pubmed.ncbi.nlm.nih.gov

Experiments indicating a second hydrogen ordered phase of ice VI - PubMed

  • ️Mon Jan 01 2018

. 2018 Mar 26;9(18):4224-4234.

doi: 10.1039/c8sc00135a. eCollection 2018 May 14.

Affiliations

Experiments indicating a second hydrogen ordered phase of ice VI

Tobias M Gasser et al. Chem Sci. 2018.

Abstract

In the last twelve years five new ice phases were experimentally prepared. Two of them are empty clathrate hydrates and three of them represent hydrogen ordered counterparts of previously known disordered ice phases. Here, we report on hydrogen ordering in ice VI samples produced by cooling at pressures up to 2.00 GPa. Based on results from calorimetry, dielectric relaxation spectroscopy, Raman spectroscopy, and powder X-ray diffraction the existence of a second hydrogen ordered polymorph related to ice VI is suggested. Powder X-ray data show the oxygen network to be the one of ice VI. For the 1.80 GPa sample the activation energy from dielectric spectroscopy is 45 kJ mol-1, which is much larger than for the known hydrogen ordered proxy of ice VI, ice XV. Raman spectroscopy indicates the 1.80 GPa sample to be more ordered than ice XV. It is further distinct from ice XV in that it experiences hydrogen disordering above ≈103 K which is 26 K below the ice XV to ice VI disordering transition. Consequently, below 103 K it is thermodynamically more stable than ice XV, adding a stability region to the phase diagram of water. For the time being we suggest to call this new phase ice β-XV and to relabel it ice XVIII once its crystal structure is known.

PubMed Disclaimer

Figures

Fig. 1
Fig. 1. The equilibrium phase diagram of H2O including the colored area in which the new hydrogen ordered phase, tentatively called ice β-XV, is stable. The blue arrow at 1.8 GPa shows the cooling process of ice VI from 255 K leading to the new polymorph at 77 K, which is metastable in the ice VIII domain. Dashed lines are extrapolations from experimentally determined phase boundaries (solid lines).
Fig. 2
Fig. 2. Differential scanning calorimetry (DSC) thermograms of samples of the new polymorph prepared at pressures of 1.00 GPa (black), 1.45 GPa (blue), 1.60 GPa (green), 1.80 GPa (violet) and 2.00 GPa (orange) and cooled at the given pressures with a rate of approximately 3 K min–1. The baselines were corrected and shifted to 0 J mol–1 K–1 at 96 K. The indicated enthalpies are average values from 2 (orange), 19 (violet), 4 (green), 5 (blue) and 8 (black) scans. Heating rate is 10 K min–1 for all DSC traces.
Fig. 3
Fig. 3. Influence of recooling at ambient pressure on the 1.80 GPa sample (ice β-XV). DSC thermograms of samples prepared at 1.80 GPa heated to 117 K (black trace), recooled to 93 K at 1 K min–1 and heated a second time to 140 K (blue trace). Heating rate is 10 K min–1 for all DSC traces. The baselines were corrected and shifted to 0 J mol–1 K–1 at 97 K.
Fig. 4
Fig. 4. Dielectric loss spectra of the 1.80 GPa sample (ice β-XV) during first heating. Spectra acquired for T ≤ 114 K are shown in panel (a) and data referring to higher temperatures are shown in panel (b). Please note that the frequency axes of the two panels are different.
Fig. 5
Fig. 5. Dielectric loss spectra of the reference ice XV sample produced at 1.00 GPa, recorded upon heating. As in Fig. 4 spectra for T ≤ 114 K are shown in panel (a) and those for T ≥ 114 K are displayed in panel (b). Please note that the frequency axes of the two panels are different.
Fig. 6
Fig. 6. Dielectric loss spectra of the same 1.80 GPa sample (ice β-XV) for which data are shown in Fig. 4. (a) Spectra measured upon first heating to 114 K are re-plotted here with a zoomed frequency scale to facilitate comparison with subsequent measurements on this sample. (b) Triangles represent data obtained during subsequent cooling which yielded sample 1.80 GParec(114 K). The spectra recorded upon heating of this sample are represented in (b) as asterisks. The dashed lines refer to the amplitude of the dielectric loss spectrum measured at 102 K for the 1.80 GPa sample upon first heating. To guide the eye, data points are connected by lines.
Fig. 7
Fig. 7. Arrhenius diagram reflecting data recorded upon heating of (from top to bottom) the reference XV sample (cf.Fig. 5b), sample 1.80 GParec(114 K) (cf.Fig. 6b) and the 1.80 GPa sample (ice β-XV). To enhance the clarity of presentation, the relaxation times for sample 1.80 GParec(114 K) are shifted up by 1.5 decades and those for the reference ice XV sample by 3 decades. The region of the low-temperature phase transition of the 1.80 GPa sample was examined in three independent measurements (blue symbols). The open stars correspond to the first heating run of the 1.80 GPa sample (cf.Fig. 6a) up to 114 K (indicated by the blue arrow), whereas the closed stars refer to the heating of the 1.80 GParec(114 K) sample (second heating, cf.Fig. 6b). Dotted lines indicate phase transition temperatures for the 1.80 GPa sample. The solid lines correspond to Arrhenius laws with the activation energies indicated in the figure. The uncertainty of the various activation energies is estimated to be about ±2 kJ mol–1.
Fig. 8
Fig. 8. Raman spectra of (a) the 1.80 GPa sample, tentatively designated ice β-XV as well as of (b) ice XVrec(130 K) and (c) ice VI samples, containing 9 mol% HDO. The ice XV sample was recooled from 130 K in the microstat at 10 mbar as described by Whale et al. All spectra were recorded at 84 K. Spectra were smoothed by a 7-point Savitzky–Golay filter and normalized to the most intense peak for each individual trace.
Fig. 9
Fig. 9. (a) Raman spectra of ice β-XV acquired upon heating. (b) Raman spectra of ice XV (recooled from 120 K at 1 bar) at different temperatures. Spectra are normalized to the most intense peak in each individual trace. A Savitzky–Golay filter was not applied for the processing of these spectra. (c) Relative intensity RI450 of the librational band at 450 cm–1 (see text for definition). Error bars are on the order of the size of the symbols.
Fig. 10
Fig. 10. (a) Powder X-ray diffractograms of ice β-XV recorded at 20 K. Crossed out peaks stem from the sample holder. (b) Comparison of diffraction pattern for ice VI, ice XV, and ice β-XV at 80 K in the range highlighted by the orange rectangle in (a). (c) Peak positions for ice β-XV at different temperatures (black trace), and for ice β-XV after recooling from 118 K, i.e., for ice XV (blue trace). The lines are a guide to the eye. Note that the diffractograms in (a) were recorded on the XPERT III, and the diffractograms for (b) and (c) on the D5000 diffractometer.

Similar articles

Cited by

References

    1. del Rosso L., Grazzi F., Celli M., Colognesi D., Garcia-Sakai V., Ulivi L. J. Phys. Chem. C. 2016;120:26955–26959.
    1. Salzmann C. G., Radaelli P. G., Slater B., Finney J. L. Phys. Chem. Chem. Phys. 2011;13:18468–18480. - PubMed
    1. Komatsu K., Noritake F., Machida S., Sano-Furukawa A., Hattori T., Yamane R., Kagi H. Sci. Rep. 2016;6:28920. - PMC - PubMed
    1. Bridgman P. W. Proc. Am. Acad. Arts Sci. 1911;47:319–343.
    1. Kamb B. Science. 1965;150:205–209. - PubMed