17:15 〜 19:15
[SIT20-P02] Ab initio structural models and infrared spectra of hydrous bridgmanite
キーワード：第一原理計算、ブリッジマナイト、IRスペクトル
Nominal anhydrous minerals (NAMs) that contain hydrogen as an impurity are the main form of water in the Earth's interior. The water in NAMs affects various physical properties, such as melting relations (Inoue, 1994), rheological properties (Karato & Jung, 2023), and electrical conductivity (Karato & Wang, 2013). Therefore, elucidating the structure and properties of NAMs is important for understanding the evolution and dynamics of the Earth.
Bridgmanite, the most abundant mineral in the Earth is reported to contain 0.1 wt% H2O, based on recent Fourier transform infrared spectroscopy (FTIR) and NanoSIMS results (Fu et al., 2019). On the other hand, Bolfan-Casanova et al. (2003) reported a water content of 0-2 ppm, and there has been no consensus on the water content in bridgmanite so far. In addition, indirect information on the structure has been obtained using infrared spectroscopy. Furthermore, these experiments analyzed samples recovered at ambient conditions, and the state of water under high pressure remains unclear.
The aim of this study is to determine the hydrogen defect structure of hydrous bridgmanite. To this end, we first assumed various cation substitution forms and performed structural optimization up to the lower mantle pressure using first-principles calculations. Then, we used the structural models to determine the vibrational properties using density functional perturbation theory. From the obtained OH vibrational properties, we generated infrared spectra at ambient and high pressure conditions and compared them with those obtained in the experiment at ambient conditions.
The structures used in the calculations in this study are supercells of 2x2x1 MgSiO3 bridgmanite in which Mg atoms are replaced by two hydrogen atoms (Mg2+<->2H+;Mg15Si16O48H2), Si atoms are replaced by four hydrogen atoms (Si4+<->4H+;Mg16Si15O48H4), and Si atoms are replaced by Al and H atoms (Si4+<->Al3++H+;Mg16Si15AlO48H). Structural optimizations were carried out for all structural parameters at static 0 K conditions until the residual force was less than 1.0x10-5 Ry/a.u. Pressure calculations were carried out from 0 to 120 GPa.
The results of the calculations at ambient pressure showed that multiple stable hydrogen positions were obtained for each substitution type. It was also found that the OH stretching frequencies differed depending on the position of the hydrogen. Under high pressure, the covalent OH distance (ROH) increased and the hydrogen bond distance (RO...H) decreased in all defect structures. In addition, hydrogen bond symmetrization was observed in some Mg defect structures above 80 GPa. The results of the calculations under high pressure showed a decrease in the OH stretching frequency as the hydrogen bonding strength increased. This trend is generally in line with the empirical rule of Libowitzky(1999). However, an increase in the OH stretching frequency was observed above the pressure at which hydrogen bonding symmetry occurs.
Some of the OH stretching frequencies with Mg and Si defects were similar to those of the experiments at 0 GPa, However, they could not fully explain the experimental results as the calculations produced additional peaks that were not observed in the experiment. As a next step, we need to consider the effects such as incorporation of Fe and temperature, which were not included in the calculations.
Bridgmanite, the most abundant mineral in the Earth is reported to contain 0.1 wt% H2O, based on recent Fourier transform infrared spectroscopy (FTIR) and NanoSIMS results (Fu et al., 2019). On the other hand, Bolfan-Casanova et al. (2003) reported a water content of 0-2 ppm, and there has been no consensus on the water content in bridgmanite so far. In addition, indirect information on the structure has been obtained using infrared spectroscopy. Furthermore, these experiments analyzed samples recovered at ambient conditions, and the state of water under high pressure remains unclear.
The aim of this study is to determine the hydrogen defect structure of hydrous bridgmanite. To this end, we first assumed various cation substitution forms and performed structural optimization up to the lower mantle pressure using first-principles calculations. Then, we used the structural models to determine the vibrational properties using density functional perturbation theory. From the obtained OH vibrational properties, we generated infrared spectra at ambient and high pressure conditions and compared them with those obtained in the experiment at ambient conditions.
The structures used in the calculations in this study are supercells of 2x2x1 MgSiO3 bridgmanite in which Mg atoms are replaced by two hydrogen atoms (Mg2+<->2H+;Mg15Si16O48H2), Si atoms are replaced by four hydrogen atoms (Si4+<->4H+;Mg16Si15O48H4), and Si atoms are replaced by Al and H atoms (Si4+<->Al3++H+;Mg16Si15AlO48H). Structural optimizations were carried out for all structural parameters at static 0 K conditions until the residual force was less than 1.0x10-5 Ry/a.u. Pressure calculations were carried out from 0 to 120 GPa.
The results of the calculations at ambient pressure showed that multiple stable hydrogen positions were obtained for each substitution type. It was also found that the OH stretching frequencies differed depending on the position of the hydrogen. Under high pressure, the covalent OH distance (ROH) increased and the hydrogen bond distance (RO...H) decreased in all defect structures. In addition, hydrogen bond symmetrization was observed in some Mg defect structures above 80 GPa. The results of the calculations under high pressure showed a decrease in the OH stretching frequency as the hydrogen bonding strength increased. This trend is generally in line with the empirical rule of Libowitzky(1999). However, an increase in the OH stretching frequency was observed above the pressure at which hydrogen bonding symmetry occurs.
Some of the OH stretching frequencies with Mg and Si defects were similar to those of the experiments at 0 GPa, However, they could not fully explain the experimental results as the calculations produced additional peaks that were not observed in the experiment. As a next step, we need to consider the effects such as incorporation of Fe and temperature, which were not included in the calculations.