(521d) Slug Flow Reactor Simulation for Controllable Residence Time Distribution Using Computational Fluid Dynamics.

In the pharmaceutical industry, the process has traditionally operated in batch type.¹ However, the batch process is not easy to control, which can be challenging to achieve product consistency. Also, batch crystallization has been existed the problems of inefficiency during the process and batch-to-batch variability.² Compared to the batch crystallizer, continuous crystallization processes have been expected to improve product quality consistency and manufacturing process efficiency as well as lower capital and production costs.^3-5 Accordingly, many continuous crystallizer designs have been suggested with the aim of controlling the product specification and generating uniform crystals.⁶ Among them, slug flow crystallizer does not require additional equipment (e.g., mixer) and reduces common issues (e.g., fouling, clogging) in tubular reactors. For scaling up, the slug flow crystallizer only needs to increase the number of slugs by using longer tubes with the same equipment.⁶ Many studies of residence time distribution (RTD) have been conducted on liquid-liquid and liquid-gas slug flow according to the slug size and flow rate.^7,8 However, there are few studies on the effect of RTD due to diffusion through the liquid film, which is also generated depending on the surface tension of the liquid used.

Here, we implemented a gas-liquid slug flow with thin liquid film using computational fluid dynamics (CFD) simulation with multi-phase consideration. Through simulation, the liquid film phenomenon that is generated in a solvent with low surface tension and affects RTD was also considered. Our CFD model was applied the Volume of Fluid (VOF) method that is applicable to form an interface between multiphase flows. The proposed CFD model was verified under the same conditions as the actual experiment. For the reactor, a silicone tubing (Masferflex Transfer Tubing, C-Flex) with an inner diameter of 2.38 mm was used. To capture the moving slugs for validation, a digital single-lens reflex (DSLR) camera (Nikon D5600) was placed 10 cm after the T-junction. CFD model predicted the shape of slug and liquid film accurately. We first discussed the effects of slug flow and liquid film on RTD. Obviously, the slug flow had a narrower RTD than the non-slug flow (one phase laminar flow), and when the thin liquid film is present in the slug flow, it can be confirmed that the RTD spreads slightly due to the diffusion phenomenon through the thin film between each slug. In addition, according to the difference in surface tension, RTD sensitivity analysis was carried out by setting the flow rate as variables for isopropanol and water. The higher the flow rate, the smaller the variance of RTD, but the slug formation was unstable and broken. This phenomenon was observed in the isopropanol solvent in which the film was formed rather than in the water in which the film was not formed, along with the uncertainty of the RTD. Accordingly, we finally performed a comparison of RTD variability according to solvents with various surface tensions. As a result, it was found that the lower the surface tension, the greater the diffusion by the film, and the larger the RTD variance. Through this study, we analyzed the slug flow crystallizer as RTD using CFD simulation. In order to be possible, the controllable RTD, which is considered a benefit of the slug flow, each slug should act as an independent batch rector. The thin film should be minimized by using a solvent with the large surface tension as possible and that an appropriate flow rate should be used for stable slug formation.

¹ Paul, E. L.; Tung, H.-H.; Midler, M. Organic Crystallization Processes. Powder Technol. 2005, 150, 133–143.

² Zhang, D., Xu, S., Du, S., Wang, J., & Gong, J. (2017). Progress of pharmaceutical continuous crystallization. Engineering, 3(3), 354-364.

³ Alvarez, A. J.; Myerson, A. S. Continuous Plug Flow Crystallization of Pharmaceutical Compounds. Cryst. Growth Des. 2010, 10 (5), 2219−2228.

⁴ Lawton, S.; Steele, G.; Shering, P.; Zhao, L.; Laird, I.; Ni, X. Continuous Crystallization of Pharmaceuticals using a Continuous Oscillatory Baffled Crystallizer. Org. Process Res. Dev. 2009, 13 (6), 1357−1363.

⁵ Ferguson, S.; Morris, G.; Hao, H.; Barrett, M.; Glennon, B. InSitu Monitoring and Characterization of Plug Flow Crystallizers. Chem. Eng. Sci. 2012, 77, 105−111.

⁶ Jiang, Mo, and Richard D. Braatz. "Designs of continuous-flow pharmaceutical crystallizers: developments and practice." CrystEngComm 21.23 (2019): 3534-3551.

⁷ Qian, D., & Lawal, A. (2006). Numerical study on gas and liquid slugs for Taylor flow in a T-junction microchannel. Chemical Engineering Science, 61(23), 7609-7625.

⁸ Günther, A., Khan, S. A., Thiemann, M., Trachsel, F., & Jensen, K. F. (2004). Transport and reaction in microscale segmented gas–liquid flow. Lab on a Chip, 4(4), 278-286.