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

[6] Samples from the boundary layer were first collected from a monolith cut out of an exposure and provided for us by courtesy of the administration of the Museum of Natural History in Vienna. Additional samples were then collected at exposures to ensure the reproducibility of the results.

Figure 2

[7] The monolith had a height of 46 cm, a width of 30 cm in its bottom portion and 22 cm at the top, and a thickness of 4 cm (Figure 2). For its comprehensive examination, the monolith was divided (vertically, from bottom to top) into layers 2 cm thick (which were labeled A, B, C, ldots

, W), and each of them was then subdivided into fragments by lines spaced 2 cm apart (the fragments were labeled 1, 2, 3, etc.). Layer J (2 cm thick) at the K/T boundary was examined more thoroughly and subdivided into six units approximately 3 mm each [see Grachev et al., 2005 for details].

[8] In exposure Gams 2, where the thickness of the boundary layer reaches 5 cm, samples were collected in the form of continuous blocks, which were then cut into individual slices, with cutting surfaces spaced approximately 1 cm apart.

[9] Samples 10-15 g in mass were crushed in a porcelain mortar and sieved through a 0.25-mm screen. Heavy-fraction minerals were separated from the carbonate-clay mass in heavy liquids (bromoform, density 2.89 g cm^-3 ). The heavy and light fractions were washed in alcohol and purified of the magnetic fraction (magnetite) with the use of a simple magnet. The heavy-fraction minerals were separated on a magnetic separator according to their electric conductivity into a non-electromagnetic, weakly electromagnetic, and electromagnetic fractions.

[10] Other minerals were hand-picked under a binocular magnifier, using a needle and were glued to glass platelets with pits. The glue was Sherlok, which can be dissolved in alcohol. Quartz was obtained from the light fraction ( < 2.89 g cm ^-3 ) by electromagnetic separation. In order to analyze the composition and microstructures of ferromagnetic minerals in the rocks, the samples were examined on a Camebax microprobe equipped with three wave-dispersive spectrometers and then on a Tescan Vega I and Tescan Vega II microprobes equipped with energy- and wave-dispersive spectrometers. Samples were usually mounted onto a pellet 26 mm in diameter using Wood's metal, carefully polished and finished with diamond pastes, and then sputter-coated with carbon. The analyses were conducted at an accelerating voltage of 20 kV and a beam current of 10 nA. The elements determined using their characteristic X-ray radiation ranged from Na to U. The effective diameter of the beam was thereby close to 1-2 μm and was systematically tested at small grains. We carried out quantitative analysis of ore minerals for TiO₂, FeO, MgO, MnO, Cr₂O₃, and Al₂O₃; preliminary analysis was carried out for all elements.

[11] Selected magnetic mineral grains were prepared in compliance with a specially designed technique that ensured the ultrafine polishing of the material with the removal of its layer no thicker than 5 μm during the whole polishing process. The removed 5- to 10- μm-thick layer made it possible to expose pores in iron spherules 10-20 μm. Polishing sometimes exposed hollow metallic cosmic spherules with walls 0.5-2 μm thick. To achieve this, grains separated with the use of a powerful hand-held Nb-B-Fe magnet were placed onto a flat (polished) surface and covered with epoxy resin. The plastic was selected in such a way to ensure the minimum adhesion of epoxy resin to it. This allowed us to easily remove the sample with the examined particles after the solidification of the epoxy resin. Preparatorily to its use, the epoxy resin was degassed in vacuum to minimize the amount of bubbles in it. The sample was polished by ASM 2/1, 1/0, and 0.5/0 diamond pastes and suspension of ultrafinely dispersed diamonds on Montasupal polishing machines (on broadcloth and felt).

[12] The minerals separated in heavy liquids were mounted on a double-layer conducting carbon adhesive tape. Each specimen comprised a few hundred mineral grains ranging from fractions of a micrometer to hundreds of micrometers in size. The grains were first examined on a Camebax microprobe equipped with an optical microscope and three wave-dispersive spectrometers to rapidly locate luminescent grains (of, for example, diamond and moissanite). It was, however, hard to conduct qualitative analysis of each grain because of the very long required instrument time. Because of this, as soon as a new Tescan Vega II microprobe became available for us, we re-examined all the specimens on it. All of the identified grains were analyzed for all possible elements, from Be to Pu. The high sensitivity of the new detector made it possible to significantly decrease the electron-beam current: the analyses were carried out mostly at a current of 200 pA and 20 kV accelerating voltage. In complicated situations, the energy- and wave-dispersive spectrometers were utilized simultaneously to improve the reliability of the measurements. In the course of the re-examination of the specimens, we identified new minerals, which were often submicrometer-sized and had not been detected on the Camebax microprobe.

[13] Magnetic minerals were separated from the crushed samples by a powerful hand-held Nd-Fe-B magnet. In order to separate magnetic particles from the fine clay fraction, we applied ultrasonic pulverization in water. The operating regimes of the generator were selected in such a manner that small (no larger than a few fractions of a micrometer) magnetic particles were reliably recovered but not destroyed in the course of ultrasonic cleaning.