Ultrasound-assisted continuous crystallization of metastable polymorphic pharmaceutical in a slug-flow tubular crystallizer

Highlights

• •Prepared metastable CVD form Ⅱ crystals using a tubular crystallizer continuously.
• •Optimized the crystallization process of the tubular crystallizer using air bubble segments and ultrasonic irradiation strategies.
• •Two strategies were used to prepare CVD metastable form II crystals: tubular and batch crystallizers.
• •Solved the phenomenon of clogging and wall sticking in a tubular crystallizer.
• •Effectively circumvented the issue of polymorphic transformation during the CVD production process.

Abstract

Metastable polymorphic pharmaceuticals have garnered significant attention in recent years due to their enhanced physicochemical properties, including solubility, bioavailability, and intellectual property considerations. However, the manufacturing of metastable form pharmaceuticals remains a formidable challenge. The stable preparation of metastable carvedilol (CVD) form Ⅱ crystals during CVD production is elusive, leading to substantial inconsistencies in product quality and regulatory compliance. In this study, we successfully prepared metastable CVD Form Ⅱ crystals using a continuous tubular crystallizer. Our findings demonstrate that the tubular crystallizer exhibits high efficiency and robustness for generating metastable crystal Form Ⅱ. We optimized the crystallization process by incorporating air bubble segments and employing ultrasonic irradiation strategies to overcome blockages and wall sticking issues encountered during operation. Ultimately, we developed an ultrasound-assisted continuous slug-flow tubular crystallization method and evaluated its performance. The results indicate that the CVD crystals produced through this process are resilient, sustainable, and uninterrupted products with promising potential for producing metastable polymorphic pharmaceuticals while effectively addressing encrustation problems associated with continuous tubular crystallization.

1. Introduction

Developing metastable crystals for active pharmaceutical ingredients is an effective method for improving the bioavailability of insoluble drugs. However, metastable crystals are characterized by their thermodynamic instability, which increases the difficulty of utilizing metastable crystal drugs [1]. Thermodynamic instability leads to problems with polymorphic transformation during the preparation of metastable crystals. Metastable crystals can improve the physicochemical properties of drugs, and a metastable crystal form often has a higher solubility than its stable crystal form. The mechanical and biological properties of the two also differ[2]. Adjusting the crystal form of the drug can control the solubility and stability, thus affecting the shelf life of the drug preparation. Metastable crystalline drugs can be developed as new intellectual property rights. Wilson et al.[3] studied metastable polymorphs of paracetamol with high purity by continuous crystallization. The prepared metastable paracetamol crystal form II has more substantial solubility and compressibility than stable crystal form I. To avoid the polymorph transformation and gelation of Clopidogrel hydrogen, Gong et al.[4] designed a Clopidogrel hydrogen crystallization process that could prepare high-quality spherulites of form I. The prepared Clopidogrel hydrogen metastable form I has higher solubility and bioavailability.

The metastable polycrystalline drug crystallization process has attracted overt attention in recent years. Fang et al.[5] prepared metastable form III of aripiprazole using ultrasound-assisted and continuously controlled air slug flow with a tubular cooling crystallization method. Lauren et al.[3] prepared metastable form II of paracetamol using a continuous oscillation baffle crystallizer and mixed suspension mixed product removal (MSMPR) system. Tsai-Ta et al.[6] studied the relationship between the crystal pattern dynamics of L-glutamic acid, residence time, and temperature during the crystallization of mixed product removal from continuously mixed suspensions. We could control temperature and residence time by selectively generating metastable or stable crystals in an MSMPR mold. We propose a method of rapidly preparing metastable CVD form II crystals by tubular crystallizer. As two operating forms of industrial crystallization, continuous crystallization has higher efficiency and an easier scale-up feasibility than the batch crystallization process [7], [8], [9], [10]. Due to its high efficiency and flexibility, continuous crystallization is becoming increasingly popular in medicine production [11], [12], [13], [14]. Tubular crystallizers in continuous crystallization have captured the attention of the pharmaceutical industry because they allow precise control of mean particle size, particle size distribution, and polymorphism [15], [16]. A plug flow can be approximated in tubular crystallizers, where crystals have a narrower residence time distribution and are thus easier to scale up. However, tubular crystallizers have disadvantages such as scaling blockage and high equipment maintenance cost [17], [18]. Therefore, some process-strengthening methods would improve continuous tubular crystallization’s economic benefit and product quality. Ultrasound strongly affects the tubular crystallization process. In recent years, ultrasound-assisted technology has been widely used in the pharmaceutical industry [19], [20]. Savvopoulos et al. [21] proposed and evaluated a theoretical feedback control scheme for ultrasound-assisted aspirin continuous anti-solvent crystallization in a tubular crystallizer. They determined that ultrasound-assisted crystallization was an effective method for preparing crystals with size distribution width [22], [23], [24]. The induction time and metastable zone width (MSZW) of the batch crystallization and flow system could be effectively shortened by ultrasound irradiation during the continuous crystallization of drugs [25]. Ultrasound vibrations can remove particles from the wall surface, reduce particle agglomeration and crystal breakage and optimize crystal habits to avoid clogging caused by tubular crystallizers. Ultrasound has been one of the essential tools for controlling the nucleation of continuous tubular crystallizers[26]. Ultrasound-assisted tubular crystallization cleans tubes, prevents plugging, and improves the quality of crystal products [25], [27], [28], [29], [30].

In drug crystallization production, crystal polymorphism transformation causes significant losses to pharmaceutical manufacturers. Pataki et al. [31] studied the solvent-mediated crystallization of CVD using real-time Raman spectroscopy. During the phase transition mediated by the ethyl acetate solvent, the kinetically preferred form II was transformed into the thermodynamically stable form I [32]. During the production process, CVD metastable crystals adopt traditional intermittent crystallizer cooling crystallization methods. Therefore, CVD metastable crystals cannot be repeatedly and stably generated, reducing their production efficiency.

We selected carvedilol CVD 1-(9H-carbozol-4-ryloxy)-3-[[2-(2-methoxy phenoxy)ethyl]amino]-2-propanol; Fig. 1) as the model compound, which is used to clinically treat congestive heart failure[33], hypertension[34], and myocardial infarction[35]. CVD is a third-generation non-selective beta and alpha-1 blocker belonging to BCS class II APIs [34], [35], [36], [37], [38], as shown in Fig. 1. CVD has three solvates (V, VI, III) and five polycrystalline forms (I, II, IV, VII, IX).

Fig. 1.

Chemical structure of CVD − 1-(9H-carbozol-4-ryloxy)-3-[[2-(2-methoxy phenoxy)ethyl]amino]-2-propanol.

A series of batch cooling crystallization experiments was conducted as a control to address the challenges associated with polymorphic transformation in the manufacturing process. The tubular crystallizer was improved for clogging and tube sticking to achieve stable and continuous preparation of CVD metastable crystal Form Ⅱ. Additionally, air bubble- and ultrasound-assisted experiments were performed. The utilization of the continuous slug flow tube technique facilitated robust, sustainable, and uninterrupted crystallization of CVD products. The CVD product prepared using the novel approach offers a convenient process methodology for practical pharmaceutical production, effectively preventing the issue of polymorphic transformation during CVD production and facilitating scale-up manufacturing.

2. Materials and methods

2.1. Materials

CVD (molecular formula: C₂₄H₂₆N₂O₄, purity > 99.6 %) was provided by Zhejiang Huahai Pharmaceutical Co., Ltd., China. Ethyl acetate, deionized water, and all organic solvents (analytical reagent grade) were purchased from Tianjin Jiangtian Chemical Co., Ltd., China.

2.2. Batch crystallization apparatus process and monitoring

2.2.1. Batch crystallization apparatus process

As shown in Fig. 2, the double-jacketed crystallizer has a height-diameter ratio of 4:5 and the effective volume is 50 mL. The temperature in the crystallizer was regulated by a jacketed system using the circulating medium (CF41, Julabo, Germany). The stirring speed was controlled by a DC motor and stirring controller.

Fig. 2.

The batch experimental apparatus of the CVD solution crystallization process includes a crystallizer, stirring, feeding, and temperature control systems.

The crystallization point of CVD in ethyl acetate was 320.15 ± 0.02 k by multiple experiments without adding seed. We obtained the CVD solution by dissolving CVD in ethyl acetate solution at 347.15 ± 0.02 K. The dissolution time was 15 min, and the CVD concentration was 0.143 g/mL. The temperature was cooled to 320.15 ± 0.02 K and kept for 4 h. Then, we conducted a gradient cooling process to decrease the temperature to 270.15 ± 0.02 K for 2 h. The batch crystallization process experiment was conducted at least six times for consistency.

2.2.2. Batch crystallization apparatus monitoring with Raman spectrum online experiment

We prepared the CVD solution by dissolving CVD in ethyl acetate solution at 347.15 ± 0.02 K. The dissolution time was 15 min, and the CVD concentration was 0.143 g/mL after CVD was completely dissolved in the solution. The temperature was cooled to 320.15 ± 0.02 K and kept for 10 h. A Raman spectrometer (Raman RXN2, Kaiser Optical Systems Inc., USA) was used to track the CVD characteristic peaks at constant temperature processes of 320.15 ± 0.02 K and 290.15 ± 0.02 K, respectively, which helped reveal structural changes at the molecular level during the transition process. The Raman probe was immersed in the CVD solution, and the spectrum wavelength of the Raman probe is 785 nm. The Raman shift data were collected at intervals of 30 s using an exposure time of 5 s with six accumulations over a range of 100–3200 cm⁻¹ with a resolution of 1 cm⁻¹. The uncertainty of the Raman shift for the characteristic peak was 1 cm⁻¹. We performed the monitoring experiment at least six times with the same trend.

2.3. Tubular crystallization apparatus process

The preparation procedures of the CVD solution with the concentration of 0.143 g/ml were the same as with batch crystallization processes. The tubular crystallizer, as shown in Fig. 3, was immersed in an ultrasound water bath, and a copper tube with an inner diameter (ID) of 8 mm and an outer diameter (OD) of 10 mm was also immersed in an ultrasound water bath and connected to a thermostatic controller to control the temperature of the tubular crystallizer, thereby regulating the supersaturation of the crystallization process. The residence time of slurry in the process of tubular crystallization was controlled by adjusting the flow rate of the peristaltic pump. We conducted the tubular crystallization processes experiment at least six times for consistency.

Fig. 3.

The tubular experimental apparatus of the CVD solution crystallization process includes a tubular crystallizer system, ultrasound system, feeding system, and temperature control system.

As shown in Fig. 3, the tubular crystallizer is a custom-made glass U-tube with a 4 mm ID, 7 mm OD, and 110 cm length for continuous flow crystallization. The ultrasound water bath (Ningbo Xinzhi Biotechnology Co., Ltd. SB25-12DTD, China) was rebuilt so that the U-tubular crystallizer could fit inside the ultrasound irradiation chamber. The solution was pumped with a peristaltic pump (Longer, YZ1515X, BT-100-1F, China) using a primary silicone tubular with an ID of 3.2 mm and OD of 6.4 mm (Fig. 3(b)). In the slug flow crystallization experiments, we installed a secondary silicone tubular of 2 mm ID and 4 mm OD co-axially in the primary silicone tubular to supply air. For all sonicated experiments, the ultrasound electrical power was 450 W (±2 W) and the frequency was 40 kHz.

2.4. Slug-flow tubular crystallization with and without ultrasound irradiation

2.4.1. Air bubble-assisted tubular crystallization process

The calibrated two-channel peristaltic pump (Longer, YZ1515X, BT-100-1F, China) transmitted the bulk drug solution from the batch crystallizer to the tubular crystallizer through a silicone tubular with an ID of 2 mm and OD of 4 mm. Air was fed into the crystallizer through a peristaltic pump. The CVD solution was mixed with the air slug flow at the entrance of the tubular crystallizer using a polyethylene Y-form mixer. Therefore, the gas–liquid two-phase flowed through the Y-shaped mixer outlet into the tubular crystallizer, forming a stable slug flow.

2.4.2. Ultrasound-assisted continuous slug-flow tubular crystallization process

Air was fed into the crystallizer through a peristaltic pump. The CVD solution was mixed with the air slug flow at the entrance of the tubular crystallizer using a polyethylene Y-form mixer. The gas–liquid two-phase flowed through the Y-shaped mixer outlet into the tubular crystallizer, forming a stable slug flow. In this work, an ultrasound water bath (SB25-12 DTD, SCIENTZ, China) provided an ultrasound field for the tubular crystallization process. The frequency and power of the ultrasound water bath were set to 40 kHz and 564 W, respectively.

2.5. Characterization of CVD

The CVD was dried at 333.15 ± 0.5 K (inherent error ± 0.1 K) for 12 h in a vacuum oven until it reached a constant weight. The crystal products were also characterized and analyzed by a Raman spectrometer (Raman RXN2, Kaiser Optical Systems Inc., USA), differential scanning calorimetry (Mettler DSC 1, Mettler Toledo, Switzerland), powder X-ray diffraction (Model D/MAX 2500, Rigaku, Japan), and scanning electron microscopy (TM3000, Hitachi, Japan).

3. Results and discussion

3.1. Batch crystallization and polymorphic transformation monitoring

3.1.1. Monitoring results of batch crystallization CVD transformation at 323.15 K

Fig. 4 shows the real-time Raman spectrum of the CVD batch crystallization process at 323.15 K, which compares the real-time Raman spectrum of the CVD batch crystallization process for 0 min and two hours. The wave numbers of 664 cm⁻¹ and 1722 cm⁻¹ had significant changes. Fig. 5 shows the development of peak strength values, 664 cm⁻¹ and 1722 cm⁻¹, during CVD batch crystallization at 323.15 K. Under the constant temperature of 323.15 K for 20 min, the characteristic peaks of 664 cm⁻¹ and 1722 cm⁻¹ shifted, and 664 cm⁻¹ showed a decreasing trend with time, whereas 1722 cm⁻¹ showed an increasing trend. The wave number 664 cm⁻¹ corresponds to the two-carbon bond vibration in the CVD solution. The wave number 1722 cm⁻¹ corresponds to the change in the CVD solution’s aldehyde group. Fig. 4, Fig. 5 show that the CVD crystallization transformation occurs during the constant temperature crystallization process at 323.15 K and from 20 min onwards.

Fig. 4.

Real-time Raman spectrum during batch crystallization of CVD at 323.15 K.

Fig. 5.

Development of intensity values at peaks of 664 cm⁻¹ and 1722 cm⁻¹ during batch crystallization of CVD at 323.15 K.

3.1.2. Monitoring results of batch crystallization CVD transformation at 293.15 K

The real-time Raman spectrum of the CVD batch crystallization process at 293.15 K was shown in Fig. 6. As shown in Fig. 6, comparing the real-time Raman spectrum of the CVD batch crystallization process at 293.15 K at 0 min and 3 h 30 min, the wave number 1409 cm⁻¹ changed slightly. The development of the peak strength value of 1409 cm⁻¹ during the CVD batch crystallization at 293.15 K was shown in Fig. 7. Under the constant temperature of 293.15 k for 150 min, the characteristic peak of 1409 cm⁻¹ changed, and 1409 cm⁻¹ showed an increasing trend with the time change. The wave number 1409 cm⁻¹ corresponded to the vibration of nitrogen-containing aromatic groups in the CVD solution. It can be seen from Fig. 6, Fig. 7 that the crystallization transformation of CVD occurred during the 293.15 K constant temperature crystallization process, and from 150 min onwards.

Fig. 6.

Real-time Raman spectrum during batch crystallization of CVD at 293.15 K.

Fig. 7.

Development of intensity values for the peak at 1409 cm⁻¹ during batch crystallization of CVD at 293.15 K.

According to Fig. 4, Fig. 5, Fig. 6, Fig. 7, a polymorphic transformation of CVD was prone to occur during the traditional cooling crystallization process in the bath crystallizer to produce CVD crystal Form Ⅱ. In turn, the CVD metastable form II cannot be repeatedly and stably generated, resulting in low production efficiency, unstable product quality, and other industrial problems.

3.2. Tubular crystallization process

Since tubular crystallization can overcome the bath crystallization problem mentioned above, we used the tubular crystallization process strategy for metastable crystal form preparation. The benefit of moving from batch to continuous operation was a temperature and concentration distribution unaffected by time. This strategy gave the product a more stable and repeatable character and avoided inconsistency between batches. We proposed rapidly preparing metastable form II crystals by tubular crystallizers based on the intermittent crystallization method of CVD solution. Due to the short CVD induction time, it was suitable for cooling crystallization in tubular crystals, avoiding the high probability of tubular plugging caused by long tubular. A tubular crystallizing device was designed to continuously and stably prepare the pure metastable crystal form Ⅱ of CVD.

We set the initial temperature of the solution to 323.15 K and the ultrasound water bath to 293.15 K. We selected the optimum solution concentration within the operating range to avoid explosive nucleation and unsaturation. The optimized initial concentration was 0.1428 g/g for cooling crystallization. During the tubular crystallization process, the high flow rate ensured a low residence time, thus preventing the tubular crystallizer from clogging, and the flow rate should not be higher than 20 mL/min. Otherwise, the solution will not crystallize in the 100 cm U-tubular crystallizer. We selected an optimal flow rate was 15 mL/min and the residence time was 55 s.

3.3. Characterization of CVD in two crystallization strategies

Fig. 8, Fig. 9, Fig. 10 show CVD crystals characterization by microscope, SEM, XRD, and DSC. The comparison between the microscope and SEM images shows obvious differences in the crystal habit of the CVD crystallization products prepared by the two crystallizers. According to the microscope and SEM images, the CVD crystallization products prepared under the tubular crystallizer were rod-like compared to the batch crystallizer, which was lamellar-like.

Fig. 8.

Pictures of CVD characterization. (a) Microscope image of CVD tubular crystals; (b) microscope image of CVD batch crystals; (c) SEM image of CVD tubular crystals; (d) SEM image of CVD batch crystals.

Fig. 9.

PXRD patterns of CVD in two crystallization strategies.

Fig. 10.

DSC of CVD in two crystallization strategies.

Fig. 9 shows XRD images of CVD crystallization products prepared by the two crystallizers. The positions of the corresponding diffraction peaks (12.89°, 14.73°, 17.24°, 24.20°, 26.18°) in the tubular crystallizer were close to the central diffraction peaks (12.81°, 14.67°, 17.34°, 24.12°, 26.04°) of the CVD crystal in batch. However, the intensity of the two diffraction peaks was different by observing the ordinate, indicating that the crystal habit of the CVD crystallization products prepared by the two crystallizers slightly changed. However, since the CVD II crystal form was metastable, we inferred that the CVD prepared in the pot crystallizer had formed a mixed crystal form. Therefore, the XRD pattern was not apparent. CVD Form I was more stable (melting point: 123–126 °C), but Form II (melting point: 114–115 °C) was used in formulations sold on the market [39]. Fig. 10 shows that the melting point of CVD crystalline products prepared by the tubular crystallizer was 114.42 ℃. The temperature range of 114–115 ℃ of crystal form II indicates that the CVD prepared by the tubular crystallizer was the crystal form II. The melting point of the CVD crystallization product prepared by the batch crystallizer was 116.31 ℃, indicating that the CVD prepared by the batch crystallizer forms a mixed crystal Form I and II. According to Fig. 10, CVD crystal form II can be produced stably using tubular crystallizers, and the produced CVD crystal form II can meet industrial production needs.

3.4. Performance of ultrasound-assisted slug-flow crystallization to prevent encrustation

3.4.1. Promotion of nucleation

CVD continuous slug-flow tubular crystallization was conducted with (b) and without (a) ultrasound irradiation. In Fig. 11 (a), the nucleation site where the solution began to cloud was 100 cm away from the inlet, whereas, in Fig. 11 (b), it was 55 cm away from the inlet. CVD continuous slug-flow tubular crystallization without ultrasound irradiation has an obvious nucleation-promoting effect. Therefore, the nucleation process can be promoted and optimized using ultrasound irradiation.

Fig. 11.

Pictures for CVD continuous slug-flow tubular crystallization. (a) No ultrasound irradiation; (b) ultrasound irradiation.

3.4.2. Prevention of tubular crystallizer blockage

As shown in Fig. 12 (a) and (b), CVD crystals are fluffy in the tube, and localized accumulations occur in tubular flow crystallization. Fig. 12 (a) and Fig. 12 (b) were based on which tubular crystallization process without ultrasound irradiation and air bubble-assisted. Our results showed that clogging and wall sticking often occur in tubular crystallizers. Tubular crystallizers have been improved to achieve stable and continuous CVD metastable crystal form Ⅱ preparation. Ultrasound can significantly improve the crystal nucleation process as an effective process-strengthening tool. In addition, ultrasound can improve dispersion and reduce wall nucleation, thus avoiding the problem of crystal particles sticking to the tube. Each air solution segment in the system represents a separate cell, which can be seen as a functional crystallizer. Slug flow crystallization uses air as a carrier, allowing the liquid to flow at the same velocity as the solid particles. We combined ultrasound irradiation and air segmental slug flow to investigate their improvement effects.

Fig. 12.

Pictures for CVD tubular flow crystallization. (a) Crystal precipitation in the tubular; (b) the crystal was blocked in the tubular; (c) the air bubble-assisted tubular crystallization process has severe wall adhesion; (d) the ultrasound-assisted continuous slug flow tubular crystallization process has no wall adhesion and blockage.

In the slug flow-assisted tubular crystallization process (without ultrasound), the flow rate of the crystal was slower than that of ordinary fluid, and the crystal adhered to the pipe wall (Fig. 12 (c)), resulting in pipe blockage.

As shown in Fig. 12 (d), there was no blockage in the tubular crystallization process of continuous slug flow assisted by ultrasound. In addition, when we increased the amount of CVD to 20 g (240 g of ethyl acetate, flow rate of 25 mL/min), the crystallization process lasted 12 min, and CVD form II was continuously and stably generated without blocking the pipeline.

The ultrasound-assisted continuous slug flow tubular crystallization process has two advantages:

• (1)It provides a new strategy for avoiding CVD solution-mediated polymorphic transformation. A tubular reactor can be selected to avoid the risk of CVD crystallization during the crystallization process. The tubular reactor was a plug flow reactor that ensured a steady-state crystallization process, avoided the polymorphic transformation of CVD, and could stably prepare CVD metastable form II crystals;
• (2)Ultrasound inhibits tubular mold blockage and promotes nucleation. Crystals are also more complete with slug flow.

Fig. 4, Fig. 5, Fig. 6, Fig. 7 show the solvent-mediated polymorphic transformation experiment of CVD during batch crystallization. The polymorphic transformation of CVD occurred during the crystallization process of 323.15 K and began at 20 min. The polymorphic transformation process of CVD occurred during the crystallization process of 293.15 K and began at 150 min. Therefore, the solvent-mediated polymorphic transformation would not occur when 323.15 K CVD was injected into 293.15 K tubular crystallizer within 20 min. At the same time, CVD crystallization did not occur in the 293.15 K tubular crystallizer within 150 min. After dissolution, the CVD crystal solution was first cooled to 323.15 K. Then, the slurry was driven through the slug flow into the tubular crystallizer at a constant temperature of 293.15 K. Under the ultrasound wave, the CVD crystal had stable precipitation, the product crystal Form II was stable, and the operation was continuous to avoid clogging. At the same time, the benefit of moving from a batch operation to a continuous operation was a temperature and concentration distribution unaffected by time. This transition gave the product more stable and repeatable characteristics and avoided the problem of inconsistent changes from batch to batch.

CVD’s metastable form II crystals can be obtained continuously and stably during tubular crystallization. In addition, the ultrasound-assisted slug flow tubular crystallization method has the potential to transition from the laboratory to the industrial level and achieve stable preparation of pure metastable crystalline products.

4. Conclusions

In this study, we successfully prepared metastable form II crystals of CVD using a continuous mode tubular crystallizer. Two strategies, namely tubular crystallizers and batch crystallizers, were employed for the preparation of CVD metastable form II crystals. The resulting CVD crystalline products obtained from both methods were subjected to comprehensive characterization. According to the solvent-mediated suspension experiments in batch crystallization, the polymorphic transformation of CVD initiated after 20 min during the constant temperature crystallization process at 323.15 K. The onset of polymorphic transformation for CVD occurred at 150 min during the constant temperature crystallization process at 293.15 K. We propose a method for rapidly preparing metastable II crystals with a tubular crystallizer based on the intermittent crystallization mechanism of the CVD solution. The CVD metastable crystal form II was prepared using tubular and batch crystallizers to evaluate the effectiveness of the two preparation strategies. The results showed that the tubular crystallizer could produce metastable form II crystals. We successfully used air bubble- and ultrasound-assisted experiments to optimize the tubular crystallization process and solve the phenomenon of clogging and wall sticking in a tubular crystallizer. Finally, we developed the ultrasound-assisted continuous slug flow tubular crystallization process and evaluated its effects. Our experimental results showed that CVD products prepared by ultrasound-assisted continuous slug flow tubular crystallization could achieve robust, sustainable, and uninterrupted CVD products. The utilization of this approach effectively circumvented the issue of polymorphic transformation during the CVD production process, thereby demonstrating its efficacy in the practical implementation of metastable pharmaceutical manufacturing.

CRediT authorship contribution statement

Huadong Liao: Data curation, Conceptualization, Writing – original draft. Wenfeng Huang: Data curation, Validation, Investigation. Ling Zhou: Data curation, Conceptualization, Validation, Investigation. Lan Fang: Conceptualization, Investigation. Zhenguo Gao: Conceptualization, Resources, Writing – review & editing, Project administration, Funding acquisition. Qiuxiang Yin: Conceptualization, Resources, Project administration, Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.