Abstract
We characterize shapes and volumes of droplets generated in PDMS T-junctions and assess the use of this type of microfluidic device to generate droplets suitable for the study of nucleation. Water droplets were formed in oil in a PDMS T-junction and subsequently stored. Droplet volume reproducibility and stability were investigated from acquired micrographs. By theoretically analyzing the influence of the mean volume of a population of droplets on the estimation of nucleation rates, we have shown that deviations in mean volumes can seriously affect the estimates, unless such deviation is smaller than 10%. This condition is fulfilled if experiments are repeated using the same microdevice. Measured droplet polydispersity remained low enough to treat the droplets as monodisperse. Immersing the microdevice in a water bath mitigates solvent evaporation, and allows for very accurate temperature control. Finally, a screening procedure was used to select the ideal operating conditions to obtain droplets with the desired sizes. Applying this method in devices with increasing T-junction cross sectional area, we have demonstrated a scaling-up of droplet volumes close to an order of magnitude while tuning the droplet shape, i.e., the average length to width ratio, at values between 1 and 1.2.
Highlights
Nucleation is considered the most critical step in crystallization processes, because it impacts several crystal properties, including crystal size distribution and polymorphism
We develop a theoretical analysis of how the distribution of droplet volumes affects the nucleation process and the estimate of the nucleation rate
In this work we have assessed the reproducibility and stability of droplet formation and storage in PDMS microdevices in view of carrying out nucleation experiments in such droplets in a follow-up study
Summary
Nucleation is considered the most critical step in crystallization processes, because it impacts several crystal properties, including crystal size distribution and polymorphism. Nucleation is known to have a stochastic nature (Kulkarni et al, 2013; Goh et al, 2010; Kubota, 2015) This characteristic poses challenges to obtaining reproducible experimental data, and numerous identical experiments are essential to correctly build probability distributions for studying the stochastic behaviour of primary nucleation. Besides carrying out numerous experiments, reproducibility of volumes is fundamental to avoid introducing further variability in the probability distributions associated with the repeated nucleation measurements. If this is achieved, robust statistics can be obtained and nucleation mechanism and kinetics can reliably be determined
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
