5.1 Nucleation of Ice
115
than half a century since the initial report, the underlying mechanisms are still under
active investigations.
An issue that further complicates the matter is an impact of surface roughness
on the heterogeneous nucleation potency of a material. Purely geometrical argument shows that surface roughness enhances heterogeneous nucleation potency of a
substrate [13]. Nanoscale roughness in the form of etch pits on a substrate surface
can have a profound effect on the kinetics of heterogeneous nucleation of ice [9].
Another surprising finding is that, in nano-porous alumina, liquid water subcools
to −42 °C and then freezes to metastable I c , not to thermodynamically stable I h [14].
I c thus formed remains stable during hearing, up to 273 K at which point it melts
to liquid water [14]. How could this happen? Nano-porous alumina had one of the
highest surface-to-volume ratios of any system studied to date and heterogeneous
nucleation of ice ought to be expected more than ever, and yet liquid water confined
inside nano-porous alumina subcooled to −42 °C. Of course, if the pores were
so small that the space left was insufficient for a long-range positional order (that
defines a crystalline phase) to manifest then no freezing of water would have been
expected. However, the pores in the nano-porous alumina were not that small; indeed,
a metastable I c crystalline phase could still form inside the nano-porous alumina [14].
These are all important outstanding issues that have not been resolved to date.
The so-called “Contact nucleation” refers to a nucleation process of ice on a solid
wall at an air–liquid interface (like around the three-phase lines of a meniscus). Such
contact nucleation has been found to be more efficient than “immersion nucleation”
(nucleation of ice on a solid wall that is totally immersed in liquid water) [1]. In
the presence of a potent heterogeneous nucleation substrate, surface nucleation preferentially occurred at the three-phase lines where the air, liquid water, and a solid
wall met. It was also found that the nanoscopic (nanometer scale) surface roughness
played a significant role while the microscopic (micrometer scale) surface roughness did not [1]. It has further been shown that rough substrates enhanced contact
nucleation while smooth ones did not [1]. It appears that steps, pores, cracks, or
other surface features of the sizes of the order of the critical nucleus may effectively
promote nucleation by lowering the free energy barrier [1, 15]. These findings are
most likely relevant to clathrate hydrate nucleation as we will see in the next section.
5.1.4 Nucleation of Ice from Water Vapor
Nucleation of ice from water vapor is relatively less understood compared to nucleation of ice from bulk liquid water despite more direct relevance to cloud seeding. A
two-step mechanism of (1) condensation of water vapor followed by (2) freezing of
the condensed water [16] and a three-step mechanism of (1) condensation of water
vapor which was followed by (2) freezing of the condensed water which in turn was
followed by (3) growth of bulk ice from the ice nucleus that requires surmounting of
an additional activation barrier [17] have been proposed.
115
than half a century since the initial report, the underlying mechanisms are still under
active investigations.
An issue that further complicates the matter is an impact of surface roughness
on the heterogeneous nucleation potency of a material. Purely geometrical argument shows that surface roughness enhances heterogeneous nucleation potency of a
substrate [13]. Nanoscale roughness in the form of etch pits on a substrate surface
can have a profound effect on the kinetics of heterogeneous nucleation of ice [9].
Another surprising finding is that, in nano-porous alumina, liquid water subcools
to −42 °C and then freezes to metastable I c , not to thermodynamically stable I h [14].
I c thus formed remains stable during hearing, up to 273 K at which point it melts
to liquid water [14]. How could this happen? Nano-porous alumina had one of the
highest surface-to-volume ratios of any system studied to date and heterogeneous
nucleation of ice ought to be expected more than ever, and yet liquid water confined
inside nano-porous alumina subcooled to −42 °C. Of course, if the pores were
so small that the space left was insufficient for a long-range positional order (that
defines a crystalline phase) to manifest then no freezing of water would have been
expected. However, the pores in the nano-porous alumina were not that small; indeed,
a metastable I c crystalline phase could still form inside the nano-porous alumina [14].
These are all important outstanding issues that have not been resolved to date.
The so-called “Contact nucleation” refers to a nucleation process of ice on a solid
wall at an air–liquid interface (like around the three-phase lines of a meniscus). Such
contact nucleation has been found to be more efficient than “immersion nucleation”
(nucleation of ice on a solid wall that is totally immersed in liquid water) [1]. In
the presence of a potent heterogeneous nucleation substrate, surface nucleation preferentially occurred at the three-phase lines where the air, liquid water, and a solid
wall met. It was also found that the nanoscopic (nanometer scale) surface roughness
played a significant role while the microscopic (micrometer scale) surface roughness did not [1]. It has further been shown that rough substrates enhanced contact
nucleation while smooth ones did not [1]. It appears that steps, pores, cracks, or
other surface features of the sizes of the order of the critical nucleus may effectively
promote nucleation by lowering the free energy barrier [1, 15]. These findings are
most likely relevant to clathrate hydrate nucleation as we will see in the next section.
5.1.4 Nucleation of Ice from Water Vapor
Nucleation of ice from water vapor is relatively less understood compared to nucleation of ice from bulk liquid water despite more direct relevance to cloud seeding. A
two-step mechanism of (1) condensation of water vapor followed by (2) freezing of
the condensed water [16] and a three-step mechanism of (1) condensation of water
vapor which was followed by (2) freezing of the condensed water which in turn was
followed by (3) growth of bulk ice from the ice nucleus that requires surmounting of
an additional activation barrier [17] have been proposed.