Transformations in silicate phases define the major petrographic divisions of chondrites [1,2]. The corresponding transformations of opaque phases throughout the metamorphic sequence have been comparatively little studied [e.g. 3] and Perron et al. [4] showed that some textural changes took place from type 3 to type 6. We here attempt to systematically describe the changes undergone by opaque minerals, and establish a metamorphic scale parallel to that based on silicates.

We hope to derive:

(i) a better understanding of the state of opaque minerals at peak temperatures in order to better constrain cooling rate estimations,

(ii) an adequate way of comparing experimental metamorphic analogs with metamorphosed chondrites (since measured/estimated temperatures are insufficient due to the timescale differences involved).

We have chosen to limit ourselves here to the main textural transformations, though behavior of Ni, Co, and trace elements [5] is also relevant.

METHOD:

We have studied the textural appearance and relationship of metal and sulfides in a series of sections of ordinary chondrites of various petrographic types. Because of the lack of H chondrites of the lowest petrographic types and of the superimposition of brecciation effects in most LLs of the highest ones, we have chosen to break our series into two overlapping parts: metamorphic effects in type 3 chondrites have been studied in L/LL chondrites of types ranging from 3.0 to 4, whereas the effects from 3.4 to 6 have been studied in H chondrites.

RESULTS:

Our results for the two metamorphic series are displayed in Fig. 1 and 2 respectively. The detailed study of opaque mineral distribution in 3.0 Semarkona indicates that, except for a few selected locations, metal and sulfide phases are constantly and intimately associated, a likely result of nebular corrosion effects of H₂S gas on kamacite partly before but mostly after chondrule formation [6]. Apart from the fine-grained matrix and a few tetrataenite grains also found in the matrix, metal grains with no associated sulfides are only found inside of type I (FeO-poor) chondrules. On the other hand, sulfide grains with no metal are restricted to the interior of type II (FeO-rich) chondrules and a few isolated sulfide grains in the matrix. Figure 1a shows that the metal/sulfide contact is very highly contorted in such a material. This spatial relationship of metal and sulfide is gradually changed by metamorphism.

We identify three main steps:

I) up to 3.5 (Fig. 1b-d), metal and sulfides stay closely associated but their contact progressively changes from contorted to straight. Metal grains have a round shape both inside and outside chondrules. Sulfide appears in increasing quantities associated with the metal beads inside chondrules.

II) from 3.5 to 4 (Fig. 2a-b) metal and sulfide start separating from one another and sulfide grains tend to connect together (this effect becomes more obvious by 3.7). Opaque grains inside of chondrules lose their round shape to become more angular. The first zoned taenite grains appear and metal grains start coalescing [4,7] by 3.8 (Fig. 2a).

III) From 4 to 5 (Fig. 2c-d), the separation of metal and sulfide progresses almost to completion and adjacent metal grains between chondrules gradually coalesce. All the silicate material intially enclosed by the coalescing grains eventually gets expelled and the grains acquire a uniform clean, clear cut aspect. No round shape survives and the shape of the opaque minerals seems to be governed by the interstices they fill between the silicates.

With the disappearance of chondrules by type 6, the grains become more evenly dispersed and smaller than the large metal grains of types 4 and 5.

CONCLUSIONS:

Opaque phases in chondrites undergo changes with metamorphism that can be followed step by step, allowing the optical determination of a petrographic type matching that deduced from the silicate phases with a reasonable precision. In the first approximation, there is no observable difference between the opaque mineral textures observed in H and LL chondrites, which indicates that our method could potentially also be applied to metal-bearing carbonaceous chondrites.