1.2 Bragg Scattering and Bernal–Fowler Water
15
model of water, which corresponds to the X-ray and neutron scattering data, is still
developing [33, 34, 36, 43]. Obviously, water is more disordered than the ideal
tetrahedral arrangements of amorphous SiO 2 . Though the RDF analysis roughly
confirms the trigonal geometry of water molecules in the liquid state and shows that
time-averaged four-coordinated structures dominante [12, 44].
Ice generally reproduces the data for water, but there are several additionally
resolved maxima near 2.2, 3.4, 5.2 Å. These maxima show that ice has a long-order
and a higher degree of tetrahedrality compared to liquid water. Why is the local
diffusion-averaged structure of water different from that of ice?
An analysis of the fine structure of the water RDF made by Samoilov [45] showed
that there is an additional smaller maximum at 3.5 Å, which is not expected from the
purely tetrahedral arrangement of H 2 O molecules. He assumed that this maximum
corresponds to the interstitial molecules trapped in the hexagonal rings (see Fig. 1.9).
Samoilov used a pair distribution function, defined as 4πr
2
ρ(r ), where ρ(r ) is the
density, which can be understood as the deviation of the local density of water in
comparison with the averaged density. Figure 1.11 shows (a) the experimental and
(b) the calculated distribution functions. A relative subtraction shows that there is
an additional density peak near 3.5 Å, which does not appear in the ideal structure.
Samoilov found that several percent of water molecules are temporarily trapped in the
interstitial cavities. This assumption explains the deviation from the ideal structure
of Bernal and Fowler, but does not explain why the interstitial states differ from
the lattice states. Most recent studies of the radial distribution function [42] show
that phenomenological constants allow a description of the experimental data with
satisfying accuracy, and no further improvement of the experimental data is needed.
However, there is still no simple structural model, which describes the RDF and
satisfies the scope of the thermodynamics, electrodynamics, and structural properties
of water.
Fig. 1.11 A comparison of
a experimental and b
theoretical curves of the
radial distribution of oxygen
atomic density for water at
275 K. The function
difference c shows a hidden
density maximum at 3.5 Å,
which is not expected from
the ideal tetrahedral
structure. It corresponds to
the molecules trapped in the
cavities of the hexagonal
rings (interstitial molecules)
15
model of water, which corresponds to the X-ray and neutron scattering data, is still
developing [33, 34, 36, 43]. Obviously, water is more disordered than the ideal
tetrahedral arrangements of amorphous SiO 2 . Though the RDF analysis roughly
confirms the trigonal geometry of water molecules in the liquid state and shows that
time-averaged four-coordinated structures dominante [12, 44].
Ice generally reproduces the data for water, but there are several additionally
resolved maxima near 2.2, 3.4, 5.2 Å. These maxima show that ice has a long-order
and a higher degree of tetrahedrality compared to liquid water. Why is the local
diffusion-averaged structure of water different from that of ice?
An analysis of the fine structure of the water RDF made by Samoilov [45] showed
that there is an additional smaller maximum at 3.5 Å, which is not expected from the
purely tetrahedral arrangement of H 2 O molecules. He assumed that this maximum
corresponds to the interstitial molecules trapped in the hexagonal rings (see Fig. 1.9).
Samoilov used a pair distribution function, defined as 4πr
2
ρ(r ), where ρ(r ) is the
density, which can be understood as the deviation of the local density of water in
comparison with the averaged density. Figure 1.11 shows (a) the experimental and
(b) the calculated distribution functions. A relative subtraction shows that there is
an additional density peak near 3.5 Å, which does not appear in the ideal structure.
Samoilov found that several percent of water molecules are temporarily trapped in the
interstitial cavities. This assumption explains the deviation from the ideal structure
of Bernal and Fowler, but does not explain why the interstitial states differ from
the lattice states. Most recent studies of the radial distribution function [42] show
that phenomenological constants allow a description of the experimental data with
satisfying accuracy, and no further improvement of the experimental data is needed.
However, there is still no simple structural model, which describes the RDF and
satisfies the scope of the thermodynamics, electrodynamics, and structural properties
of water.
Fig. 1.11 A comparison of
a experimental and b
theoretical curves of the
radial distribution of oxygen
atomic density for water at
275 K. The function
difference c shows a hidden
density maximum at 3.5 Å,
which is not expected from
the ideal tetrahedral
structure. It corresponds to
the molecules trapped in the
cavities of the hexagonal
rings (interstitial molecules)