Workshop on Parent-Body and Nebular Modification of Chondritic Materials
ALTERATION OF CAIs: TIMES AND PLACES.
S. S. Russell and G. J. MacPherson, Department of Mineral Sciences, MRC NHB-119, U.S. Museum of Natural History, Smithsonian Institution, Washington DC 20560, USA.
E-mail: srussell@volcano.si.edu
Calcium- Aluminium- rich inclusions (CAIs) commonly contain a distinctive suite of secondary minerals. The chemical and isotopic compositions of these minerals can be used to constrain the site and timing of the alteration event. The style of alteration in CAIs is strongly dependent on the meteorite group in which they are found.
CV meteorites: CAIs from the oxidised subgroup (e.g. Allende) show extensive signs of secondary alkali- and iron- enrichment. The fine grained secondary minerals (typically <10-20>monticellite, hedenbergite, andradite, and grossular; these typically embay primary minerals and fill cross- cutting veins within the CAIs. Some euhedral whiskers of wollastonite and nepheline are located within cavities. In addition, multi-layered Wark-Lovering rim sequences on CAIs clearly postdate the CAI interior, and in that sense can be considered secondary. Fine-grained inclusions are typically more altered than coarser grained ones: alteration in these inclusions consists of feldspathoid layers surrounding primary spinel. The temperature of melilite + anorthite breakdown to grossular + monticellite in type B Allende CAIs has been estimated to be 668 ºC [1]. Hutcheon and Newton argued that the temperature must have remained around this value for a “prolonged period” to allow formation of large grossular grains, but the timing of the high temperature alteration event was probably minor to 100,000 years, otherwise Mg diffusion would ensure the CAI anorthite no longer retains a Mg-26 excess [2].
The location of the alkali-iron alteration has been widely debated. Most workers believe the alteration took place in a nebular setting. The sequence of alteration is compatible with equilibration with a cooling, oxidised solar nebula gas [3]. Wark [4] argued for pre-accretionary alteration because of the presence of alkali-rich halos in the meteorite matrix surrounding some CAIs. Sodium mapping of Allende CAIs shows that the sodium is enriched in accretionary rims, suggesting CAIs became alkali- rich prior to incorporation in the parent body. Veins cross-cutting CAIs typically do not extend into the meteorite matrix, indicating they did not form in situ. Euhedral wollastonite whiskers, nepheline needles, and grossular in CAI cavities are indicative of condensation from a vapour, and these grains probably formed in the nebula. An alternative viewpoint, championed by Krot et al., argues that the alteration of CV CAIs can be explained by a parent body process of alteration by alkaline-rich fluids followed by dehydration [5]. This process is postulated to have affected the more oxidised CV meteorites, such as Allende, more than the other CVs, a conclusion also reached by some other studies [e.g. 6].
CO meteorites: CAIs in CO3 chondrites have experienced considerable secondary alteration, both before and after accretion [7, 8]. The presence of altered CAIs in unmetamorphosed CO3s indicates some events occurred in the nebula: formation of Wark-Lovering rims, melilite and anorthite breakdown, and iron enrichment of spinels in hibonite-rich inclusions. In contrast, correlations between petrologic type of the host meteorite with iron content and melilite breakdown in Type A and spinel-pyroxene CAIs suggest some alteration occurred during parent body metamorphism [8]. Hibonite seems to be unaffected by the metamorphism experienced by CO3s.
CM meteorites: CAIs in CM chondrites have suffered ubiquitous aqueous alteration. The CAI primary mineralogy has been altered to pyllosilicates (Fe- and Mg-serpentines) and tochilinite, calcite and calcium sulphate. Secondary minerals typically occur in a layer immediately beneath the rim sequence. Some CM CAIs have also suffered fragmentation and recrystallisation. It is not clear what the phyllosilicate is replacing: anorthite is a possibility. Greenwood et al. [9] suggest that nebula processes caused fragmentation of CAIs, whereas aqueous alteration took place over a protracted period of time on the parent body. In contrast, MacPherson and Davis [10] argued that the CAIs were not altered in the environment in which they are now found, and many are too fragile to have been moved to their current location by recycling in the regolith, so they favoured formation of hydrous secondary minerals in a nebula environment.
CR meteorites: CAIs in the CR chondrite Acfer 059 shows no evidence of alteration [11], whereas inclusions in Renazzo and Al Rais contain some secondary calcite [12].
CH meteorites: Some CAIs in the CH chondrite ALH 85085 show evidence of recrystallisation due to reheating [13]. In contrast, inclusions from PCA 91467 and Acfer 182 appear unaltered [14].
Unequilibrated ordinary chondrites (UOC): CAIs in ordinary chondrites are rare. UOC CAIs are often rimmed, and secondary feldspathoids are occasionally present. In one Semarkona CAI, melilite has been partially replaced by sodalite [15].
CONCLUSIONS:
Times: While most CAIs are believed to have formed at around the same time, their alteration was an on-going process that took place over several million years. I-Xe dating suggests that the alteration took place up to >10Myr after initial CAI formation [16]. Al-Mg studies of grossular in CV CAIs also indicate formation > 2.4 Myr after CAI production [17], and a Al-Mg analysis of a recrystallised CAI from CH chondrite indicates a heating event > 2 Myr after CAI production [13]. Chemical exchange between anorthite and melilite in Type B inclusions appears to have occurred >2-3 Myr after CAI formation [18]. Similarly, sodalite in fine grained CV inclusions is postulated to have formed after Al-26 decay, ie., several Myr after CAI formation [19]. In contrast, sodalite in a Semarkona (LL3.0) inclusion apparently formed very quickly after CAI formation [15].
Places: Many CAIs are pristine, but some underwent several heating events. The presence of altered plus pristine CAIs close together in some meteorites (e.g. CMs) suggests that some alteration occurred before they reached their current site in the parent body (although this may reflect post- accretionary brecciation). Many features of alteration appear to have occurred in the nebula. Wark-Lovering rims predate accretion into the present parent bodies. Some primary minerals exchanged with a nebula gas, and some secondary minerals condensed from a vapour. Sodalite in Semarkona [15] probably formed in the nebula, since Al-26 dating suggests it formed before the accretion of the asteroids. In contrast, the long time span of alteration suggested by I-Xe dating for Allende CAIs has been used to argue in favour of alteration in a parent body [16]. In addition to nebula processes, metamorphism in parent bodies tended to equilibrate CAIs with their host rock. Aqueous processing in some meteorites may have affected CAIs in parent bodies. The location of the event responsible for incorporation of alkalis into CAIs, however, is still highly contentious.
References:
[1] Hutcheon and Newton (1981), LPSC XII 491-493;
[2] LaTourette and Wasserburg (1997) LPSC 28 781-782;
[3] Hashimoto and Grossman (1987) GCA 51 1685-1704;
[4] Wark (1981) LPSC XII 1145-1147;
[5] Krot et al., (1995) Meteoritics 30 748-775;
[6] McSween (1977) GCA 41 1777-1790;
[7] Greenwood et al., (1992) Meteoritics 27 229;
[8] Russell et al., (1997) GCA, submitted;
[9] Greenwood et al., (1994) GCA 58 1913-1935;
[10] MacPherson and Davis (1994) GCA 58 5599-5625;
[11] Weber and Bischoff (1997) Chem. Erde 57 1-24;
[12] Weisburg et al., (1993) GCA 57, 1567-1586;
[13] Kimura et al., (1993) GCA 57 2329-2360;
[14] Weber et al., (1995) LPSC XXVI 1475-1476;
[15] Russell et al., (1997) LPSC XXVIII, 1209-1210;
[16] Swindle et al., (1998) GCA 52 2215-2227;
