4:00 PM - 4:15 PM
[SCG48-14] Pressure-induced phase transitions in hemimorphite: A low-frequency Raman spectroscopic study
Keywords:hemimorphite, pressure-induced phase transition, Raman spectroscopy, soft mode
Hemimorphite (Zn4Si2O7(OH)2*H2O, hereafter referred to as phase I) is known to undergo a transition to a high-pressure phase (phase II) at about 2.5 GPa from single-crystal X-ray structure studies (Seryotkin and Bakakin, 2011; Okamoto et al., 2021). Hemimorphite contains both water molecules and OH groups in its structure, and the pressure-induced structural changes and transitions are thought to be closely related to the hydrogen bonds they form. Yamaguchi et al. (2010) reported two phase transitions at higher pressures than the I/II transition, based on high-pressure IR in-situ observation etc. However, these transitions have not yet been well investigated. Soft modes may be observed in the I/II transition. In this study, three transitions, including the I/II transition, were investigated using a micro Raman spectrometer. To observe soft modes, the instrument was tuned so that the low-frequency region could be measured down to 10 cm-1. Using a diamond anvil cell, apiece of natural hemimorphite crystal was pressurized to just under 8 GPa at room temperature. In addition, vibrational calculations were performed using Quantum Espresso.
As phase I was pressurized, a very strong Raman peak appeared at about 12 cm-1 at 2.67 GPa. This peak became the strongest Raman peak in phase II and shifted sharply to the high frequency side with pressure. Vibrational calculations at 3 GPa for phase II indicate that the A1 mode is a rotational vibration of the Zn4Si2O7(OH)2 unit, which coincides exactly with the atomic displacement toward the phase I structure. This is considered to correspond to the soft mode observed in Raman. On the other hand, no such mode was found in phase I from either calculations or experiments. The pressure dependence of the wavenumber w of the soft mode was fitted using w = a|Pc-P|b, yielding a transition pressure Pc of 2.56(1) GPa and a critical exponent b = 0.44(1) (see Figure for fitting). The soft mode disappeared during the decompression process from phase II at about the same pressure as the transition pressure during the pressurization process, confirming a transition to phase I. In phase I, a broad peak centered at 3500 cm-1 is observed due to the presence of H2O molecules in the OH stretching vibration, which broadens further with increasing pressure and increases in intensity around 3600 cm-1. When transformed to II, a new peak at 3500 cm-1 is observed, and this peak moves to the lower frequency side with increasing pressure.
With further compression of II, the soft mode disappears around 5.6 GPa, and a transition to another phase is observed. This is thought to correspond to the transition at 6 GPa described by Yamaguchi et al. When the pressure was reduced from phase III, the transition to phase II occurred at about 4.6 GPa and showed hysteresis of about 1 GPa. This is referred to as phase IV. This transition seems to correspond to that observed at 7 GPa in Yamaguchi et al. On depressurization from phase IV, the transition occurred directly to phase II between 3.7 and 2.7 GPa (bypassing phase III), and finally phase I was recovered. phase I was recovered. These results suggest that phase IV may be poorly crystalline, but not amorphous.
It is difficult to determine the state of hydrogen bonding in heteropolar ores from the spectra because extremely broad peaks due to water molecules appear in the OH stretching vibration region. However, the relatively narrow peaks that appeared in the transition studied here are thought to be due to OH groups. X-ray structure analysis shows that the pressure-induced shortening of the oxygen-oxygen distance at which OH groups are hydrogen bonded in phase II can explain the observed shift of the OH stretching frequency to lower frequencies in phase IV. In phase IV, the hydrogen bonding of the OH groups weakens. For further investigation, it is first necessary to clarify the structure of phases III and IV.
By heating the anode to 500-600 oC, only the H2O molecules can be expelled without destroying the structure. A similar study of "anode ores without water molecules" is underway. Comparison of the two should provide a better understanding of the behavior of hydrogen bonding in these structures.
References:
Okamoto et al. (2021) J. Miner. Petrol. Sciences, 116, 251-262
Seryotkin and Bakakin (2011) Phys. Chem. Minerals, 38, 679-684
Y. Yamaguchi et al. (2010) Abstract, Annual Meeting of the Geochemical Society of Japan, 3D13 09-09
As phase I was pressurized, a very strong Raman peak appeared at about 12 cm-1 at 2.67 GPa. This peak became the strongest Raman peak in phase II and shifted sharply to the high frequency side with pressure. Vibrational calculations at 3 GPa for phase II indicate that the A1 mode is a rotational vibration of the Zn4Si2O7(OH)2 unit, which coincides exactly with the atomic displacement toward the phase I structure. This is considered to correspond to the soft mode observed in Raman. On the other hand, no such mode was found in phase I from either calculations or experiments. The pressure dependence of the wavenumber w of the soft mode was fitted using w = a|Pc-P|b, yielding a transition pressure Pc of 2.56(1) GPa and a critical exponent b = 0.44(1) (see Figure for fitting). The soft mode disappeared during the decompression process from phase II at about the same pressure as the transition pressure during the pressurization process, confirming a transition to phase I. In phase I, a broad peak centered at 3500 cm-1 is observed due to the presence of H2O molecules in the OH stretching vibration, which broadens further with increasing pressure and increases in intensity around 3600 cm-1. When transformed to II, a new peak at 3500 cm-1 is observed, and this peak moves to the lower frequency side with increasing pressure.
With further compression of II, the soft mode disappears around 5.6 GPa, and a transition to another phase is observed. This is thought to correspond to the transition at 6 GPa described by Yamaguchi et al. When the pressure was reduced from phase III, the transition to phase II occurred at about 4.6 GPa and showed hysteresis of about 1 GPa. This is referred to as phase IV. This transition seems to correspond to that observed at 7 GPa in Yamaguchi et al. On depressurization from phase IV, the transition occurred directly to phase II between 3.7 and 2.7 GPa (bypassing phase III), and finally phase I was recovered. phase I was recovered. These results suggest that phase IV may be poorly crystalline, but not amorphous.
It is difficult to determine the state of hydrogen bonding in heteropolar ores from the spectra because extremely broad peaks due to water molecules appear in the OH stretching vibration region. However, the relatively narrow peaks that appeared in the transition studied here are thought to be due to OH groups. X-ray structure analysis shows that the pressure-induced shortening of the oxygen-oxygen distance at which OH groups are hydrogen bonded in phase II can explain the observed shift of the OH stretching frequency to lower frequencies in phase IV. In phase IV, the hydrogen bonding of the OH groups weakens. For further investigation, it is first necessary to clarify the structure of phases III and IV.
By heating the anode to 500-600 oC, only the H2O molecules can be expelled without destroying the structure. A similar study of "anode ores without water molecules" is underway. Comparison of the two should provide a better understanding of the behavior of hydrogen bonding in these structures.
References:
Okamoto et al. (2021) J. Miner. Petrol. Sciences, 116, 251-262
Seryotkin and Bakakin (2011) Phys. Chem. Minerals, 38, 679-684
Y. Yamaguchi et al. (2010) Abstract, Annual Meeting of the Geochemical Society of Japan, 3D13 09-09