Two spinel populations from the Cretaceous-Paleogene (K/T)...

Outline of Spinel Chemistry

[17] Minerals of the spinel group were found in the Gams sequence in the K/T boundary layer of clay and in the overlying Maestrichtian clays. We determined that the rocks contain two spinel populations: Cr-spinel with a low Ni concentration ( < 0.1%) and spinel rich in Ni ( > 2%) (Tables 1 and 2). The former was found throughout the whole vertical section of the Gams boundary layer, whereas the latter is contained not only in all portions of transitional layer, but also below in calcareous marlstones. The lower portion of the boundary layer commonly contains spinel in association with single grains of magnesiochromite and chromite.

Figure 3

[18] The two spinel populations also show different morphologies of their grains. The Cr-spinel typically occurs as angular grains with clearly pronounced serrated edges, a feature suggesting the absence of abrasion (Figure 3a). These grains range from 20 to 100-150 μm in size. The Ni-bearing spinel was found in the form of octahedral crystals up to 10-15 μm (Figure 3 b, c, d).

[19] The Cr-spinel is noted for high and relatively little varying Cr₂O₃ concentrations (from 57% to 79%) and high variations in MgO (3-20%) and FeO (13-40%). Some analyses indicate the presence of TiO₂ (up to 0.6%), MnO (up to 1.6%), ZnO (up to 0.8%), V₂O₅ (up to 0.3%), and NiO (up to 0.1%) (see Table 1).

Figure 4

[20] Spinel compositions are commonly classified on the basis of their Cr/(Cr + Al) and Mg/(Mg + Fe²⁺ ) ratios. As follows from the diagram in Figure 4, spinel from the boundary layer of the Gams sequence has very high and little varying Cr# values (

0.8) and strongly varying Mg#.

[21] It has been demonstrated [Pober and Faupl, 1988] that spinel from sedimentary rocks of the Upper Gosau Group shows broad variations in its Cr# (from 0.15 to 0.85) at insignificantly varying Mg#, and its compositional points plot within the field of spinel from harzburgites and lherzolites. As can be seen in Figure 4, the compositional fields of spinel form the Gamns sequence and Gosau Group are clearly separated in the diagram, which testifies that the erosion of Alpine-type peridotites of an ophiolite association could not provide detrital spinel for the Gams boundary layer.

[22] The origin of Cr-spinel in the clays could have been related with volcanic processes, because the influence of a mantle plume was proved for the boundary layer in the Gams sequence [Grachev et al., 2005, 2007], but our samples are characterized by low TiO ₂ contents ( < 1%). The Cretaceous Tainba Formation in southern Tibet, which has a similar age and composition, contains detrital spinel borrowed from basalts related to a mantle plume and contains up to 4% TiO ₂ [Zhu et al., 2004], and this mineral from Siberian and Deccan flood basalts contains 18-20% TiO ₂ [Melluso et al., 1995; Zolotukhin et al., 1989].

[23] An appealing idea seemed to be searching for possible genetic relations between detrital spinel in the Gams sequence and chromitites in ultrabasic rocks of the Speick Complex (PR₃-PZ₁ ) at a distance of approximately 100 km southeast of the village of Gams [Melcher and Meisel, 2004]. We sampled the chromitites and examined the composition of Cr-spinel from the Kraubath Massif and used the analyses published in [Melcher and Meisel, 2004] for comparative analysis.

[24] The compositional fields of spinel from the Kraubath and Hochgrossen massifs, both of which belong to the Speick Complex, partly overlap the field of spinel from the Gams sequence, but their general Cr# vs. Mg# dependences are exactly opposite. Moreover, spinel from the Speick Complex is higher in TiO ₂ (up to 3%) [Melcher and Meisel, 2004].

Figure 5

[25] As follows from Figures 4 and 5, the closest analogue of the Gams Cr-spinel is spinel from kimberlites and inclusions in diamonds [Barnes and Roeder, 2001; Kamensky et al., 2002; Sobolev, 1974]. In a Cr# vs. Mg# diagram (Figure 4), the composition of Cr-spinel from the Gams sequence overlaps the compositions of spinel inclusions in Yakutian and South African diamonds [Sobolev, 1974; Sobolev et al., 2004].

Figure 6

Figure 7

Figure 8

Figure 9

[26] The Cr₂O₃ vs. MgO diagram in Figure 6 accentuates this similarity revealing a common "peridotite'' trend with a strong correlation between Cr and Al typical of Cr-spinel in kimberlites [Barnes and Roeder, 2001; Kamensky et al., 2002]. The analysis of other diagrams corroborates this conclusion (Figures 7, 8, and 9).

[27] Judging from the data presented above, the detrital Cr-spinel could originate from metamorphic rocks of eclogite composition, which are widespread in Eastern Alps [Sassi et al., 2004; and others]. Single spinel grains were recently found in eclogites from Pohorje [Janak et al., 2006]. Diamonds were found in such rocks in many complexes throughout the world (in Kazakhstan, China, Norway, Greece, and Germany) [Mposkos and Kostopoulos, 2001; Schulze et al., 1996; Sobolev and Shatsky, 1990; Xu et al., 1992; and others].

[28] Ni-bearing spinel was fist found in marine deposits at the K/T boundary in Furlo and Petriccio deposits in Italy and in Hole 577 in the Pacific Ocean [Montanari et al., 1983; Smit and Kyte, 1984] and then elsewhere, in rocks of various ages, from Archean rocks to modern cosmic dust [Ben Abdelkader et al., 1997; Bohor et al., 1986; Byerly and Lowe, 1994; Kyte and Bohor, 1995; Robin et al., 1992, 1991; and others]. To distinguish this spinel from spinel from magmatic rocks, the former was soon referred to as cosmic spinel [Robin et al., 1992]. The typical composition of this spinel from various deposits is reported in Table 3.

[29] Since cosmic spinel was first found in rocks containing impact quartz and high Ir concentrations, the genesis of this mineral was thought to be related to the impact process [Smit and Kyte, 1984; and others]. However, finds of Ni spinel in rocks showing no traces of impact events, for example, in cosmic dust and micrometeorites, gave rise to an alternative viewpoint, whose proponents believed that the origin of Ni spinel was related to the ablation of a cosmic body when it entered the atmosphere [Robin et al., 1992; Toppani and Libourel, 2003].

[30] The Gams stratigraphic sequence is principally important in the context of the genesis of Ni spinel, because spinel from these rocks contains more than 2% NiO and was detected not only in the lower part (so called rust layer) but also in the upper and middle vertical portions of the boundary layer.

[31] The chemical composition of Ni spinel from the Gams sequence is close to the composition of this mineral from elsewhere (compare Table 2 and Table 3). The NiO concentration varies from 2% to 8%, FeO from 54% to 77%, Al₂O₃ from 1% to 10%, and Cr₂O₃ from 5% to 13%. Note that the mineral pervasively contains ZnO (up to 5%).

[32] The Cr# vs. Mg# diagram in Figure 4 shows the overlap of the compositional fields of Ni spinel from the Gams sequence and other regions, which is also illustrated in other classification diagrams (Figures 5, 6, 7, 8, and 9). Cosmic spinel differs from terrestrial spinel in having high concentrations of Ni and Fe²⁺ and displays significant compositional variations.

[33] Another no less important identification feature is the significant difference between the Fe³⁺/(Fe²⁺ + Fe³⁺ ) ratios (which were calculated from stoichiometric considerations, see Table 3) of magmatic and cosmic spinel. This ratio is known to be utilized as an indicator of oxygen fugacity and is important for identifying the conditions under which the melt crystallized [Toppani and Libourel, 2003]. As follows from Tables 1 and 3, these differences are very significant for spinel from the Gams sequence: the Ni spinel has high ratios (0.62-0.98), whereas this parameter of the Cr-spinel ranges from 0.06 and 0.48. These data are consistent with data on the composition of cosmic spinel from the fusion crusts of meteorites, in which the Fe³⁺/(Fe²⁺ + Fe³⁺ ) ratio varies from 0.75 to 0.90 [Robin et al., 1992].