Chemistry Reference
In-Depth Information
differences when compared to the spectra of ice Ih, thus casting doubt onto the ear-
lier Antarctic studies [112]. Additional diffraction studies on Antarctic ice samples
concur that it is unlikely to observe a proton-ordered arrangement under such condi-
tions [113]. In addition to the numerous experimental reports, there have also been
theoretical investigations of this proton-ordering transition. In 1981, Minagawa
[114] calculated a Curie-Weiss transition at 69 K to a ferroelectric structure based
on an electrostatic model that reproduced the experimental dipole and quadrupole
of water molecules, and included only nearest neighbor interactions.
The unit cell of ice Ih is hexagonal with space group
P
6
3
/mmc
. The symmetry
of the low-temperature proton-ordered configuration, ice XI, shown in Fig. 2a,
is orthorhombic, space group
Cmc
2
1
, as indicated by neutron scattering [10, 11,
14, 115] and thermal depolarization experiments [13, 116] on KOH doped ice Ih.
Bonds that are oriented parallel to the
c
-axis all point in the same direction. The
ab
layers, composed of bonds oriented perpendicular to the
c
-axis are polarized
parallel to the
b
-axis with alternating layers oppositely aligned. Thus, the struc-
ture is overall antiferroelectric in the
a
and
b
directions and ferroelectric in the
c
direction. This antiferroelectric arrangement of the
ab
layers gives rise to a slight
displacement of the oxygen lattice parallel to the
b
-axis in the direction of the
polarization.
This view has been contested: Iedema et al. [18] referred to more recent claims
as “UFI citings (underidentified ferroelectric ices) in the literature”. Even if the
Cmc
2
1
structure proves to be correct, there is some justification for characterizing
the current state of knowledge of low-temperature ice Ih/XI as “underidentified”.
While a mechanism has been proposed for incomplete conversion of ice Ih to ice
XI [117], several features of the presumed ice Ih/XI transition are not understood:
While the calorimetric signature of the Ih/XI transition is remarkably insensitive
to KOH concentration, the amount of conversion, as measured by the total heat of
transformation, is strongly concentration dependent. If KOH truly acts as a catalyst
and samples have adequate time to equilibrate, there should be no concentration
dependence. The KOH seems to be playing another role, perhaps related to the
crystal strain discussed by Johari [117]. Furthermore, there are reports that protons
in ice become immobile below a certain temperature due to being trapped by
the defects present in ice [118]. Wooldridge and Devlin [43] performed Fourier
transform infrared (FTIR) experiments that indicated that proton motion comes
to a halt below 100 K. More recently, “soft-landing” experiments by Cowin et al.
[44] indicate that hydronium ions are in fact immobile at all temperatures below
190 K. If hydroxide is as immobile as excess protons at low temperature, then
the basis for the catalytic role of hydroxide would be thrown in doubt. Recent
dielectric and calorimetric experiments [119] indicate that the alkali hydroxide
dopants polarize nearby water molecules to promote orientational ordering at low
temperatures that may explain the observed weak concentration dependence on
the amount of transformation achieved.
