Biodegradable PLGA Microsphere Formation Mechanisms in Electrosprayed Liquid Droplets †

Microspheres composed of poly (lactic-co-glycolic acid) (PLGA) were formed in liquid droplets using the electrospray technique. The structure of the microspheres was controlled by changing the electric voltage of the electrospray. PLGA microspheres with porous structures and micro-sized nanocomposite particles comprising PLGA nanosphere aggregates were formed at 5.0–7.0 kV and 2.5–3.5 kV, respectively. The structural change was related to the extent of evaporation of the solvent from the droplets during their flight. When the evaporation was completed in the relatively small droplets, the microspheres with porous structures were formed in the droplets. To study the mechanism, we observed the effects of the electric voltage of the electrospray, PLGA concentration, flight distance of the droplets, and molecular weight of PLGA on the structure of the PLGA particles. The novelty of this study is the analysis of the size and structure of the PLGA microparticles, which were controlled by the electrospray technique. Therefore, this research has important implications for the structural design and preparation of nanocomposite particles.


Introduction
The emulsion solvent diffusion (ESD) method in water devised by Kawashima et al. (Kawashima et al., 1982;1998) is an established method for producing drug-loaded poly (lactic-co-glycolic acid) (PLGA) nanospheres. PLGA and the drug are dissolved in a mixture of acetone and ethanol and dropped into a stirred polyvinyl alcohol (PVA) aqueous solution. Nano-sized emulsion droplets are formed by spontaneous emulsification associated with the rapid diffusion of acetone into the external aqueous phase after dropping. Then, mutual diffusion of the solvents causes PLGA and the drug to precipitate as the solubility decreases. At this time, the external PVA phase is adsorbed onto the surface of the emulsion droplets and prevents the coalescence of emulsion droplets. This principle makes this method suitable for the production of ~200-nm-sized drug-loaded PLGA nanospheres, with a uniform particle size distribution, and it has been researched extensively as a new drug delivery system (DDS) technology (Kimura et al., 2008;Nakano et al., 2009;Hara et al., 2008;Tsujimoto et al., 2007a;2007b).
In this study, the ESD method was scaled (crystallizer size, 100 L) at the Pharmaceutical/Beauty Science Research Center of the Hosokawa Micron Corporation. The following steps were completed: (1) mass production of PLGA nanospheres loaded with useful ingredients for cosmetics and quasi-drugs (Charlesworth et al., 1960), and manufacturing and sale of unique skincare products and hair growth formulas containing these PLGA nanospheres; (2) development of a Good Manufacturing Practice (GMP) production method (Tsujimoto et al., 2004;2007b) for ~80-nm-sized PLGA nanospheres that can undergo sterile pressure filtration (Tsukada et al., 2009), for use in injectable pharmaceutical formulations. The study developed a production method for the practical use of PLGA nanospheres. However, there were issues in applying the ESD method regarding (1) the formation of PLGA microspheres and (2) the formation of PLGA nanocomposites. For example, in the manufacture of the PLGA microspheres, it is difficult to apply methods other than a W/O/W emulsion solvent evaporation method (in-liquid drying method) (Toguchi et al., 1991;Okada et al., 1997) using a strong solvent (chloroform, dichloromethane); however, the method carries a high environmental load. Regarding PLGA nanocomposites, an example of typical use is dry powder inhaler (DPI) formulations. Microcomposite particles comprised of drugloaded PLGA nanospheres possess both the flowability (capsule fillability and alveolar delivery) and redispersibility into nanospheres (absorption into alveolar tissue) required of DPI formulations (Tsujimoto et al., 2005;Yamamoto et al., 2004;2007). However, the manufacturing process requires the formation of PLGA nanospheres using the conventional ESD method and a subsequent mechanical compounding process (Yamamoto et al., 2004;2007) which can be time consuming and labor intensive. Therefore, a new manufacturing process is desirable.
The electrospray fine particle preparation method is attracting attention because particle size can be controlled by varying the electrostatic potential between the counter electrode and liquid (Kobara et al., 2019). In this study, as a novel method, we used an electrospray of the PLGA solution onto the surface of the PVA aqueous solution phase. From the results obtained, it can be inferred that, unlike in the conventional ESD method, it is possible to control the particle size from submicron to several micrometers with the same solvent composition without using a chlorinated solvent. In particular, it was found that the characteristic PLGA microspheres, such as PLGA microspheres with a porous structure and micro-sized nanocomposite particles, comprising PLGA nanosphere aggregates, can be produced by changing the preparation conditions. This result is important for the structural design and preparation of nanocomposite particles when applying PLGA microspheres as various platforms for DDS.
We would like to report herein this paper the mechanism for the PLGA microspheres formed in the electrosprayed liquid droplets, and the structure-controlling factors such as electric voltage, concentration of PLGA, flight distance of the electrosprayed droplets and molecular weight of PLGA.

Apparatus and method
In this experiment, three types of PLGA (Sample A, B: Fujifilm Wako Pure Chemicals, Sample C: Mitsui Chemicals) were used without further purification. Table 1 lists the basic chemical properties of the materials. Sample B (molecular weight (MW): 20,000 Da) was used for the experiments except for the experiment concerning the effect of the molecular weight. Fig. 1 shows a schematic of the experimental setup for the PLGA particle formation. A glass beaker (cylindrical shape, outer diameter of 53 mm, height of 70 mm) was filled with 50 ml of 0.25 wt% PVA aqueous solution, and placed on a temperature-controlled stirrer. A stainless-steel nozzle (outer diameter of 0.26 mm, inner diameter of 0.13 mm) was situated above the PVA aqueous solution surface. The distance between the nozzle tip and PVA aqueous solution surface was set to 20 mm as the standard condition. A high voltage for the electrospray was supplied between this nozzle and the counter electrode through a high-voltage DC power supply unit (HJPQ-10P3, HJPQ-10N3 (Matsusada Precision Inc.)). Because the counter electrode was in the aqueous solution, the surface of the aqueous solution acted as the counter electrode for the electrospray, as shown in Fig. 1.
For the preparation of PLGA particles, PLGA (33 mg/ml) was dissolved in an acetone/ethanol mixture with a 2:1 volume ratio. The PLGA solution was pumped with a constant flow rate of 0.03 ml/min using a microsyringe pump (YSP-101, YMC Co., Ltd.) to the electrospray nozzle, which was supplied with a high voltage. The concentration of the PLGA solution (33 mg/ml) was the same as that used in the conventional ESD method.
At the nozzle tip, the PLGA solution was fragmented into charged droplets according to the electrospray  principle. The charged droplets were continuously sprayed onto the surface of the PVA aqueous solution, which was gently stirred using a magnetic stirrer (stirrer size: 8 × 25 mm, cylinder type, and 800 rpm). For each experiment, the PLGA solution was electrosprayed at 0.03 ml/min for 60 min, i.e., 1.8 ml of PLGA solution was sprayed onto the surface of the PVA aqueous solution. This experiment was conducted at room temperature (288-293 K). After spraying, the PLGA particles suspended in the PVA solution were analyzed using the following procedures: (i) centrifugation of the PVA solution containing PLGA particles (48,000 × g, 4 °C, 30 min); (ii) removal of supernatant; (iii) redispersion with purified wa-ter and observation with a scanning electron microscope. Fig. 2 shows a typical SEM image of the PLGA particles produced using Run 4 ( Table 2). The Feret diameters of the particles were measured from the SEM images (magnification: 500-2000×). The average particle size was calculated by measuring the Feret diameters of 20-30 particles.

Liquid droplets generated by electrospray
The PLGA solution was electrosprayed in a strong electric field between the nozzle tip and counter electrode. The electrospray was started at 2.5 kV, and the electric voltage was increased slowly, while the PLGA solution was fed into the strong electric field at 0.03 ml/min through the nozzle, as shown in Fig. 3. At 2.5 kV, a Taylor cone and relatively large liquid droplets, which fragmented from the Taylor cone, were observed ( Fig. 3(a)). The formation of the Taylor cone, which is a characteristic of the electrospray, indicates that the electrospraying was accurately performed. The liquid droplets were generated by the electrostatic repulsion between the electric charges formed on the Taylor cone, which is called "coulombic explosion." Coulombic explosion can occur several times when the charge density is increased, owing to solvent evaporation from the droplets. With the increase in electric voltage, from 2.5 kV to 7.0 kV, the size of the liquid droplets decreased, owing to more coulombic explosion events. When the electric voltage was higher than 3.5 kV, multiple Taylor cones were formed (Figs. 3(b), (c), (d)). The droplet size was too small to be observed through a high-speed camera  (Photron, 100,000 fps). When the voltage was increased above a certain level, the size of the liquid droplet was assumed to be ≤ 10 μm based on the high-speed camera resolution.

Particleization of PLGA using electrospray
When the droplets of the PLGA solution were generated via the electrospray, the solvent in the droplets was vaporized. This increased the PLGA concentration during the flight of the droplets from the nozzle tip to the surface of the PVA aqueous solution. The solidification to form the PLGA particles was related to this condensation process. To investigate the effect of condensation on the electrosprayed liquid droplets and particleization, the following experimental conditions were observed, as listed in Table 2: (i) electrospray electric voltage (Run 1, 2, 3, 4, 5); (ii) PLGA concentration (Run 4,6,7,8,9,10); (iii) distance between the nozzle tip and surface of the PVA aqueous solution (Run 4, 11, 12); (iv) molecular weight of PLGA (Run 4,13,14). In each experiment, Run 4 was considered the standard condition, because the PLGA particles formed in Run 4 reflected the original characteristics of the electrospray. Other experimental conditions, such as the flow rate of the PLGA solution, aqueous PVA solution (concentration and volume), and stirring conditions, were fixed.

Effect of electric voltage of the electrospray on the formation of PLGA particles
In the conventional ESD method, the PLGA solution with the acetone/ethanol mixed solvent (volume ratio of 2:1) is mixed with water, containing PVA as a stabilizer. The ESD method uses the difference in solubility of PLGA in the acetone/ethanol mixture and water. PLGA is highly soluble in the acetone/ethanol mixture. In contrast, PLGA is hardly soluble in water. Therefore, precipitation of PLGA occurs when the two solutions are mixed, which results in the formation of nanospheres. In this study, electrospray was used to mix these two solutions; however, evaporation of the solvent from the electrosprayed droplets was found to play an important role in particleization, as discussed below.
As shown in Fig. 1, the PLGA solution with the acetone/ethanol mixed solvent was electrosprayed, and the resulting liquid droplets traveled to the surface of the PVA aqueous solution through the electric field. PLGA particles were formed in the aqueous PVA solution owing to the mixing of the two solutions. The microscopic structures of the resulting particles were observed through SEM. As shown in Fig. 4, the PLGA particle formation process was classified into three groups, depending on the electric voltage of the electrospray.
Group I at low electric voltages, 0 to +2.0 kV: PLGA nanospheres similar to the nanospheres formed in the conventional ESD method were observed (Figs. 4(a), (b)).
The effect of the supplied electric voltage on the structure of the PLGA particles is schematically depicted in Fig. 5.
In Group I, electrospraying was performed at low electric voltages. Relatively large droplets of the PLGA solution were fed into the PVA aqueous solution. The droplets were diffused into the PVA aqueous solution, which resulted in the formation of the PLGA nanospheres owing to the decrease in solubility of PLGA through interaction with water. This mechanism is the same as that of the conventional ESD method, reported previously (Kawashima et al., 1998).
In Group II, electrospraying was performed at electric voltages of +2.5 kV to +3.5 kV, and electrospray-generated fine droplets were formed. The solvent molecules of acetone and ethanol were evaporated out of the liquid droplets of the PLGA solution during the flight of the droplets from the nozzle to the surface of the PVA aqueous solution.
Owing to the evaporation of the solvent molecules, the PLGA concentration increased, which solidified the PLGA partially in the droplets and especially on the surface of the   droplets. This led to the formation of microspheres through the aggregation of PLGA nanospheres. In Fig. 4(c), the SEM image, which was magnified 30,000 times, shows a microsphere formed through the aggregation of PLGA nanospheres. The specific surface area of the microspheres was measured using the gas adsorption method, and found to be larger than that of the microspheres in Group III, as listed in Table 3. The microspheres in Group II had nanocomposite structures. The size of the microparticles formed in Group II was in the range of 2-10 μm. This size of the microparticles was in good agreement with that of the electrosprayed liquid droplets, suggesting that one microparticle was formed per liquid droplet.
In Group III, owing to high electric voltages (+5.0 kV to +7.0 kV), much smaller liquid droplets were generated. During the flight of the droplets, the solvent molecules of acetone and ethanol evaporated, leading to the precipitation of PLGA, and microspheres with a porous structure were formed. In Fig. 4(e), the SEM image, which was magnified 100,000 times, shows such a microsphere with a porous structure. The size of the microspheres in Group III was smaller than that in Group II. This was in agreement with the decrease in the size of the liquid droplets at higher electric voltages. This also supports the hypothesis that one microsphere of PLGA was formed per liquid droplet.
In the conventional ESD method, the number of microspheres formed was negligible. In contrast, by using electrospray, microspheres were formed when the solvent molecules were evaporated out of the electrosprayed liquid droplets.

Effect of PLGA concentration on the formation of PLGA particles
As shown in Figs. 4 and 5, the formation of PLGA microspheres was promoted by the solidification of the electrosprayed liquid droplets through the evaporation of the solvent molecules. To verify the mechanism in Fig. 5, the effect of the PLGA concentration on the formation of PLGA particles was examined. The concentrations of PLGA in the acetone/ethanol mixture (volume ratio: 2:1) were 3, 7, 17, 33, 67, and 133 mg/ml. The concentration used in the experiments on the effect of the electric voltage (Fig. 4) was 33 mg/ml. The SEM images of the resulting PLGA particles are shown in Fig. 6.
It was observed that the particle structures were dependent on the PLGA concentration; they were classified into three categories. At lower concentrations (3 and 7 mg/ml), microspheres were not formed. Instead, film-like structures were formed. At suitable concentrations of 17, 33, and 67 mg/ml, porous microspheres were generated. At higher concentrations, such as 133 mg/ml, microspheres were formed, but they did not have porous structures. They were dense and partially dimpled.
The observed effect of concentration is schematically illustrated in Fig. 7. At lower concentrations (Fig. 7(a)), during the flight of the electrosprayed liquid droplets, the solvent molecules were evaporated out of the liquid droplets; however, the concentration of PLGA, required for solidification, was insufficient to form the outer shell structure. Owing to the hydrophobicity of the PLGA solution, the electrosprayed fine droplets were eager to stay on the surface of the aqueous PVA solution (Fig. 1). Thus, the nanospheres resulting from the interaction of PLGA in the droplets with water on the surface of the PVA aqueous solution were present on the surface of the PVA aqueous solution. The resulting nanospheres aggregated on the surface of the PVA aqueous solution to form a thin film structure. Note that the acetone/ethanol solvent was also present around the surface of the aqueous PVA solution. This acetone/ethanol mixture may have contributed to the formation of the PLGA thin-film structure. It was assumed that the aggregating PLGA nanospheres were transformed into thin-film structures through the partial dissolution of the acetone/ethanol mixture. At suitable concentrations, porous microspheres were formed; moreover, the porosity and size of the microspheres decreased and increased, respectively, with increasing PLGA concentration (Figs. 6(c), (d), (e)). This was in agreement with the measurements of the specific surface area, which is listed in Table 3. When the PLGA concentration in the solution was increased from 17 to 33 mg/ml, the specific surface area decreased from 162 to 80 m 2 /g. This indicates a decrease in porosity with an increase in the PLGA concentration. Furthermore, the size of the microspheres increased with increasing PLGA concentration in this concentration range. This is related to the increase in size of the liquid droplets with the increase in the concentration. Because the viscosity of the PLGA solution increased with increasing PLGA concentration (Table 4), the size of the electrosprayed liquid droplets also increased.
At higher concentrations, such as 133 mg/ml, porous structures were difficult to generate. Moreover, numerous PLGA molecules in the droplets formed denser structures.
We observed that porous structures were generated for a specific range of PLGA concentrations. This also supports the hypothesis that one microsphere was formed per liquid droplet. Various studies on particle formation in emulsion droplets in a liquid have been conducted (Mersmann, 1998;Kawashima et al., 2002;Yoshizawa, 2004). However, the particle formation mechanism in the proposed electrospray method is better equipped to create various particle structures than the particle formation mechanisms reported previously. This is because the solvent diffuses from the emulsion droplets in the liquid and evaporates from the surface of the droplet in the air.

Effect of flight distance of the electrosprayed droplets on the formation of PLGA particles
Fig. 8 shows the results of particle formation by changing the distance between the nozzle tip and surface of the PVA aqueous solution. When the distance was increased, the droplet size also increased owing to the decrease in the electric field. The SEM images (Fig. 8) show that larger particles with porous structures were produced as the distance increased. Although the size of the droplet increased, solidification in the air occurred during long-distance flights.

Effect of PLGA molecular weight on the formation of PLGA particles
PLGA with a molecular weight of 20,000 Da was used. The size of the PLGA particles, resulting from the electrospray method, depended on the molecular weight of PLGA. Fig. 9 shows the observed SEM images of the PLGA particles, generated from the PLGAs with molecular weights of 5,000, 20,000, and 50,000. Porous structures were observed for these three molecular weights, and the average diameter increased with an increase in the molecu-lar weight. Because the viscosity and solubility also change with the molecular weight, the effect of the molecular weight was studied more systematically.

Conclusions and future prospects
(1) We found that PLGA microparticles were formed in the liquid droplets using the proposed electrospray technique. Porous PLGA microparticles and microparticles composing PLGA nanospheres, which cannot be produced by the conventional ESD method, were generated. The formation mechanism of these PLGA particles was classified into three categories depending on the extent of solvent evaporation from the droplets. (i) The evaporation of the droplets was completed in the air, leading to the formation of microspheres with a porous structure. (ii) Owing to the evaporation of the solvent from relatively large  droplets, solidification began on the surface of the droplets, leading to the formation of microparticles composed of nanospheres. (iii) The evaporation was insufficient for solidification in large droplets, leading to the formation of nanospheres similar to those in the conventional ESD method.
(2) If these microparticles can be loaded with drugs, they can be used for solid pharmaceutical preparations. The possibility of application in DPI and subcutaneous injection formulations will be the subject of future studies.
(3) Currently, we are also investigating the following subjects: (i) the mass production procedure of PLGA microparticles using electrospray; (ii) the size-distribution control of PLGA nanospheres using electrospray in a liquid medium instead of air. We will report these results in the future.

Moe Tanaka
Moe Tanaka received her bachelor's degree in pharmacy in 2019 from Osaka University of Pharmaceutical Sciences. As a pharmacist and researcher, she has joined the Pharmaceutical /Beauty Science Research Center in Material Business Division of Hosokawa Micron Corporation in 2019 and engaged in the research and development of skincare and haircare products containing PLGA particles. Her current research interests are in formulation design using PLGA and its application to functional cosmetics.

Ayaka Ochi
Ayaka Ochi received her bachelor's degree in 2018 and master's degree in fermentation chemistry in 2020 from Ehime University, Japan. As a researcher, she has joined the Pharmaceutical /Beauty Science Research Center in Material Business Division of Hosokawa Micron Corporation in 2020. Her current research interests are in skincare and aging care utilizing PLGA nanosphere technology and the development of functional cosmetics as an application of those technologies.

Aiko Sasai
Aiko Sasai is the vice director of the Pharmaceutical /Beauty Science Research Center in Material Business Division of Hosokawa Micron Corporation. She earned her Ph.D. degree in material engineering from Iwate University, Japan in 2009. Since joining Hosokawa Micron Corp. in 2009, she has been engaged in research on PLGA nanoparticles, and has been developing many cosmetics applying PLGA nanoparticle technology.