Three Macroscopic Antifreeze Properties

In the last section we discussed some of the crystal growth behaviours of ice in the absence of additives. In the presence of additives, crystal ice growth behaves di↵erently. In this section we present three macroscopic properties of antifreeze molecules, which we will later use to characterize the antifreeze behaviour of our polymers and compare to what has already been done in experiments. The three properties are ice recrystallization inhibition (IRI), thermal hysteresis (TH) and

dynamic ice shaping (DIS).

Ice recrystallization inhibition

During thawing or storage in a freezer, larger ice crystals normally grow at the expense of neighbouring smaller ones in order to reduce the interfacial energy of the system. This process is known as ice recrystallization and is often likened to a form of Ostwald ripening. Antifreeze molecules that inhibit or slow this process give

rise to the term ice IRI.[3,46]This particular antifreeze property is thought to be the

most useful of the three, particularly for cryopreservation, because it can potentially reduce ice crystals to manageable sizes.

Ice recrystallisation is a↵ected by surface adsorbing polymers or proteins, which interfere with the movement of water molecules within the ice boundaries,

e.g. ice surface modifications (e.g. during surface adsorption), disruption of wa-

ter di↵usion rates and manipulation of liquid channel thickness between the ice grains.[3,46,47] Because even non-ice-adsorbing macromolecules are sometimes able to manipulate water di↵usion rates, the inclusion of small solutes like salt or sugars are often used to di↵erentiate between adsorbing and non-adsorbing polymers whilst testing for IRI. The reason for this is that salts like NaCl are able to increase the size of the liquid channels, counteracting any non-specific, retarding e↵ects on water mobility.[3,46–48]

IRI is widely investigated using the “splat” assay first developed by Knight

et al[22]_{, although other techniques have also been developed.}[49] _{The set-up of a}

splat assay is shown in Fig 1.2. In this test a syringe containing the solution of inter- est is suspended approximately 2 m above a 193.15 K chilled glass plate and a small drop is released. On immediate contact the drop is quickly frozen, and annealed at a warmer temperature close to 267.15 K, yielding a polycrystalline wafer after 30 minutes. A high-resolution camera is used to continuously monitor this droplet and its crystals from start to finish. At the end of the experiment, image analysis tools like ImageJ[50] _{are used to determine the mean ice grain size (MIGS) or the mean} largest grain size (MLGS) from a number of randomly selected or largest crystals in the field of view respectively. The results are then compared to a negative con- trol, typically a phosphate-bu↵ered saline (PBS) solution, which serves as a solvent for the most part. The final results are normally reported as a percentage of the negative control,[1,7,8,22,51,52]_{and sometimes by quoting the lowest concentration at} which IRI activity is observed.

Figure 1.2: Splat assay set up. A. syringe, B. distance, C. chilled glass plate, D. cooling aluminium plate, E. stand, F. imaging system G. schematic of micrograph.

Thermal hysteresis

TH is possibly the most studied of the three antifreeze properties. Ordinarily the melting and freezing temperature of ice crystals are one and the same. In the presence of antifreeze molecules this might change: thermal hysteresis describes the scenario when the freezing temperature falls beneath the melting temperature and the di↵erence between the two is called the TH gap. TH is thought to be caused

by direct and irreversible adsorption onto specific planes of a growing ice crystal

surface.[1] As a result of direct pinning onto the surface, the local area becomes microscopically curved and energetically unfavorable to grow any further (Fig. 1.3). This is known as the Kelvin e↵ect or sometimes Gibbs-Thomson e↵ect.[53]_{The size} of the TH gap varies from molecule to molecule, however at temperatures within the TH region, ice crystals does not grow or shrink.[3] _{In the literature, TH is usu-} ally described as non-colligative, which means that TH antifreeze molecules does not lower the freezing point in proportion to their concentrations.[54] _{Rather, they} typically work more e↵ectively than expected at lower concentrations.

In order to assess this antifreeze property, a nanolitre osmometer is also used to accurately monitor the temperature and ice crystal size at time intervals.[3] Al- ternatively a di↵erential scanning calorimeter can be used to automate the process, without the need for microscopic observation.[55]

Figure 1.3: A schematic illustrating the accumulation of water (green) onto the ice surface (blue) in the presence of antifreeze molecules (white). The blue arrow highlights this process between adjacent antifreeze molecules. A. antifreeze molecule, B. growing ice front, C. accumulating water molecules.

Dynamic ice shaping

As mentioned earlier, ice crystals typically grow out into spherical shapes at tem-

peratures close to the Tm (up to∆T⇡-2 K). However in the presence of antifreeze

molecules the ice crystal shape may become altered as it grows within the TH region

(or at temperatures close to the Tm), because certain antifreeze molecules have the

ability to interact closely with certain planes of ice.[3] _{The name for this process is}

dynamic ice shaping (DIS), and it can be monitored using a range of techniques,

mainly hemispheral etching and by use of a nanolitre osmometer.[3]