11:00 〜 13:00
[SGC36-P02] 3D analysis of melt inclusions containing platinum group elements in Tahitian harzburgite xenolith with X-ray nanotomography
キーワード：放射光ナノXCT、白金族元素、メルト包有物、Fe-Ni-Cu硫化物、白金族鉱物
Platinum group elements (PGEs: Ru, Rh, Pd, Os, Ir, Pt) are strongly partitioned into metallic phases, and thus are believed to have strongly partitioned into the core during core-mantle segregation. However, the PGE abundances of the primitive upper mantle (PUM: a hypothetical upper mantle reservoir) are abnormally high and their relative ratios don't coincide with those inferred from experimental data. To elucidate the cause driving the discrepancy, one of the diagnoses is detailed investigation of platinum group minerals (PGMs) and base metal sulfides (BMSs), which are accessary mineral phases in the mantle peridotites, but are major PGE-host minerals. PGMs and BMSs are identified as granular form or interstitial form in the mantle peridotites. Compared to the interstitial form, granular form if armored by silicate mineral (inclusion) has significant importance as they are protected from later processes such as weathering and alteration (Akizawa et al., 2020). Since the inclusions of PGM and BMS are usually very small (down to the sub-micrometer scale), sub-micrometer-scale intense investigations are necessarily required to reveal the PGE behavior in the Earth's mantle. Akizawa et al (2017, 2020) observed 3D distribution of a planer inclusion array hosted by clinopyroxene (Cpx) in H3-001_TK mantle peridotite from Tahiti Island with synchrotron radiation-based X-ray computed nanotomography (SR-nanoXCT). Akizawa et al (2017, 2020) analyzed one of the inclusions with transmission electron microscope (TEM). The STEM-EDS analysis revealed that Ir, Pt, and Rh are hosted in a PGM phase, not in BMS. Irrespective of such a series of intensive investigations, 3D distribution of the PGEs in the inclusion array is unclear because PGM and BMS were not distinguishable in their CT-data. To understand the PGE behavior in the Earth's interior, 3D distribution of the PGEs in the inclusion array has been desired.
In this study, we combined two SR-nanoXCT techniques (dual energy tomography (DET) (Tsuchiyama et al., 2013) and scanning-imaging X-ray microscopy (SIXM) (Takeuchi et al., 2014)) to discriminate PGM and BMS in the 3D visualization of the same sample used in Akizawa et al (2017). We conducted DET using X-ray energy of 7 keV and 7.35 keV, and SIXM using X-ray energy of 8 keV at BL47XU of SPring-8, a synchrotron radiation facility in Japan. The voxel sizes in DET were 47.17 nm and 49.44 nm at 7 and 7.35 keV respectively, which are significantly smaller than that of Akizawa et al (2017), and that of SIXM was 111.06 nm to x and y directions and 108.1 nm to z direction.
In total, 35 inclusions were investigated herein. The inclusions ranging in volume from 0.01–19.45 µm3 include PGM, BMS, silicate glass, and vapor phase, which were identified based on the two-dimensional plot of linear attenuation coefficient (LAC) and refractive index decrement (RID) values. The investigated inclusion array consists of 0.91 vol% PGM, 16 vol% BMS, 72 vol% silicate glass, and 11 vol% vapor phase. Considering the constituent-phase of each inclusion, the inclusions were classified into 4 types: (1) vapor phase dominant (V, 5 inclusions), (2) sulfide (= BMS + PGM) + vapor phase (S+V, 8 inclusions), (3) silicate glass + vapor phase (G+V, 14 inclusions), and (4) sulfide + silicate glass + vapor phase (S+G+V, 8 inclusions). Type V inclusions contain more than 40 vol% vapor phase, while inclusions of the other types contain less than 15 vol% vapor phase. In type S+V inclusions, more than 90 vol% of the solid phases is sulfide, while type G+V inclusions contain more than 90 vol% glass as the solid phase. Type S+G+V inclusions contain 20–74 vol% sulfide as the solid phase. Type S+V and S+G+V inclusions have higher fractions of vapor phase (9.0 vol% on average of the biggest 7 inclusions) than type G+V inclusions (4.3 vol% on average of 7 bigger inclusions out of total 14 inclusions). PGM is included in all the sulfides, and the volume ratio of PGM/sulfide is constant particularly among the big sulfides (4.4 ± 1.1 vol% from the biggest 8 sulfide inclusions).
Our new data suggest that not only the inclusion investigated with TEM by Akizawa et al (2017, 2020), but all sulfides in the inclusion array contain PGEs as PGM phase. Since the sulfide-bearing inclusions are more abundant in vapor phase compared to the other sulfide-free inclusions, the sulfide melt that contained PGEs probably exsolved more vapor phase than the silicate melt after being trapped by the host Cpx. The 3D analysis of SR-nanoXCT employed herein for the detailed investigation of sub-micrometer-sized inclusion containing PGEs is a powerful tool to elucidate PGE behavior in the Earth's interior.
In this study, we combined two SR-nanoXCT techniques (dual energy tomography (DET) (Tsuchiyama et al., 2013) and scanning-imaging X-ray microscopy (SIXM) (Takeuchi et al., 2014)) to discriminate PGM and BMS in the 3D visualization of the same sample used in Akizawa et al (2017). We conducted DET using X-ray energy of 7 keV and 7.35 keV, and SIXM using X-ray energy of 8 keV at BL47XU of SPring-8, a synchrotron radiation facility in Japan. The voxel sizes in DET were 47.17 nm and 49.44 nm at 7 and 7.35 keV respectively, which are significantly smaller than that of Akizawa et al (2017), and that of SIXM was 111.06 nm to x and y directions and 108.1 nm to z direction.
In total, 35 inclusions were investigated herein. The inclusions ranging in volume from 0.01–19.45 µm3 include PGM, BMS, silicate glass, and vapor phase, which were identified based on the two-dimensional plot of linear attenuation coefficient (LAC) and refractive index decrement (RID) values. The investigated inclusion array consists of 0.91 vol% PGM, 16 vol% BMS, 72 vol% silicate glass, and 11 vol% vapor phase. Considering the constituent-phase of each inclusion, the inclusions were classified into 4 types: (1) vapor phase dominant (V, 5 inclusions), (2) sulfide (= BMS + PGM) + vapor phase (S+V, 8 inclusions), (3) silicate glass + vapor phase (G+V, 14 inclusions), and (4) sulfide + silicate glass + vapor phase (S+G+V, 8 inclusions). Type V inclusions contain more than 40 vol% vapor phase, while inclusions of the other types contain less than 15 vol% vapor phase. In type S+V inclusions, more than 90 vol% of the solid phases is sulfide, while type G+V inclusions contain more than 90 vol% glass as the solid phase. Type S+G+V inclusions contain 20–74 vol% sulfide as the solid phase. Type S+V and S+G+V inclusions have higher fractions of vapor phase (9.0 vol% on average of the biggest 7 inclusions) than type G+V inclusions (4.3 vol% on average of 7 bigger inclusions out of total 14 inclusions). PGM is included in all the sulfides, and the volume ratio of PGM/sulfide is constant particularly among the big sulfides (4.4 ± 1.1 vol% from the biggest 8 sulfide inclusions).
Our new data suggest that not only the inclusion investigated with TEM by Akizawa et al (2017, 2020), but all sulfides in the inclusion array contain PGEs as PGM phase. Since the sulfide-bearing inclusions are more abundant in vapor phase compared to the other sulfide-free inclusions, the sulfide melt that contained PGEs probably exsolved more vapor phase than the silicate melt after being trapped by the host Cpx. The 3D analysis of SR-nanoXCT employed herein for the detailed investigation of sub-micrometer-sized inclusion containing PGEs is a powerful tool to elucidate PGE behavior in the Earth's interior.