atmospheric entry velocities usually asso- ciated with a cometary origin. However, once released from their parent body, the trajectories of dust within that size range substantially evolve Fig. 3. NanoSIMS-50 (secondary ion mass spectrometry) isotopic and elemental maps of UCAMMs.

(A and B) dD (‰) (13) (A) and C/H atomic ratio (B) of particle 19. The contour in (A) indicates a region with low D/H ratio (Fig. 4) (13). (C) dD (‰) map of particle 119. (D) Higher-magnification dD (‰) map of the zone indicated by the white rectangle in (C).

REPORTS

on May 7, 2010

www.sciencemag.org

Downloaded from

(Fig. 4). It has been suggested that this main trend in UOCs may be due to a parent body process (10). However, it is difficult to conceive that the main trend from the UCAMMs may result from the same process, because (i) the UCAMMs and UOCs strongly differ both chem- ically and mineralogically [in particular, the UOC organic matter concentration (<0.5 wt %) (10)

is much lower than that observed in UCAMMs]; (ii) we do not observe any sign of thermal pro- cessing of the UCAMMs and, by contrast to the UOCs, we observe for C/H > 3 a constant D/H plateau rather than a correlated increase of D/H with C/H (fig. S2) (13); and (iii) the data presented here (Fig. 3) show that the whole range of var- iation of the main trend coexists within a few tens

of square micrometers, whereas the UOC main trend concerns IOM residues from the bulk mete- orites. Therefore, the UCAMM main trend seems compatible with the sampling of a heterogeneous organic matter reservoir.

Above the main trend, the data broadly spread toward extreme D/H ratios with D/H > 2.5 × 10−3

and C/H = 2 to 6 (Fig. 4). The D-rich hot spots in IDPs for which C/H ratios have also been reported span a large range (0 < C/H < 3) (4), whereas the hot spots from IOMs of CR2 prim- itive chondrites have C/H ratios limited to a more restricted zone (1 < C/H < 1.5) (5). The UCAMM extreme D/H component seems to extend the high C/H trend observed in IDPs, including particles collected during the meteor shower associated with comet 26P/Grigg/ Skjellerup (17).

High D excesses observed in interplanetary materials have long been attributed to interstellar chemistry, because large D enrichments (D/H > 0.01) are observed in the gas phase of cold molecular clouds (7). However, there is a strict upper limit on the fraction of crystalline relative to amorphous silicates in the interstellar medium (<0.2% by mass) (18). If the organic matter from the UCAMMs was a direct heritage of interstellar origin, one would expect the associated minerals to be dominated by amorphous silicates, which is not the case. Quite the opposite, the organic matter of the UCAMMs contains crystalline phases typical of silicates processed within the accretion disk (19), such as those observed both in anhydrous IDPs (20) and in the fine-grained fraction of Wild 2 particles (21). Therefore, the UCAMMs cannot be considered as a direct interstellar heritage but most probably sampled material (organic matter and minerals) from the protoplanetary disk itself.

Substantial D excesses have been identified at the molecular level in IOM from the Orgueil and Murchison meteorites, supporting an exchange mechanism between the organic matter and a local gaseous D-rich reservoir within the nascent solar system (9). Numerous astronomical obser- vations demonstrate the occurrence of deuterated molecules in protoplanetary disks, some of them exhibiting large D/H variations (0.01 < D/H < 0.1) for radial distances between 30 and 70 AU (22). The large range of D/H ratios observed in the UCAMMs may be reminiscent of the D/H gradient that once existed at several tens of as- tronomical units from the young Sun.

Other than the bona fide Wild 2 particles returned by the Stardust mission, the assignment of a cometary or asteroidal origin to a given interplanetary dust particle remains speculative. The unmelted nature of the UCAMMs precludes high atmospheric entry velocities usually asso- ciated with a cometary origin. However, once released from their parent body, the trajectories of dust within that size range substantially evolve (as a result of resonances with giant planets, radiation pressure, the Poynting-Robertson ef- fect, and solar wind drag), and cometary dust can Fig. 4. Distribution of D/H versus C/H atomic ratios in particles 19 and 119 (13). The data from bulk IOM

from CR (white diamonds), UOCs (black dots) (10), and the range of D-rich hot spots observed in primitive chondrites (5) (white arrow) and that from IDPs (4, 17) (gray surface) are reported.

Fig. 3. NanoSIMS-50 (secondary ion mass spectrometry) isotopic and elemental maps of UCAMMs. (A and B) dD (‰) (13) (A) and C/H atomic ratio (B) of particle 19. The contour in (A) indicates a region with low D/H ratio (Fig. 4) (13). (C) dD (‰) map of particle 119. (D) Higher-magnification dD (‰) map of the zone indicated by the white rectangle in (C).

7 MAY 2010 VOL 328 SCIENCE www.sciencemag.org

744

REPORTS

on May 7, 2010

www.sciencemag.org

Downloaded from

Figure 19. Image nanoSIMS du rapport D/H dans une micrométéorite antarctique ultra carbonée (UCAMM). On remarque la présence de zone micrométriques très enrichies en

deutérium (tiré de Duprat et al. 2010)

Fig. 1.16 – Image nanoSIMS du rapport D/H dans une microm´et´eorite antarctique ultra carbon´ee (UCAMM). On remarque la pr´esence de zones microm´etriques tr`es enrichies en deut´erium (extrait de Duprat et al. 2010).

De mˆeme que pour les IDPs, de tr`es forts enrichissements en deuterium ont ´et´e mesur´es 29

dans les UCAMMs, jusque 30000‡ (Figure 1.16, Duprat et al. 2010). Cette signature tr`es enrichie en deut´erium n’est pas mesur´ee dans les microm´et´eorites Fg ou Sc (Engrand et al. 1999). R´ecemment, il a ´et´e propos´e que leur tr`es fort enrichissement en deut´erium (jusque 5400±2200 ‡) et jusque δD > 20,000‡ dans certaines zones microm´etriques et en15N (jusque d15N = 95±30 ‡) pourrait ˆetre expliqu´es par l’irradiation ionisante dans les zones tr`es externes du disque (et donc tr`es froides) de glace form´ee sur le corps parents des com`etes et riche en azote et carbone, pour des dur´ees d’expositions aux rayonnements stellaires allant jusque plusieurs milliards d’ann´ees (Dartois et al. 2013).

1.4

Les silicates hydrat´es dans le disque protoplan´e-

taire

La question des silicates hydrat´es dans le disque est avant tout celle de l’origine de l’eau terrestre. En effet, d’apr`es les multiples scenarii existants, l’eau proviendrait soit d’un apport endog`ene au moment de l’accr´etion du corps plan´etaire (voir Drake 2005), soit d’un apport tardif, par l’interm´ediaire des com`etes ou de mat´eriel chondritique (voir Albar`ede 2009). Cependant, si les com`etes semblent constituer des candidates id´eales, elles pourraient ne pas avoir particip´e pour plus que 10% au budget total de l’eau terrestre comme le montre l’´etude de Dauphas and Marty (2002) `a partir des calculs de bilans de masses des gaz rares entre la Terre et les possibles impacteurs (com`etes et ast´ero¨ıdes). La Terre pourrait ´egalement avoir acquis une majorit´e de son eau lors de son accr´etion, les poussi`eres l’ayant form´e pouvant contenir jusque 4 oc´eans, en fonction de la temp´erature r´egnant dans la zone o`u s’est form´ee la Terre (King et al. 2010). La question est alors : Est-ce que des silicates hydrat´es sont aptes `a se former dans le disque, avant l’accr´etion, et demeure t-il une trace de leur pr´esence pass´ee ?

Les calculs th´eoriques de Prinn et Feigley (1989) ont montr´e que la condensation d’un phyllosilicate dans des conditions de la n´ebuleuse solaire ´etait un processus trop long au regard de la dur´ee de vie du disque. Pour expliquer la pr´esence de phyllosilicates dans les chondrites carbon´ee, les auteurs ont donc propos´e une alt´eration dans le corps parent `

a partir d’enstatite et de forst´erite. Cette alt´eration est possible pour des plan´et´esimaux suffisamment gros pour avoir de l’eau circulant sous forme liquide. Des phyllosilicates comme la serpentine, la brucite et le talc seraient alors aptes `a se former (Fegley, 2000). Cependant, les ´energies d’activation des r´eactions entre min´eraux primaires et secondaires sont possiblement surestim´ees (Bose et Ganguly 1995). De plus, les auteurs ne s’int´eressent qu’`a des phases cristallis´ees. Or il a ´et´e montr´e depuis que les silicates nourrissant le disque protoplan´etaire, issus du milieu interstellaire, sont tr`es majoritairement amorphes et leur possible hydratation dans le disque, est tr`es peu ´etudi´ee. La possibilit´e d’une origine pr´e- accr´etionnelle est renforc´ee par la pr´esence autour de certains chondres d’une couronne de grains riches en phyllosilicates (Metzler et al. 1992). La pr´esence de cette couronne pourrait r´esulter d’une alt´eration pr´e-accr´etionnelle. Une condensation de phyllosilicates dans la n´ebuleuse solaire `a partir d’ondes de chocs dans le disque a ´et´e propos´ee pour expliquer ces couronnes (Ciesla et al. 2003). De tels ´ev`enements, augmentant localement les conditions P,T, seraient `a-mˆeme d’expliquer ´egalement la formation des chondres. En

Les silicates hydrat´es dans le disque protoplan´etaire

effet, les chondres montrent des indices min´eralogiques d’´episodes de chauffages ´eclairs et de refroidissements rapides, jusqu’`a 1000 K/heure (e.g., Boss et al. 1996).

...

The building blocks of planets within the ‘terrestrial’ region of protoplanetary disks

R. van Boekel1,2_{, M. Min}1_{, Ch. Leinert}3_{, L.B.F.M. Waters}1,4_{, A. Richichi}2_,

O. Chesneau3_{, C. Dominik}1_{, W. Jaffe}5_{, A. Dutrey}6_{, U. Graser}3_{, Th. Henning}3_,

J. de Jong5_{, R. Ko¨hler}3_{, A. de Koter}1_{, B. Lopez}7_{, F. Malbet}6_{, S. Morel}2_,

F. Paresce2_{, G. Perrin}8_{, Th. Preibisch}9_{, F. Przygodda}3_{, M. Scho¨ller}2

& M. Wittkowski2

1_{Astronomical Institute “Anton Pannekoek”, University of Amsterdam,}

Kruislaan 403, 1098 SJ Amsterdam, The Netherlands

2_{European Southern Observatory, Karl-Schwarzschild-Strasse 2, D-85748}

Garching, Germany

3_{Max-Planck-Institut fu¨r Astronomie Heidelberg, Ko¨nigstuhl 17, 69117}

Heidelberg, Germany

4_{Instituut voor Sterrenkunde, K.U. Leuven, Celestijnenlaan 200B, 3001 Heverlee,}

Belgium

5_{Leiden Observatory, Niels Bohrweg 2, 2333 CA Leiden, The Netherlands} 6_{Observatoire de Bordeaux 2, rue de l’Observatoire F-33270 Floirac, France} 7_{Observatoire de la Coˆte d’Azur, De´partement Fresnel UMR 6528, BP 4229,}

06034 Nice Cedex 4, France

8_{Laboratoire d’Etudes Spatiales et d’Instrumentation en Astrophysique,}

Observatoire de Paris, section de Meudon, 5 place Jule Janssen, 92190 Meudon, France

9_{Max-Planck-Institut fu¨r Radioastronomie, Auf dem Hu¨gel 69, 53121 Bonn,}

Germany

...

Our Solar System was formed from a cloud of gas and dust. Most of the dust mass is contained in amorphous silicates1

, yet crystalline silicates are abundant throughout the Solar System, reflecting the thermal and chemical alteration of solids during planet formation. (Even primitive bodies such as comets contain crystalline silicates2_{.) Little is known about the evolution of the}

dust that forms Earth-like planets. Here we report spatially resolved detections and compositional analyses of these building blocks in the innermost two astronomical units of three proto- planetary disks. We find the dust in these regions to be highly crystallized, more so than any other dust observed in young stars until now. In addition, the outer region of one star has equal amounts of pyroxene and olivine, whereas the inner regions are dominated by olivine. The spectral shape of the inner-disk spectra shows surprising similarity with Solar System comets. Radial-mixing models naturally explain this resemblance as well as the gradient in chemical composition. Our observations imply that silicates crystallize before any terrestrial planets are formed, consistent with the composition of meteorites in the Solar System.

Most young stars are surrounded by a disk of gas and dust which is a remnant of the star-formation process. This disk is formed owing to conservation of angular momentum in the collapsing proto-stellar cloud, and channels material from the cloud to the proto-star. When the material in the surrounding molecular cloud is exhausted, the disk dissipates within approximately 107

years (ref. 3). Planet formation is believed to result from the growth of submicrometre-sized interstellar dust particles4

. Therefore, changes in size but also in the chemical nature of the dust grains in the nebular disk environment trace the first steps in planet formation. For instance, crystalline silicates are formed as a result of thermal annealing of amorphous grains, or by vaporization and subsequent gas-phase condensation in the innermost disk regions. These are referred to as primary processes. After inclusion of dust in larger parent bodies such as asteroids and planets, so-called secondary processing occurs, which includes oxidation, aqueous alteration and thermal metamorphism. Asteroids and comets contain pristine

interstellar dust as well as dust which has seen substantial proces- sing5

. The reconstruction of the formation history of our Solar System depends on a better understanding of the nature of primary and secondary processes, and when and where they occurred in the proto-solar nebula.

We observed three Herbig Ae stars with the Mid-Infrared Inter- ferometric Instrument (MIDI)6

installed at the Very Large Telescope Interferometer (VLTI). The light from two 8.2-m Unit Telescopes separated by 103 m on the ground was combined, providing a spatial resolution of about 20 milli-arcseconds. This corresponds to ,1–2 astronomical units (AU) at the distance of the observed stars; an improvement of more than a factor of ten in spatial resolution compared to the largest modern-day telescopes, in this wavelength regime. The MIDI instrument measures spectrally dispersed visibilities with l/Dl¼ 30 in the 7.5–13.5-mm atmos- pheric window. The intensity distribution of circumstellar disks is strongly centrally peaked7,8_{, so the correlated spectra measured by}

the interferometer are dominated by the inner 1–2AUof the disks. We refer to these as the inner-disk spectra. In addition, spectra were obtained with a single 8.2-m telescope, in which the objects are spatially unresolved8_{. We refer to these spectra as the total-disk}

spectra. The difference between the total-and the inner-disk spectra arises mainly from a region between approximately 2 and 20AU. We will refer to these spectra as the outer-disk spectra.

Figure 1 The spectrum of the innermost disk regions of HD 142527 compared to spectra of typical dust species. From top to bottom we plot the observed inner-disk spectrum of HD 142527, the laboratory spectra of crystalline olivine and pyroxene29_{, a laboratory}

spectrum of an IDP consisting of hydrated silicates17_{, and the interstellar medium silicate}

spectrum1_{. The resolution of the laboratory data is reduced to that of the interferometric}

spectrum. The main resonances of crystalline pyroxene at 9.2 mm and crystalline olivine at 11.3 mm are clearly seen in the HD 142527 spectrum. We can exclude the possibility of a significant contribution of hydrated silicates to the spectrum in the inner-disk regions of HD 142527, which suggests that we see primary, rather than secondary dust.

NATURE | VOL 432 | 25 NOVEMBER 2004 | www.nature.com/nature _{© 2004 Nature Publishing Group} 479

Figure 21. Spectre infrarouge des régions internes de du disque associée à l’étoile HD 142527, ainsi que notamment les spectres obtenus en laboratoires pour certaines des silicates, comme l’olivine et le pyroxène cristallin, et le spectre d’un silicate hydraté (tiré de Van Boekel et al. 2004).

Fig. 1.17 – Spectre infrarouge des r´egions internes du disque associ´ee `a l’´etoile HD 142527, ainsi que notamment les spectres obtenus en laboratoire pour des silicates comme l’olivine et le pyrox`ene cristallin et le spectre d’un silicate hydrat´e (tir´e de Van Boekel et al. 2004).

Si une hydratation des poussi`eres peut se produire dans le disque, il devrait donc ˆetre possible de d´etecter ces phases hydrat´ees. Les premi`eres d´etections de silicates dans des environnements astrophysiques ont commenc´e dans les ann´ees 1970, avec la d´etection de bandes larges entre `a 9.8 et 18 µm (Gammon et al. 1972), correspondant aux liaisons Si-O et O-Si-O. L’ubiquit´e des silicates dans le disque et le MIS a ´et´e mise en ´evidence grˆace