INCRYS

Crystallization is a separation and purification process widely used in industry for drug synthesis, purification of battery-grade metals, food manufacturing, etc. Although the technique is mature, industrial crystallization still faces two main challenges: (1) controlling the crystal size distribution (CSD) to improve the efficiency of downstream processes (filtration, washing, drying), and (2) controlling the solid phase in order to optimize the physico-chemical properties of the product –e.g. solubility of active pharmaceutical ingredient (API). A better control of these features is hindered by undesired nucleation in the crystallization reactor. Nucleation is a stochastic phenomenon producing solid crystalline seeds from a supersaturated solution. In the development of crystallization processes, scale-up is often an issue because solution stirring (required for homogeneity) induces shear stress that strongly increases the nucleation rate [1] and thus prevents accurate CSD control. Industry circumvents the nucleation issue by seeding crystallizers with a specific amount of calibrated crystalline seeds at low supersaturation ratio in order to prevent undesired spontaneous nucleation. Although this technique is satisfying, it can be constraining if crystallization is performed in a sterile environment, or if the seed material is not available. Therefore, there is a need for in situ, on-demand and scalable seeding techniques allowing a high level of polymorph and size distribution controls. Alternative techniques based on external fields are currently explored but their scalability is still an issue - e.g. sonocrystallization, laser-induced nucleation [2]. The INCRYS project aims to use cold plasma generated by nanosecond discharges for seeding crystallization reactors in a non-intrusive manner, in situ, and on-demand. INCRYS will pave the way towards this potential use of plasma by addressing three main objectives: 1. Identification of the plasma-induced nucleation mechanism by measuring (i) the nucleation dynamics and (ii) the dynamics of well-known plasma nucleation agents active at specific timescales during and after the nanosecond discharge. Assuming that nucleation is synchronous with the driving plasma nucleation agents, the mechanism will be inferred. 2. Comparison of shear- and plasma-induced nucleation by measuring nucleation rates in agitated reactors. As a general trend, because the nucleation rate increases with shear stress, we expect that nucleation is driven by the plasma at low shear rates. This range of shear rates defines the operating conditions of the plasma inside stirred crystallization reactors. 3. Scalability of the nucleation method by seeding a 100 mL batch-stirred crystallization reactor. We expect that the seed loading induced by the plasma can be adjusted to have the desired effects on CSD. The objectives will be addressed using a unique microfluidic plasma reactor that will be exploited throughout the project, with a lead halide perovskite as a model compound for the nucleation studies. A consortium of partners with complementary expertise from CentraleSupélec, Ecole Normale Supérieure Paris-Saclay and CNRS will collaborate to reach the objectives. The output of the project will be a microfluidic plasma-seeding reactor and a method to implement it on a batch-stirred reactor. The experimental approach developed in the project will be applied to perovskite crystallization, but can be transposed to the crystallization of any other compounds of interest, thus opening promising perspectives for application in research and industry. [1] C. Forsyth et al., 2015, Cryst. Growth Des., 15, 94–102, https://doi.org/10.1021/cg5008878 [2] B. Clair et al., J. App. Cryst., 2014, 47: 1252. https://doi.org/10.1107/S160057671401098X