|
Titel |
Ice from Supercooled Water: Neither Hexagonal nor Cubic |
VerfasserIn |
Tamsin Malkin, Benjamin J. Murray, Andrey V. Brukhno, Jamshed Anwar, Christoph G. Salzmann |
Konferenz |
EGU General Assembly 2011
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 13 (2011) |
Datensatznummer |
250054824
|
|
|
|
Zusammenfassung |
Under atmospherically relevant conditions both the stable hexagonal phase (ice Ih) and a
metastable phase of ice I that lacks hexagonal symmetry can form. The ice which initially
crystallises from pure water or solution droplets that freeze homogeneously has been
identified in the past as being cubic ice (ice Ic) and it was suggested that this ice contains
hexagonal like stacking faults [Murray et al., 2005; Murray and Bertram, 2006]. Here we
show through X-ray diffraction experiments and a computational study that ice that initially
crystallises from water when it freezes homogeneously is neither cubic nor hexagonal. It is in
fact in fact, a mixed form of ice I, which randomly switches between sequences of cubic and
hexagonal layers alternating perpendicular to the basal face {0001}. Normally it is assumed
that a distinct phase crystallises rather than ice which lacks either cubic or hexagonal
symmetry.
Droplets of pure water suspended in an oil emulsion of median diameter 0.9 μm were cooled
down within a powder X-ray diffractometer at 30 K min-1 and froze homogeneously at
around 232 K. The diffraction pattern of these frozen droplets lacks some of the peaks
associated with hexagonal symmetry and the patterns are similar to other ice Ic patterns in the
literature. Modelling of these diffraction patterns reveals that the ice is ice I with randomly
stacked cubic and hexagonal sequences. Hansen et al. [2008 a,b] concluded that ice I
recrystallised from ice V and ice IX is also a mixed structure with 25 - 60% hexagonal
sequences in ice Ic. Our result is important because it suggests that what has been called
cubic ice is in fact ice I with random stacking sequences lacking the pure cubic and
hexagonal symmetries.
We also conducted Monte Carlo simulations of homogeneous ice formation in bulk water in
which ice I crystallisation is encouraged, but no preference is made between hexagonal and
cubic structures [Brukhno et al, 2008]. The resulting ice is most often a mixed
cubic-hexagonal structure.
The experimental and computational studies show that the growth of ice I from supercooled
water yields neither cubic nor hexagonal forms, but rather a mixed ice structure with random
alteration between cubic and hexagonal sequences. These results suggest that ice growth is
primarily governed by the nearest neighbour environment, making stacking faults a general
feature.
Murray, B. J. et al., Nature, 434, 202-204 (2005).
Murray, B. J. and A. K. Bertram, Phys. Chem. Chem. Phys,. 8, 186-192 (2006)
.
Hansen, T. C. et al., J. Phys. Condens. Matter, 20, 285104 (2008a).
Hansen, T. C. et al., J. Phys. Condens. Matter, 20, 285105 (2008b).
Brukhno, A. et al. J. Phys. Condens. Matter, 20, 494243 (2008). |
|
|
|
|
|