A statistical approach: understanding the variation in induction time to measure nucleation rates
Tutor / director / evaluatorHorst, Joop ter
Document typeBachelor thesis
Rights accessRestricted access - author's decision
Crystal nucleation is the first and most important step in crystallization process as it has strong impact on crystal product quality. An easy, reliable and accurate way to measure nucleation rate in stirred solutions is the induction time measurement method. At small scale for many systems the method makes use of variations in measurements which are due to stochastic nature of nucleation and measure nucleation rates based on probability distribution of induction times. The variations could also be because of errors in manual sampling or the analytical technique used. A statistical analysis of experimental data of induction time measurements was performed to identify experimental errors. This includes observing the variations due to different factors such as average levels of supersaturation, position in the setup, operating temperatures, etc. which can affect to the results of the process. Nicotinamide (NA) in ethanol was used as model system for performing the experiments of induction time measurements under different supersaturation conditions. It was detected that the position where the samples are hold in the setup has no significant influence on the nucleation process. It was also noticed a statistically significant decrease in the probability distribution so as the nucleation rate of NA with each repetition of experiments of the same samples. It was verified that there are no changes in the saturation temperature, which indicates that there is no degradation of the compound. A possible explanation of the behaviour of nucleation rates could therefore be the presence of active particles responsible for heterogeneous nucleation which deactivate with the cycles. Actually, the low determined thermodynamic parameter show that heterogeneous nucleation rather that homogeneous nucleation occurs. Moreover, samples of NA in ethanol were heated to different maximum temperatures like 50ºC, 60ºC and 70ºC and cooled to 25ºC to measure induction times. It was observed that the heating temperature does significantly affect to the nucleation process, as the nucleation rate J decreases with increase in the heating temperature. This also support the presence of impurities in the system. 4-hydroxyacetophenone (HAP) in ethyl acetate was also used to perform induction time measurements, and there were no significant differences in the experimental nucleation rate results between the repetitions of the experiments of the same samples. It can therefore be stated that the phenomenon is only system dependent which was observed in case of NA in ethanol.