Microparticle Production of Active Pharmaceutical Ingredient Using Supercritical Antisolvent Process: A Case Study of Allopurinol

Abstract

Allopurinol is a relatively water-insoluble drug and, consequently, its efficacy was frequently limited by the dissolution or solubility phenomena. The purpose of this study was to improve the solid-state properties and dissolution behavior of allopurinol via a supercritical antisolvent (SAS) process using CO₂ as an antisolvent. The effects of operating parameters: temperature (35–55 °C), pressure (80–100 bar), solution concentration (8–15 mg/mL), CO₂ flow rate (2–4 L/min), and solution flow rate (0.25–0.50 mL/min) were studied. Moreover, the physical properties of unprocessed and SAS-processed allopurinol were analyzed by SEM, FTIR, DSC, TGA, and PXRD. The dissolution rate of unprocessed and SAS-processed allopurinol was also investigated and compared. In this case study, allopurinol was effectively micronized from 15.3 μm to 1.35 μm at the optimal operating condition. The results verify that the solid-state properties and dissolution rate of allopurinol can be controlled and improved via the micronization process by using SAS technology.

1. Introduction

It has been reported that about 40% of newly developed drugs by the pharmaceutical industry are classified as poorly water-soluble active pharmaceutical ingredients (APIs) [1]. These APIs have a limited dissolution rate and solubility in gastrointestinal fluids, and may lead to incomplete absorption in the human body and, consequently, low bioavailability [2]. Improving the dissolution properties and bioavailability of poorly water-soluble APIs by solid-state property modification such as microparticle production is a crucial step in the pharmaceutical industry. In traditional manufacturing methods, crystallization and mechanical milling are commonly used in the microparticle production of API. However, this traditional technique has several drawbacks including residual solvent contamination, batch-to-batch deviation, and surface property damage. To develop intensified microparticle production technologies with better control over the solid-state properties of APIs, alternative crystallization processes have been designed and developed. Among the various developed processes, the supercritical antisolvent (SAS) technology has been frequently performed involving supercritical CO₂ because it offers various advantages such as being non-toxic, non-reactive, cost-effective, and nonflammable, and it demands relatively low critical conditions [3]. In the SAS process, a solute solution is introduced into a supercritical fluid that functions as an antisolvent, and this allows the solute to become supersaturated, which is compensated by nucleation and particle growth [4]. Many studies have shown that SAS can successfully reduce particle size, modify the crystal shape of polymorphic drugs, or generate amorphous particles to improve the dissolution rate and solubility of poorly water-soluble APIs [5,6,7].

In this study, microparticle production of allopurinol by using the SAS process was presented to control and modify the solid-state properties. Allopurinol, or xanthine oxidase inhibitor, is widely used in the treatment of hyperuricemic patients to prevent gout. Allopurinol can effectively reduce the production of uric acid and lower uric acid levels in the blood and urine. It has also been used to treat seizures, pain caused by pancreas disease, and certain infections. However, allopurinol is one of the poorly water-soluble drugs, having limited dissolution or solubility-limited absorption [8]. Thus, several pharmaceutical technologies have been proposed in the literature in order to enhance the dissolution properties and bioavailability of allopurinol. For example, Samy et al. [9,10] and Changdeo et al. [11] used the solid dispersion technique to enhance the dissolution profile of allopurinol. Dai et al. [12] employed the cocrystallization method to improve the solubility and permeability of allopurinol. However, these studies improved the dissolution property of allopurinol by adding additional components such as polymeric carriers or coformers. It may also bring potential risk from stability consideration. An alternative approach adopted in this study is to use the micronization technique to generate microparticles of API for dissolution improvement. To our knowledge, no research has been published on allopurinol microparticle production using the SAS technique. In addition to demonstrating the feasibility of microparticle production of allopurinol through SAS, the effect of operating parameters in the SAS process containing solvent system, operating temperature, operating pressure, solution concentration, solution flow rate, and the CO₂ flow rate were investigated and compared. The solid-state properties of micronized allopurinol were also examined and discussed. The dissolution profiles of allopurinol before and after the SAS process were finally compared.

2. Materials and Methods

2.1. Materials

Carbon dioxide (CO₂) with a minimum purity of 99.9% was provided by Cheng-Feng Gas Co. (Taiwan) and employed as the supercritical antisolvent in this study. Allopurinol was purchased from Sigma–Aldrich with a purity level of 98%, and its physical properties are given in

Table 1. Physical properties of allopurinol.

Formula	CAS No.	Mw	Molecular Structure
C5H4N4O	315-30-0	136.11

. In total, 11 organic solvents were used to screen the solvent system, including tetrahydrofuran (THF), ethyl acetate (EA), 1-methyl-2-pyrrolidinone (NMP), N, N-dimethylacetamide (DMAC), n-heptane, dichloromethane (DCM), dimethyl sulfoxide (DMSO), anisole, toluene, acetic acid, and ethylene glycol. These organic solvents with a minimum purity of 99.5% were purchased from Sigma–Aldrich, Merck, J. T. Baker, or Macron Fine Chemical. All of the chemicals were utilized in their original form.

2.2. Apparatus and Procedure

The SAS apparatus system was employed in this study to micronize allopurinol. The schematic diagram of the SAS system is shown in Figure 1. It consists of two HPLC pumps (SSI, series II), one for CO₂ delivery and the other for pumping allopurinol solution. The CO₂ flow rate was adjusted by a micro-metering valve at the precipitator’s outlet and measured by a rotameter at ambient conditions. A back pressure regulator (Tescom) and a micro-metering valve were used to control pressure in the precipitator. The temperature of the precipitator with a volume of 70 mL was controlled using an electric heating jacket. A stainless-steel frit with a pore size of 0.5 μm was installed at the bottom of the precipitator to collect the produced particles. The SAS experiment started with injecting supercritical CO₂ and pure solvent into the precipitator until the temperature, pressure, and flow rate reached a steady state. The liquid feed flow was then changed from solvent to drug solution. After the drug solution came into contact with supercritical CO₂, rapid recrystallization occurred due to the extraordinarily high supersaturation ratio. The injection was terminated when the amount of solution injected reached the desired level. The residual solvent was then removed from the inside of the precipitator using a supercritical drying method that provided supercritical CO₂ continuously for 30 min. The precipitator was depressurized after this drying stage, and the particles precipitated on the stainless-steel frit were collected for additional analysis.

The generated particles were analyzed using a scanning electron microscope (SEM), differential scanning calorimeter (DSC), powder X-ray diffractometer (PXRD), Fourier-transforms infrared spectroscopy (FTIR), and thermogravimetric analysis (TGA) to compare the solid-state properties before and after the SAS process. Scanning electron microscopy (SEM) was used to analyze the crystal habit of samples. The sample powder was attached on conductive adhesive tape and sputtered with a thin gold film. SEM instrument (Hitachi, S-3000H, Japan) was used to obtain images of the samples. The crystal structures of allopurinol crystals were spotted using the PXRD (Malvern PANalytical, X’pert, UK) with data collected from 5° to 50° at a scanning speed of 5°/min. The DSC (PerkinElmer, Jade DSC, USA) and TGA (TA, Dupont 951, USA) were used to analyze the thermal properties of crystals. For DSC measurement, the samples were heated from 20 °C to 400 °C at a heating rate of 10 °C/min, and for the TGA analysis, the samples were heated from room temperature to 800 °C at a rate of 10 °C/min. Furthermore, FTIR (PerkinElmer, Spectrum 100, USA) analysis was also used to verify the spectroscopic properties of the samples. In addition, the dissolution profiles of the allopurinol crystals before and after the SAS process were examined. A dissolution tester (Shin Kwang, DT-1, Taiwan) and the paddle method were used to determine the dissolution rate of allopurinol. UV−vis spectroscopy (ThermoFischer, Evolution 60S, USA) was employed to analyze the dissolved amount of allopurinol.

3. Results and Discussion

3.1. Solvent Screening of SAS Processing of Allopurinol

The SAS process was employed in this case study for the microparticle production of allopurinol with control over its solid-state characteristics. Solvent screening is a crucial step in SAS operation to meet the goal of satisfactory recovery and throughput. Allopurinol is a polar compound with strong intermolecular hydrogen bonds and demonstrates limited solubility in both polar and nonpolar solvents. Zheng et al. [13] reported the solubility data of allopurinol in 14 organic solvents from 278.15 to 333.15 K. However, most of the organic solvents show low solubility value, and are unfavorable for crystallization study. To further explore more solvent candidates from regulatory consideration, 11 organic solvents were selected to determine an ideal solvent for SAS operation of allopurinol, including THF, EA, NMP, DMAC, n-heptane, DCM, DMSO, anisole, toluene, acetic acid, and ethylene glycol. These organic solvents are classified in the ICH Q3C guidelines as Class 2 or Class 3 solvents; they have low toxicity and are widely used in the pharmaceutical industry. To select a suitable solvent system, the solubilities of allopurinol in these 11 organic solvents were measured at 35 °C using the method established by Lee et al. [14]. According to the determined solubility values, allopurinol was relatively insoluble in THF, EA, n-heptane, DCM, anisole, toluene, acetic acid, and ethylene glycol (<1 mg/mL). As a result, the SAS process’ throughput employing these 8 solvents is unacceptably low and inefficient. On the other hand, allopurinol was more soluble in NMP, DMSO, and DMAC. The measured solubility values of allopurinol dissolved in NMP, DMSO, and DMAC at 35 °C were 15.7, 32.6, and 11.4 mg/mL, respectively. Thus, these three solvents were selected in SAS experiments to screen the solvent system.

Figure 2 shows the comparison of crystal habit of unprocessed and SAS-processed allopurinol using NMP, DMSO, and DMAC as the solvent. The crystal habit of unprocessed allopurinol was rod-like, as shown in Figure 2a. After SAS treatments using NMP and DMAC as the solvent, the crystal habit of SAS-processed allopurinol was similar to unprocessed allopurinol in being rod-like crystals, as shown in Figure 2b,c, respectively. However, while using DMSO as the solvent, needle-like crystals were generated as shown in Figure 2d. Crystal habits with a high aspect ratio such as needle-like shape may bring problems in downstream powder handling such as particle agglomeration. Furthermore, according to FTIR, DSC, and PXRD tests, the polymorphic forms of micronized allopurinol were compatible with the unprocessed sample as Form I. For the micronization of allopurinol, in this study, NMP was selected as the ideal solvent system because the smallest allopurinol crystals and highest recovery yields (>96%) were obtained using NMP as the solvent. The operating temperature (T) and pressure (P), CO₂ flow rate (F_CO2) and solution flow rate (F), concentration (C), and spraying nozzle diameter (N) for these three solvent screening SAS experiments were set at 35 °C, 100 bar, 4 L/min, 0.25 mL/min, and 100 μm, respectively. To reach a greater throughput, the solution concentration was set at 90% of the measured saturation solubility.

3.2. Effect of Operating Parameters

The impacts of five operating parameters in the SAS process, namely, the operating temperature, operating pressure, solution concentration, CO₂ flow rate, and solution flow rate on the particle size of allopurinol, were studied using NMP as a solvent. The SAS results containing recovery, crystal habit, and mean particle size were obtained under different operating conditions and are shown in

Table 2. Experimental conditions and results of SAS-processing of allopurinol using NMP as the solvent.

Exp. no.	T (°C)	P (bar)	C (mg/mL)	FCO2 (L/min)	F (mL/min)	Nozzle (μm)	Recovery (%)	Crystal Habit	Mean Size (μm)
Ori.	-	-	-	-	-	-	-	Rod-like	15.3
1	35	100	15	4	0.25	100	96.3	Rod-like	1.41
2	45	100	15	4	0.25	100	55.9	Irregular	2.41
3	55	100	15	4	0.25	100	No particle formation
4	35	80	15	4	0.25	100	84.9	Rod-like	1.35
5	35	100	8	4	0.25	100	86.8	Needle-like	2.68
6	35	100	11	4	0.25	100	97.3	Needle-like	1.71
7	35	100	15	2	0.25	100	95.3	Needle-like	1.99
8	35	100	15	4	0.50	100	73.8	Rod-like	2.08

.

Three temperatures were employed to investigate the influence of operating temperature: 35 °C, 45 °C, and 55 °C. Figure 3a,b illustrate the SEM images of allopurinol samples after the SAS processing at different operating temperature. When the operating temperature increases from 35 °C to 45 °C, the mean particle size increases considerably from 1.41 μm (Exp. 1) to 2.41 μm (Exp. 2), and the crystal habit changes from rod-like to irregular. In addition, the recovery was significantly reduced from 96.3% to 55.9%. According to the results that were reported by Chinnarasu et al. [15], when the temperature increases, the solubility of the drug in the entire system increases, resulting in a decrease in the level of supersaturation, which is not favorable to the nucleation of particles, and is favorable to the growth of particles, thus generating larger particles. In addition, when the operating temperature was further raised to 55 °C in this study, no allopurinol crystals were produced. The effect of operating temperature for allopurinol can be interpreted by the concept of mixture critical point (MCP). The critical locus data for the binary mixture of CO₂ and NMP was reported by Rajasingam et al. [16] and used to analyze the SAS results obtained from different operating temperatures in this study. When operating at 35 °C, the SAS operation region was above the MCP, and the system reached a supercritical state, which was beneficial to the generation of smaller particles. At 45 °C, the SAS operation region was close to the critical point of the mixture; thus, the system shifted into the two-phase gas–liquid region instead of being in a supercritical state, resulting in a decrease in the antisolvent effect and the recovery. When operating at 55 °C, the SAS operation region was lower than the MCP, and the system was located in the two-phase gas–liquid region rather than in a supercritical state, which greatly reduces the anti-solvent effect and resulted in no particle formation. It can be acknowledged that the high temperature is not conducive to the generation of smaller particles. The subsequent experimental temperature was fixed at 35 °C because operating above the MCP at 35 °C is desirable for SAS recrystallization and produces small API crystals, according to the previous studies [6,17,18]. To study the effect of operating pressure (P), two pressures, 80 bar and 100 bar, were used. Table 2 depicts the effects of the operating pressure on the mean particle size of allopurinol. When the operating pressure increased from 80 bar to 100 bar, the mean particle sizes of allopurinol increased slightly from 1.35 μm (Exp. 4) to 1.41 μm (Exp. 1). The crystal habit of allopurinol remains similar to the rod-like condition (Figure 3c). When the temperature is fixed at 35 °C, the operating pressure from 80 bar to 100 bar in this study is above the MCP. When the pressure increases, the solubility of the drug in the supercritical solution increases, leading to a decrease in the degree of supersaturation, which is not conducive to generating smaller particles of allopurinol. This trend is consistent with the finding of Rajasingam et al. and Su et al. [16,19].

For studying the effect of solution concentration, three different concentrations were examined. As presented in Table 2, once the solution concentration was increased to the highest value, there were obvious decreases in the mean particle size of allopurinol and receipt of a satisfactory recovery, because the degree of supersaturation increased as the solution concentration increased. As a result, the nucleation rate was increased, resulting in the production of smaller particles [20]. Moreover, a change in the crystal habit was detected after the solution concentration decreased from 15 mg/mL (Exp. 1) to 11 mg/mL (Exp. 6) and 8 mg/mL (Exp. 5). With the decrease in solution concentration, the crystal habit of SAS-processed allopurinol was modified from rod-like (Exp. 1) to needle-like (Exp. 5 and 6), as revealed in Figure 3. The effect of the CO₂ flow rate was also confirmed in this study. The CO₂ mole fraction in the precipitator may be changed by adjusting the CO₂ flow rate. According to previous studies [21,22,23], the high CO₂ mole fraction in the precipitator has a substantial antisolvent impact and facilitates the formation of small crystals. The above-mentioned finding is applied in this work as well. When the CO₂ flow rate was reduced from 4 L/min to 2 L/min, the mean particle size increased from 1.41 μm (Exp. 1) to 1.99 μm (Exp. 7), and the crystal habit of SAS-processed allopurinol changed from rod-like to needle-like, as illustrated in Figure 3f. Finally, to determine the effect of solution flow rates on SAS micronization of allopurinol, two different solution flow rates were varied. Once the solution flow rate increased from 0.25 mL/min to 0.50 mL/min, the mean particle size increased from 1.41 (Exp. 1) μm to 2.08 μm (Exp. 8), and the recovery yield decreased from 96.3% to 73.8%. On the other hand, the crystal habit remains the same as can be observed in Figure 3g. As the solution flow rates increase, the molar ratio of carbon dioxide in the precipitator decreases, resulting in an increase in the solute’s saturated solubility, reducing supersaturation, reducing allopurinol recovery, and causing particle size growth [19].

For further comparison of the crystal forms, thermal properties, and spectrometric properties, Figure 4, Figure 5, Figure 6 and Figure 7 show the FTIR, DSC, TGA, and PXRD results for unprocessed and SAS-processed allopurinol. Figure 4 presents a comparison of FTIR absorbance spectra for the unprocessed allopurinol and the SAS-produced allopurinol. The FTIR spectrum of allopurinol from the SpectraBase database is also presented in Supporting Information (Figure S1). As can be seen, all FTIR spectra show characteristic shoulders of allopurinol at 790 and 1245 cm⁻¹ for CH in-plane deformation, 1590 cm⁻¹ for ring vibration, 1700 cm⁻¹ for CO stretching vibration of the keto form of the 4-hydroxytautomer, and 3060 cm⁻¹ for CH stretching vibrations of pyrimidine ring [9,10]. According to Figure 4 and Figure S1, the FTIR spectra of the unprocessed and SAS-processed allopurinol are consistent and show that no changes in the structure of the allopurinol occur during the SAS process. To investigate the thermal properties, the DSC measurements were conducted. The DSC analysis results indicate that both the unprocessed and SAS-processed allopurinol had the same melting behavior, as shown in Figure 5. Figure 6 presents the comparison of TGA analysis for the unprocessed and SAS-processed allopurinol. As can be observed, the thermal decomposition properties of the unprocessed and processed allopurinol are consistent. In addition, there was no weight loss before thermal decomposition, showing no residual solvent contamination in the produced crystals.

Figure 7 compares the PXRD pattern of the unprocessed and SAS-processed allopurinol. In the Cambridge Crystallographic Data Centre (CCDC) database, allopurinol has also been recognized as form I (CCDC no.:1102293), and the calculated PXRD pattern is shown in Supporting Information (Figure S2). Moreover, the crystal structure of allopurinol has also been identified as form I by Tomoda et al. [24]. As indicated in Figure 7, the PXRD pattern of the SAS-processed allopurinol is consistent with that of the unprocessed allopurinol and with the confirmation reported in the literature. In addition, the maximum intensity peak in the PXRD pattern is presented in a subfigure in Figure 7. As can be seen, the increase in full width half maximum (FWHM) of the maximum intensity peak for the SAS-processed sample indicates a loss of crystallinity through SAS processing. Based on the above experimental evidence, the analytical results of FTIR, DSC, TGA, and PXRD of unprocessed and SAS-processed allopurinol were compared and discussed. The crystal forms, thermal properties, and spectrometric properties were consistent before and after SAS operations.

3.3. Dissolution Rate

The dissolution rates of the unprocessed and SAS-processed allopurinol were finally investigated in this case study. Figure 8 illustrates the comparison of the dissolution profile and particle size distribution of unprocessed and SAS-processed allopurinol. Due to the wide range of particle size distribution of the unprocessed sample, the dissolution experiment indicates that the unprocessed sample needs a long time to completely dissolve, and the dissolution of the unprocessed sample may have a burst release phenomenon. On the other hand, the SAS processed particles with narrower size distribution can avoid the burst release and, thus, the dissolution rate can stabilize and complete in less time. The unprocessed allopurinol was dissolved for about 75% in 15 min, and the SAS-processed allopurinol was dissolved for about 90% in 15 min. However, at 30 min, the SAS-processed allopurinol was completely dissolved, and the unprocessed allopurinol was only dissolved by about 80%. In addition, the Weibull equation was also applied to correlate the dissolution profiles. The Weibull model was expressed as [25]:

$$m=1-\exp \left[\frac{-t^b}{a}\right]$$

(1)

where m is the accumulated fraction of the pharmaceutical compound in a dissolution medium at time t. a and b are two empirical parameters that were optimally fitted using the experimental data. As presented in Figure 8, the Weibull equation fit the dissolution rate data satisfactorily.

4. Conclusions

In this case study, allopurinol was successfully micronized using the SAS process. NMP was selected as the ideal solvent system. The optimum operating conditions were 35 °C, 80 bar, concentration of 15 mg/mL, CO₂ flow rate of 4 L/min, and solution flow rate of 0.25 mL/min. At the optimal operating conditions, the mean particle size of allopurinol was successfully micronized from 15.3 μm to 1.35 μm, and the recovery reached more than 84%. Moreover, the SAS-processed allopurinol presented consistent analytical results of FTIR, DSC, TGA, and PXRD compared with the unprocessed sample. The micronized allopurinol shows an enhancement in the dissolution rate compared with unprocessed allopurinol. This case study showed that the SAS method can efficiently control solid-state characteristics and improve the dissolution rate of allopurinol throughout the micronization process.