11:15 AM - 11:30 AM
[SIT14-08] Evidence of the heavy noble gas incorporation into KAlSi3O8 hollandite under lower mantle conditions
Keywords:missing xenon, lower mantle minerals, noble gases, synchrotron X-ray, mass spectrometry
Noble gases are important geochemical tracers for exploring Earth's global volatile cycles and the evolution of Earth's atmosphere. In the atmospheres of Earth and Mars, xenon is more depleted than other noble gas elements relative to chondritic meteorites[1]. The "missing Xe" is believed to be captured somewhere inside the Earth, and various materials under crustal/mantle/core conditions have been examined experimentally and computationally, resulting in no obvious evidence for the Xe reservoir so far. Although the major lower mantle minerals (ferropericlase, bridgmanite) are unlikely to account for the missing noble gases, the solubilities of heavy noble gases (Ar, Kr, and Xe) are expected to increase with pressure/depth based on recent studies of lattice strain modeling[2-4].
This study focuses on KAlSi3O8, which forms one of the most abundant minerals in the continental crust, K-rich feldspar. KAlSi3O8 is stable below ~25 GPa as the Hollandite-I phase (Holl-I; liebermannite; tetragonal) and up to 130 GPa and >3000 K as Hollandite-II (Holl-II; monoclinic)[5]. These crystal structures have large square tunnels formed by four double chains of edge-shared octahedra, accommodating large cations of K+ (with a similar size to Ar). Holl-II KAlSi3O8 is also thought to be a potential host mineral for transporting K and incompatible lithophile elements (LILEs) with large ionic radii into the lower mantle through subduction. We formerly investigated the incorporation of Ar and Xe into KAlSi3O8 using laser-heated diamond anvil cells (DAC) to explore whether noble gas storage in Holl-II under lower mantle conditions could account for Earth's missing Xe. The results suggest that Holl-I and II phases showed no obvious changes in the cell volume and Raman shifts derived from the incorporation of noble gases into the tunnel structure. However, K depletion was observed due to heating, suggesting the more K+ sites for the noble gas uptake. This study challenged the quantification of the noble gas solubility of the recovered samples from the high PT experiments using noble gas mass spectrometry and clarifying the chemical bonding state of Xe trapped in the structure via using X-ray absorption fine structure (XAFS) spectroscopy.
The micro-XAFS measurement was performed at beamline BL15A, PF, KEK. The Xe L3-edge energy (4786 eV) was applied. The bulk Xe gas was also measured for standard. From the obtained XRF map, Xe was mainly distributed around the heated area of the sample. The XANES spectra showed that the edge position slightly shifted to the lower energy (~0.25 eV) than the bulk Xe gas. This is suggested that Xe trapped in the structure would be neutral or negative charge (anionic).
Noble gas extraction and analysis were performed using a mass spectrometry system at the University of Tokyo. Recovered samples were picked out from the hole of the Re gaskets and wrapped with aluminum foil. The samples were heated in a tantalum resistance furnace to completely melt up to 1500 ºC and extract the remaining gas in the samples. Approximately 0.5 ppm of 132Xe was detected from the recovered sample with Xe as the pressure medium. In contrast, 40Ar from the sample with Ar as the pressure medium was comparable to the background level. Thus, it is clear that Ar is not readily incorporated into hollandite under high-PT, while Xe was certainly incorporated into the crystal structure. KAlSi3O8 hollandite would accommodate Xe by replacing the K+ sites in the center of the tunnel structure, although the charge balance needs to be considered from further experiments.
References
[1] Marty, EPSL 313–314, 56–66 (2012)
[2] Shcheka and Keppler, Nature 490, 531–534 (2012)
[3] Rosa et al., EPSL 532, 116032 (2020)
[4] Zhu et al., Earth Sci. Rev. 224, 103872 (2022)
[5] Hirao et al., PEPI 166, 97–104 (2008)
This study focuses on KAlSi3O8, which forms one of the most abundant minerals in the continental crust, K-rich feldspar. KAlSi3O8 is stable below ~25 GPa as the Hollandite-I phase (Holl-I; liebermannite; tetragonal) and up to 130 GPa and >3000 K as Hollandite-II (Holl-II; monoclinic)[5]. These crystal structures have large square tunnels formed by four double chains of edge-shared octahedra, accommodating large cations of K+ (with a similar size to Ar). Holl-II KAlSi3O8 is also thought to be a potential host mineral for transporting K and incompatible lithophile elements (LILEs) with large ionic radii into the lower mantle through subduction. We formerly investigated the incorporation of Ar and Xe into KAlSi3O8 using laser-heated diamond anvil cells (DAC) to explore whether noble gas storage in Holl-II under lower mantle conditions could account for Earth's missing Xe. The results suggest that Holl-I and II phases showed no obvious changes in the cell volume and Raman shifts derived from the incorporation of noble gases into the tunnel structure. However, K depletion was observed due to heating, suggesting the more K+ sites for the noble gas uptake. This study challenged the quantification of the noble gas solubility of the recovered samples from the high PT experiments using noble gas mass spectrometry and clarifying the chemical bonding state of Xe trapped in the structure via using X-ray absorption fine structure (XAFS) spectroscopy.
The micro-XAFS measurement was performed at beamline BL15A, PF, KEK. The Xe L3-edge energy (4786 eV) was applied. The bulk Xe gas was also measured for standard. From the obtained XRF map, Xe was mainly distributed around the heated area of the sample. The XANES spectra showed that the edge position slightly shifted to the lower energy (~0.25 eV) than the bulk Xe gas. This is suggested that Xe trapped in the structure would be neutral or negative charge (anionic).
Noble gas extraction and analysis were performed using a mass spectrometry system at the University of Tokyo. Recovered samples were picked out from the hole of the Re gaskets and wrapped with aluminum foil. The samples were heated in a tantalum resistance furnace to completely melt up to 1500 ºC and extract the remaining gas in the samples. Approximately 0.5 ppm of 132Xe was detected from the recovered sample with Xe as the pressure medium. In contrast, 40Ar from the sample with Ar as the pressure medium was comparable to the background level. Thus, it is clear that Ar is not readily incorporated into hollandite under high-PT, while Xe was certainly incorporated into the crystal structure. KAlSi3O8 hollandite would accommodate Xe by replacing the K+ sites in the center of the tunnel structure, although the charge balance needs to be considered from further experiments.
References
[1] Marty, EPSL 313–314, 56–66 (2012)
[2] Shcheka and Keppler, Nature 490, 531–534 (2012)
[3] Rosa et al., EPSL 532, 116032 (2020)
[4] Zhu et al., Earth Sci. Rev. 224, 103872 (2022)
[5] Hirao et al., PEPI 166, 97–104 (2008)