Impact melting, metal-silicate fractionation, and volatile element mobility on the L chondrite parent body. Workshop on Parent Body and Nebular Modification of Chondritic Materials

Norman M.D., GEMOC, Macquarie

"I think we'd all like to know what's in that rock" -Agent Mulder

The L chondrite parent body suffered a massive, possibly catastrophic impact event about 500 million years ago [1,2,3]. We are investigating the petrologic and geochemical consequences of this event through studies of impact melt and host chondrite material preserved in Chico, a 105 g L chondrite. Chico is composed of ~60% impact melt that intrudes the host chondrite as a dike, and may represent a complex of dikes injected into the wall or floor of the crater during the 500 Myr event [2].

The host chondrite is shocked to stage S6, and contains pockets of silicate melt and veins of shock-mobilized metal. Chondrules are highly deformed, and are fractured and terminated in sharp contact with the dike. The melt rock is clast-poor, and consists primarily of fine-grained, randomly oriented, euhedral olivines and pyroxenes in a matrix of interstitial glass. Metallic globules up to 2 cm in diameter are concentrated along the center of the dike. Their orientation is suggestive of flow differentiation and implies an early stage of metal-silicate immiscibility. Melt textures are present within these globules, which contain Fe-Ni metal, troilite, and schreibersite [2]. Very small (<1-5 micron diameter) pinpoints of metal commonly adhere to the surfaces of olivine grains, demonstrating imperfect segregation of metal from silicate during the melting.

Variable K/Ca and depletions of Cs, Cd, Bi, Tl, and In in the Chico impact melt relative to the host chondrite have lead to suggestions that volatile elements may also have been mobilized during the impact event [2,4].

Chico thus provides a natural laboratory in which to study element partitioning and fractionation caused by metal-silicate segregation, and volatile element mobility during a large impact event on a chondritic parent body. Depletion of volatile elements and fractionation of metal from silicate are characteristic features of the terrestrial planets and presumably many asteroids (e.g., the eucrite parent body), so, conceivably, similar processes on larger scales may have influenced the differentiation of planetesimals during the early history of the solar system.

To investigate the consequences of this impact event on the L chondrite parent body, we have conducted an INAA study of Chico using samples from the same cores that were analyzed by [2].

REE patterns of the impact melt and the host chondrite are all consistent with an undifferentiated parent body and no large scale crystal-liquid fractionation during the impact melting event (Fig. 1). Absolute concentrations of the REE in splits of the melt rock vary by about 30%, and are negatively correlated with siderophile elements such as Fe, Ni, and Ir (Fig. 2). REE contents of the host chondrite splits also range to relatively high values, and do not vary with the amount of metal.


Siderophile elements are strongly affected by the metal-silicate fractionation within the dike. Fe, Ni, Co, Ir, Au, As, and Se are progressively depleted in the melt relative to the host chondrite, and trend toward the composition of a metal fragment (M) separated from the dike (Fig. 2,3,4). Metal fractionation also drives the melt toward low Ni/Co and Ir/Au ratios (Fig. 5). Although Se is depleted in the melt, it was not detected in the metal globule and may have been lost due to weathering of sulfide.


The anti-correlation of REE and siderophile elements must reflect variable proportions of metal and silicate melt, but the melt arrays are too steep to be produced simply by removing metal from the chondrite host (Fig. 2). The dike appears to be overly enriched in incompatible lithophile elements, and the degree of this enrichment correlates with the degree of siderophile depletion. This may reflect migration

of residual silicate melt within the dike, or addition of a metal-poor silicate melt component from the host chondrite. LREE variations in the host chondrite may also reflect variable amounts of a melt component.

K and other volatile elements are depleted in the central portion of the dike [2,4], resulting in high Ca/K in some samples (Fig. 6) Most of the melt, however, appears to be enriched in K and have Ca/K lower than the host chondrite, implying possible volatile enrichment, although K may also have been added by the melt component responsible for enrichment of the REE.

Several samples of the dike have Na contents and Sm/Na ratios that cluster around the host chondrite values, while other samples range to both higher and lower values (Fig. 7). Melting or crystallization of the melt should not fractionate these elements (as long as plagioclase is not involved), so the variations in alkali element content and Sm/Na in the dike may be indicating redistribution of alkalies. Na contents lower than those of the host chondrite almost certainly reflect volatile loss as this is difficult to achieve either by melting or by crystallization.

Within the dike, there is a general correspondence between the degree of siderophile element depletion, REE enrichment, and alkali element enrichment, suggesting that the processes of metal extraction, melt migration, and volatile element mobility were physically linked during this impact event in the L chondrite parent body.

References [1] Haack et al. (1996) Icarus 119, 182-191 [2] Bogard et al. (1995) GCA 59, 1383-1399 [3]Nakamura et al. (1990) Nature 345, 51-53

[4] Yolcubal et al. (1997) JGR Planets, submitted

Back to the GEMOC Abstract Titles Page