Stability of filled-ice structure of methane hydrate and Existence of a post filled-ice structure above 40 GPa

Transcription

1 Stability of filled-ice structure of methane hydrate and Existence of a post filled-ice structure above 40 GPa S. Machida 1, H. Hirai 1*, T. Kawamura 2, Y. Yamamoto 2, and T. Yagi 3 1: Graduate School of Life and Environmental Science, University of Tsukuba Tsukuba, Ibaraki, , Japan 2: National Institute of Advanced Industrial Science and Technology, Tsukuba,Ibaraki , Japan 3: Institute for Solid State Physics, Tokyo University, Kashiwanoha, Kashiwa,Chiba , Japan abstract High-pressure X-ray diffractometry and Raman spectroscopy showed that filled-ice-ih structure of methane hydrate transformed to a new high-pressure structure at approximately 40 GPa, which survived at least up to 86 GPa. A reason for the incomparable retention of the filled-ice-ih structure up to 40 GPa was clarified, because the structures for the other hydrate are known to decompose below 6.5 GPa. Substantial softening and splitting of vibration modes of methane molecule indicate increased intermolecular interactions, which results in excellent stabilization. Introduction Methane hydrate is expected as a clean energy resource for the next generation, while methane is an effective greenhouse gas causing global warming (Sloan 1998). In the solar system, methane hydrate is thought to be an important constituent of outer giant planets and their satellites, such as Uranus, Neptune, and Titan. Comprehensive high-pressure studies of gas hydrates performed in the last few years have greatly contributed to our knowledge of the structural changes in gas hydrates under pressures below 10 GPa. An outline of the structural changes depending of guest size and pressure has been presented within the guest size up to 4.5 A (Hirai et al. 2004). In the case of methane hydrate, the initial cubic structure transforms to a hexagonal structure at about 1 GPa, and it further transforms to a filled-ice-ih structure (FIIhS) (Chou et al. 2000, Hirai et al. 2001, Loveday et al. 2001, Shimizu et al. 2002) at about 2 GPa. The FIIhS consists of channels, not of cages as for the former two structures, formed by water molecules, and these channels are filled with guest species (Loveday et al. 2002). The other gas hydrates having guest size from argon to nitrogen finally transforms to the common FIIhS, although their initial structure and intermediate structure are different (Kurnosov et al. 2001; Hirai et al. 2002; Sasaki et al. 2003; Desgreniers et al. 2003; Hirai et al. 2004). The transition pressures from cage structures to the FIIhS are clearly correlating to the guest size; the larger the guest size, the higher the pressure at which the hydrate is kept. This suggests that the larger guests are more favorable for holding the shrinking cages during compression (Hirai et al. 2004). In contrast, as for FIIhSs, there is no dependence of stability on the guest size. The decomposition pressures of FIIhSs are various; i. e. only methane hydrate survives at least to 40 GPa (Hirai et al. 2003; Hirai et al. 2004), while the other gas hydrates decompose below 6.5 GPa. The reason for stabilizing the FIIhS of methane hydrate

2 has not been clarified so far. A theoretical study using the first principle calculation reported the retention of the FIIhS of methane hydrate until 100 GPa (Iitaka and Ebisuzaki, 2003). Here, interesting points arise whether the FIIhS of methane hydrate survives above 40 GPa or a certain post filled-ice-ih structure (post-fiihs) exists, and why only the FIIhS of methane hydrate retains under such high pressure. In this study, high-pressure experiments of methane hydrates were performed in the pressure range from 0.2 GPa to 86 GPa by using X-ray diffractometry (XRD) and Raman spectroscopy in order to clarify these subjects. Experiments A lever-and-spring type diamond anvil cell (DAC) was used in the high-pressure experiments. The ruby fluorescence method was used for the pressure measurements. The accuracy of the present measurement system is 0.1 GPa, taking the resolution of the spectrometer and the analytical procedure into account. The XRD study was performed using synchrotron radiation on BL-18C and BL-13A at the Photon Factory, High Energy Accelerator Research Organization (KEK). A monochromatized beam with a wavelength of nm was used. The initial material was methane hydrate powder, which was prepared using a conventional ice-gas interaction method under the following conditions: 15 MPa and -3. This powder consisted of almost pure methane hydrate with full occupancy, according to a combustion analysis. The sample powder was placed into a gasket hole in a vessel cooled by liquid nitrogen in order to prevent the decomposition of the sample. The sample was sealed by loading the anvils to approximately 0.2 GPa at a low temperature, and then the DAC was warmed to room temperature. The XRD and Raman studies and optical observations were conducted at room temperature in a pressure range from 0.2 GPa to 86GPa. Results Under an optical microscope, change was not detected visually after the transition from the hexagonal structure to the FIIhS at 2.0 GPa. In the XRD patterns, the typical diffraction lines such as, 011, 110, 002, 121, 112, 013, 132 and 123 of the FIIhS were observed below 40 GPa, although the XRD patterns became somewhat broad as observed in the previous study (Hirai et al. 2003). At 40 GPa, four new peaks began to appear; one peak at a little higher angle of 002 peak, one at near to 112, and two at near to 123 peak (Fig.1). The relative intensities of these peaks became stronger above 40 GPa. Instead, the remaining 002, 121 and 112 peaks weakened above 40 GPa and completely disappeared at 60GPa. At above 60 GPa, only the four peaks were observed with strong intensities (Fig.1). The XRD patterns characterized by the four peaks continued at least until to 86 GPa. With decreasing pressure, the four peaks were observed until 40 or 35 GPa, and the XRD patterns reverted to those observed below 40GPa. This indicated that the transition at 40 GPa was reversible. In order to check decomposition of the FIIhS into solid methane and ice VII, the additional high-pressure experiments were performed for solid methane in the pressure range from 1GPa to 86 GPa. The phase changes observed for solid methane were described elsewhere (Hirai et al. 2005). The observed XRD patterns for the FIIhS were completely different from those of solid methane. The relative intensities of ice VII did not increase above 40 GPa. Thus, the observed XRD patterns were intrinsic for a new high-pressure structure of methane hydrate. The heating treatments were carried out by using CO 2 -laser at 50GPa, and 86GPa for annealing the samples because the XRD patterns above 40 GPa were broad. The XRD patterns were not remarkably improved after the heating treatments. Fig.2 shows Raman spectra of intramolecule vibration modes of methane; one is

3 symmetric stretching mode Ice ν1 and the other is antisymmetric stretching mode ν3. Fig.3 shows variations in Raman shift of these modes with pressure. Changes in vibration modes were observed at approximately 15 and 40 GPa. At 14 to 18 GPa, both of ν1 and ν3 peaks began to split; a new peak appeared at a little lower frequency for these peaks, respectively (Fig.2). The new ν1 peak became gradually strong. The originally existing ν1 peak weakened. The new ν3 peak became strong above 20 GPa (Fig.2). The originally existing ν3 peak continued 86 GPa. In addition to the distinct peak splits, discontinuous changes in the slops of the Raman shift were observed at about 40GPa (Fig.3). The change was more remarkable for the new ν1 and new ν3 peaks. The peak splits and discontinuous changes in the slops were observed both in increasing and decreasing pressure at Fig.1 Representative XRD patterns with pressure change. almost same pressures, indicating that these changes occurred reversibly Discussion The XRD and Raman studies revealed that several changes took place in the methane hydrate samples at approximately 15 GPa and 40GPa. The change at about 15 GPa was observed only in Raman spectra, while the XRD patterns were unchanged at around 15 GPa. Thus, this change is interpreted as the change in the vibration modes of methane molecule within the same fundamental structure of FIIhS. The change at 40 GPa was

4 observed both in the XRD patterns and Raman spectra. This implies that a fundamental structure changed from the FIIhS to a certain post-filled ice structure (post-fiihs) at 40 GPa, and the change of vibration modes was induced in the post-fiihs by the structural change. The theoretical study using the first principle calculation reported that the FIIhS will be held up to 100 GPa (Iitaka and Ebisuzaki 2003). However, in the present study, the change in the XRD patterns was clearly observed above 40 GPa, thus the retention of the FIIhS can be said to be limited up to 40 GPa. We attempted to examine the post-fiihs using the four new diffraction lines, in spite of the small number of diffraction lines. These lines were somewhat broad but clearly observed along Debye-rings. Two possible ways of indexing were suggested. One indexing was as orthorhombic Fig. 2 Raman spectra of intramolecular vibration modes, ν1 and ν3, for methane. symmetry with the lattice parameters of a= nm, b= nm, c= nm at 50.6 GPa. In this case, the deviations between the observed d-values and calculated ones were less than 0.14 %. These lattice parameters were almost same as those calculated as the FIIhS by the theoretical study (Iitaka and Ebisuzaki 2003), so the volume was almost same. But the extinction rule of diffraction was completely different from that of FIIhS. The second indexing was as another orthorhombic symmetry with the lattice parameters of a= nm, b= nm, c= nm at 50.6 GPa. The deviations of d-values were lees than 0.29 %. In this case, two time of the length of a-axis became almost that of c-axis, indicating pseudo-tetragonal symmetry. The volumes above 50 GPa were a little larger by about 1 to 3 % than those calculated as the FIIhS above 50 GPa, but they were, of course, smaller than that of 40 GPa. In any case, the volume change was very small. And, the relative intensity-ratio of the post- FIIhS to the coexisting ice VII were almost same for that of FIIhS to

5 ice VII in the XRD patterns, i.e. the amount of ice VII did not increased or decreased at the transition. Thus, the molecular ratio of methane to water is suggested to be unchanged between both structures. Considering the unchanged molecular ratio and the small volume change, the fundamental structure of the post-fiihs might be not largely different from that of the FIIhS. The true symmetry of the post-fiihs can not be determined at present because of the limited diffraction lines. Fig.3 Variation of Raman shift of ν1 and ν3 with pressure. As described above, the FIIhSs of other gas hydrates such as argon-, krypton-, and nitrogen-hydrate decompose at various pressures below 6.5 GPa. Only methane hydrate showed remarkable stability and persisted to 40 GPa. The decomposition pressures of FIIhSs are not dependent on the guest size. In the Raman spectra, the symmetrical stretching mode, ν3, of methane molecule began to split at about 14 GPa and ν1 also did at 16 GPa. Both new vibration modes appeared with low frequencies, indicating that the softened vibration modes began to be produced in the structure at around 15 GPa. Such softening arises from increased intermolecular interaction in molecular crystals and ice materials (Goncharov et al. 1999; Aoki et al. 1996). Thus the softening of the intramolecular vibration observed in the present study can be thought as appearance of the attractive intermolecular interaction between the methane and the host water and also between methane molecules. This increased intermolecular interaction might result in additional stabilization of the FIIhS under very high pressure. Similar explanation was reported for retention of hydrogen hydrate up to 280K (Mao et al 2002). In that explanation, the softening of vibration mode of hydrogen molecule observed was an indication of increased intermolecular interaction, and this attractive interaction contributes to the additional stability even under 260 K and 200MPa (Mao et al. 2002). In order to confirm the increased interaction of the methane molecule in the FIIhS, we compared the vibration modes in the FIIhS with those in solid methane which is composed only of methane molecule. The detail Raman study of solid methane was described elsewhere (Hirai et al. 2005). Below 20 GPa the ν1 modes of the FIIhS represented higher frequencies than those of solid methane by 10 to 20 cm -1, from about 20 to 40 GPa the former was similar to or lower than the latter, and above 40 GPa the former became lower than the latter by 10 to 30 cm -1. The softening above 20 GPa evidently shows that additional increased interaction between methane and water molecules is produced in the FIIhS. The present finding of the post-fiihs of methane hydrate and its retention up to 86 GPa leads to a new development in ice-related material science. And, the observation of the increased interaction resulting in the remarkable stability of methane hydrate will help for a general understanding on gas hydrate stability. The detailed structural analyses of the

6 post-fiihs and vibration modes of methane molecules will be future subjects. REFERENCES Aoki, K., Yamawaki, H., Sakashita, M., and Fujihisa, S., Infrared absorption study of the hydrogen-bond symmetrization in ice to 110 GPa. Phys. Rev.B, 54, Chou, I-M., Sharma, A., Burruss, R. C., Shu, J., Mao, H. K., and Hemley, R. J., Transformations in methane hydrates. Proc. Nat. Acad. Sci., 97, Desgreniers, S., Flacau, R., Klig, D. D., and Tse, J. S., Dense noble gas hydrates: phase stability and crystalline structures. Abstract of the international conference of CeSMEC, Miami March Goncharov, A. F., Struzhkin, V. V., Mao, H. K., and Hemley, R. J., Raman spectroscopy of dense H 2 O and the transition to symmetric hydrogen bonds. Phys. Rev. Lett., 83, Hirai, H., Uchihara, Y., Fujihisa, H., Sakashita, M., Katoh, E., Aoki, K., Nagashima, K., Yamamoto, Y., and Yagi, T., High-pressure structures of methane hydrate observed up to 8 GPa at room temperature. J. Chem. Phys., 115, Hirai, H., Uchihara, Y., Nishimura, Y., Kawamura, T., Yamamoto, Y., and Yagi, T., Structural changes of argon hydrate under high pressure. J. Phys. Chem., B 106, Hirai, H., Tanaka, T., Kawamura, T., Yamamoto, Y., and Yagi, T., Retention of filled ice structure of methane hydrate up to 42 GPa. Phys. Rev. B, 68, Hirai, H., Tanaka, T., Kawamura, T., Yamamoto, Y., and Yagi, T., Structural changes in gas hydrates and existence of a filled ice structure of methane hydrate above 40 GPa. J. Phys. Chem. Solids, 65, Hirai, H., Konagai, K., Machida, S., Yagi, T., and Miyajima, N., 2005.Phase changes of solid methane under very high pressure. to be published. Iitaka, T. and Ebisuzaki, T., Methan hydrate under highpressure. Phys. Rev. B, 68, Kurnosov, A.V., Manakov, A. Y., Komarov, V. Y., Voronin, V. I., Teplykh, A. E., and Dyadin, Y. A., A new Gas hydrate structure. Dokl. Phys. Chem., 381, Loveday, J. S., Nelmes, R. J., Guthrle, M., Belmonte, S. A., Allan, D. R., Klug, D. D., Tse, J. S., and Handa, Y. P., Stable methane hydrate above 2 GPa and the source of Titan s atmospheric methane. Nature, 410, Loveday, J. S., Nelmes, R. J., Guthrie, M., Klug, D. D., and Tse, J. S., Transition from cage clathrate to filled ice: the structure of methane hydrate III. Phys. Rev. Lett., 87, Mao, W. L., Mao, H. K., Goncharov, A. F., Struzhkin, V. V., Guo, Q., and Hemley, R. J., Hydrogen clusters in clathrate hydrate. Science, 297, Sasaki, S., Hori, S., Kume, T., and Shimizu, H., Microscopic observation and in situ Raman scattering studies on high-pressure phase transformations of a synthetic nitrogen hydrate. J. Phys. Chem., 118, Shimizu, H., Kumazaki, T., Kume, T., and Sasaki, S., In-situ observations of high-pressure phases transformations in a synthetic methane hydrate. J. Phys. Chem., 106, Sloan, E. D., 1998.Clathrate Hydrates of Natural Gases, 2nd ed, Marcel Dekker Inc., New York.