NUCLEATION AND GROWTH OF ICE CRYSTALS IN CONCENTRATED-SOLUTIONS OF ETHYLENE-GLYCOL

Ice formation in concentrated solutions of ethylene glycol (EG; 48 to 53 wt%) was studied after various thermal histories. To characterize the rate of ice nucleation as a function of temperature (to -150 degrees C), the solutions were cooled at a constant rate to the nadir temperature and then warmed to a storage temperature (T-s) at which ice formation was studied under isothermal conditions using differential scanning calorimetry (DSC) and cryomicroscopy. The DSC data were analyzed using the Johnson-Mehl-Avrami (JMA) model of isothermal transformation. The model contains two parameters: a kinetic constant K and a reaction-order parameter n. The kinetic constant was used to quantify the relative stability of the amorphous vs frozen state and rates of ice nucleation and growth in highly supercooled solutions of EG. Our theoretical analysis suggests that the kinetic constant is proportional to the diffusion coefficient D of water molecules and to the 2/3 power of the concentration of ice nuclei that form in a solution prior to storage at T-s. Experimentally, the kinetic constant is shown to be the product of two values, the first of which is dependent on the nadir temperature and the second on storage temperature. Increases in the kinetic constant corresponding to two nucleation events occurred with decreasing nadir temperature. During cooling the second ice nucleation event began at a temperature approximately 20 degrees C lower than that at which the first nucleation event was completed. For example, in solutions containing 50 to 51.5 wt% EG, the first nucleation event occurred at temperatures between -80 and -100 degrees C and the second nucleation event occurred at temperatures between - 120 and - 135 degrees C, which is in the vicinity of the glass formation temperature (T-g). Cryomicroscopic measurements of the spherulite concentration confirmed the appearance of the first and second nucleation events in EG solutions. DSC studies showed that the total amount of nucleation increased approximately 20-fold if the concentration of EG was decreased from 51 to 48 wt%, Only about 10% of this increase was associated with the first nucleation event and the remainder was the result of the second nucleation event. Furthermore, the proportion of nucleation attributable to the first and second nucleation events was dependent on the EG concentration. At concentrations greater than or equal to 51 wt% EG, the majority of ice nuclei formed during the first nucleation event. However, in a 48 wt% EG solution the kinetic constant measured after cooling to T-n greater than or equal to - 100 degrees C, when only the first nucleation event occurred, was about 16 times smaller than that measured after cooling to T-n = - 150 degrees C, when both the first and second nucleation events occurred. For this reason, the second nucleation event is the principal cause of the decrease in stability of the amorphous state in EG solutions in which the concentration is decreased from 51 to 48 wt%. Therefore, the minimum concentration of cryoprotectant needed for vitrification at a given cooling and warming rate can be substantially reduced if the rate of ice nucleation can be decreased. Addition of bovine serum albumin (BSA) decreased the value of the kinetic constant determined for concentrated solutions of EG; however, the effect of BSA was small in comparison to that which could be effected by a small increase in the concentration of EG. For example, the decrease in the nucleation parameter, which was affected by addition of 6 g of BSA to 100 g of a 48 wt% EG solution, could also be achieved by increasing the EG concentration by 0.4wt%. (C) 1995 Academic Press, Inc.