Continuous-flow, Well-mixed, Microfluidic Devices for Screening of Crystalline Materials
thesisposted on 01.12.2021, 00:00 by Paria Coliaie
Screening crystal polymorphs and morphology and measuring crystallization kinetics in a controlled supersaturated environment is crucial for developing crystallization processes for pharmaceuticals, agrochemicals, semiconductors, catalysts, and other specialty chemicals. The current tools, such as microtiter plates and droplet-based microfluidic devices, are susceptible to supersaturation depletion in small compartments due to crystallization and nucleation. Such variation in supersaturation affects the outcome and leads to impediments during the scale-up of the crystallizer. Here we develop an innovative technique using cyclone mixer designs to study antisolvent crystallization at constant supersaturation maintained by a continuous solution flow. The developed continuous-flow microfluidic mixer device is used to analyze the polymorphs and morphology of o-aminobenzoic acid (o-ABA) at different supersaturations. A comparative study is done with a microtiter plate assay to demonstrate the effect of the supersaturation depletion on the polymorphs and morphology in the batch screening systems. To further extend the throughput scale of the continuous-flow screening, a multi-well microfluidic device is proposed for parallel continuous-flow screening. A flow-controlled microfluidic device for parallel and combinatorial screening of crystalline materials can profoundly impact the discovery and development of active pharmaceutical ingredients and other crystalline materials. While the existing continuous-flow microfluidic devices allow crystals to nucleate under controlled conditions in the channels, their growth consumes solute from the solution leading to variation in the downstream composition. The materials screened under such varying conditions are not quite reproducible in large-scale synthesis. There exists no continuous-flow microfluidic device that traps and grows crystals under controlled conditions for parallel screening. Here, we illustrate the blueprint of such a microfluidic device that has parallel-connected micromixers to trap and grow crystals under multiple conditions simultaneously. The efficacy of a multi-well microfluidic device is demonstrated to screen polymorphs, morphology, and growth rates of L- histidine via antisolvent crystallization at eight different solution conditions, including variation in molar concentration, vol% of ethanol, and supersaturation. The developed continuous-flow microfluidic device is then modified to be implemented for cooling crystallization. Cooling crystallization is a well-established technique for developing active pharmaceutical ingredients, driven by the significant solubility difference of active pharmaceutical ingredients (APIs) at higher and lower temperatures. The biggest challenge of continuous cooling crystallization is growing and trapping crystals under controlled temperatures. In the existing continuous-flow microfluidic devices, either the temperature control is not precise, and a temperature gradient exists during the crystal growth, or the crystal growth consumes solute from the solution and leads to variation in the downstream composition. The obtained kinetic data from such varying conditions are not reliable for large-scale synthesis. To overcome this issue, two temperature control strategies are incorporated within the developed continuous-flow microfluidic device to enable continuous-flow cooling crystallization under controlled conditions. The first strategy is implementing a cooling jacket around the micromixer to manipulate the temperature of the entering streams and generate the necessary temperature gradient and driving force for nucleation. In the second strategy, saturated hot and cold streams of API are pumped into the micromixer, and the temperature of the mixture will be lower than the hot saturated stream. The efficacy of these strategies is compared by measuring the equilibrium temperature profile from computational modeling analysis. Additionally, the feasibility of the two temperature control strategies is demonstrated with screening polymorphs, morphology, and growth rates of cooling crystallization of L-glutamic acid. Moving toward a higher-throughput screening system, we develop a snap-on adaptor that converts every well in a conventional batch screening assay (microtiter plates) into a continuous-flow microfluidic device. Microtiter plate assay is a conventional and standard tool for high-throughput (HT) screening that allows growing, harvesting, and screening crystals. The microtiter plate screening assays require a small amount of APIs in each experiment which is adequate for a solid- state crystal analysis such as X-ray diffraction (XRD) or Raman spectroscopy. Despite the advantages of these high throughput assays, the batch operational nature results in a continuous decrease in the supersaturation during the screening procedure due to crystal nucleation and growth. To benefit from the advantages of these HT screening tools while avoiding the batch type of operations, we developed a snap-on adapter that converts a microtiter plate assay into a continuous-flow microfluidic device. The developed snap-on adaptor is plugged in a 24-well plate assay and implemented for salt screening of naproxen. Microfluidic sensors are integrated into the developed continuous-flow microfluidic device for extending its application. Sensors are significant assets for quick and sensitive detection inside miniaturized devices where the sample is small. The incorporation of sensors into continuous-flow microfluidic systems allows for precise control of experimental conditions. One of the potential applications of a sensor-integrated microfluidic system is to measure the concentration of the solution. In this study, we have paired the previously developed continuous-flow microfluidic mixer with an electrochemical sensor to enable in-situ measurement of the supersaturation during an antisolvent crystallization of L-histidine in a water-ethanol mixture. The sensitivity of different metals towards L-histidine is measured in a batch system to choose the best material for the sensor. The chosen metal is subsequently used as the working electrode of a three-electrode surface printed sensor attached below the continuous-flow microfluidic device. The sensor is externally connected to a potentiostat device, and anodic cyclic voltammetry slopes are measured. The sensor is calibrated for different ratios of antisolvent and concentrations of L-histidine. Additionally, machine learning algorithms were implemented to develop a supersaturation prediction model from the experimental data. One of the other sensors integrated within the microfluidic platform is a photosensitive turbidity sensor that can identify the liquid-liquid phase separation (LLPS) boundaries. This phenomenon is also known as oiling-out, which is the appearance of a droplet in a supersaturated solution. Oiling-out is a common phenomenon among hydrophobic pharmaceutical compounds. However, the spontaneous nature of the transformation of oil droplets into solids makes LLPS very hard to control. As a result, it is considered an undesired phenomenon and is typically avoided during crystallization. Consequently, the most critical step is to obtain the phase diagram to ensure the crystallization does not occur within the LLPS zone. The developed microfluidic mixer is coupled with an Arduino chip, a photoresist, and an LED light to detect the formation of the oil droplets within the mixture. This sensor is implemented to detect the LLPS boundaries of the beta-alanine in water and IPA. The continuous-flow microfluidic device is proposed to be integrated with different sonication tools to be integrated with a prediction model for breakage. An innovative solution to the breakage population balance problem is developed in the final section of this study. Milling or attrition of particles is an important part of many applications, and there is not much of a comprehensive understanding of the mechanisms of this process, and there is no refined analytical approach for the simple or complex kernels. Through predictive modeling, we demonstrate an innovative approach to understanding milling mechanics.