A High-resolution Field Emission Gun Scanning Electron Microscope (FEG-SEM, 3D FEG, Quanta, United States) at 20 kV was used to investigate the microparticle’s morphology. Powder samples (PCL-MPs and AZM-PCL-MPs) were loaded by light dusting onto a metal stub with a double-sided adhesive carbon tape. The samples were coated with palladium (10 nm) with a sputter coater (TN-1100X-SPC-16M, Canada), using an electric potential of 2.0 kV for 2 min. Microscopic images were randomly captured at various magnifications from 500 to 25,000x.

The AZM-PCL-MPs and PCL-MPs formulations were resuspended (ca. 0.4 wt%) in distilled water as a dispersing medium, with a refractive index of 1.33. The analyses were carried out in Mastersizer MicroPlus (Malvern, United Kingdom) using a Hydro 2000 G (A) dispersion cell.

DSC analyzes were carried out using a DSC822 (Mettler-Toledo, Switzerland). Approximately 3–5 mg of the powdered sample (AZM-PCL-MP) and pure excipients were sealed into standard aluminum pans using a pin holed standard aluminum lid. All samples were scanned in the range of 25°C–500°C at a scanning speed of 10°C/min under 50 mL/min nitrogen gas purging using an empty aluminum pan as reference. Thermograms data were collected using the STARe SW thermal analysis software (Mettler Toledo, Switzerland).

TGA were conducted using a TGA/DSC 1 STARe System (Mettler-Toledo, Switzerland). Samples of approximately 10 mg of AZM-PCL-MPs were loaded into the platinum pan and analyzed in ramp heating mode at 10°C/min from 25°C to 500°C under nitrogen gas purging of 50 mL/min. The percentage of weight variation upon heating was estimated using the Universal Analysis 2000 software (Mettler Toledo, Switzerland).

The crystallinity of the microparticles was analyzed using an X-ray diffractometer (Rigaku, United States) equipped with a Cu-Kα radiation source and a fixed monochromator. The powder samples of AZM-PCL-MPs were placed in glass holders and X-ray diffraction was measured in the range of 5°–40° 2θ at a scan speed of 5°/min and a step of 0.02°. The voltage and current used were 40 kV and 30 mA, respectively.

ATR-FTIR spectra were acquired using a Nicolet Is10 Spectrometer (ThermoFisher, United States) in the % transmittance mode. The samples (AZM-PCL-MPs and pure excipients) were placed on the ATR diamond crystal with the pressure arm positioned over the sample and spectra were collected at a resolution of 4 cm⁻¹ with 64 scans per spectrum over the range 4,000–700 cm⁻¹. An unsampled background spectrum was collected before each test. All spectra data were collected using ResolutionPro (5.2.0, Agilent Technologies software).

Petri dishes containing nutrient agar and potato-sucrose-agar were seeded with 10⁶ cells of Staphylococcus aureus and Streptococcus pneumoniae, respectively. In sequence, 6 mm discs impregnated with 20 µL of the respective serial dilutions from 25 to 1,600 μg/mL of AZM, AZM-PCL-NP, PCL-NP were placed in Petri dishes and kept at 37°C for 18 h to analyze bacterial growth and halo formation. Microbiological tests were performed in triplicate [22].

Samples of 10⁶ bacterial cells were plated in 96-well plates and submitted to AZM, AZM-PCL-NP, PCL-NP. The drug concentration ranged from 25 to 1,600 μg/mL and the strains were kept in an incubator at 37°C for 24 h. Two hundred microliters of sterile water was added to all perimeter wells to avoid evaporation during the incubation. After 24 h, 10 µL of resazurin (1 wt%) was added to each well and after 1 h cell viability was observed through color change. A change in color from blue (oxidized state non-viable cell) to pink (i.e., reduced viable cell) indicated bacterial growth [23, 24].

To determine the minimum inhibitory concentration (MIC) of AZM and AZM-PCL-NP formulation, microdilution was performed in 96-well plates containing 10⁶ cells/well of S. aureus [20, 25] and S. pneumoniae. The AZM concentration ranged from 25 to 1,600 μg/mL. After microdilution in broth, each well with the respective treatment was seeded in Petri dishes with TSA and nutrient agar, for S. pneumoniae and S. aureus, respectively. After incubation for 24 h at 37°C, the colony forming unit (CFU) count was performed and the minimum concentration required to inhibit bacterial (i.e., CIM) in both strains was calculated.

The scattering profiles of PCL-NP containing or not AZM are presented in Figure 2A. As can be seen, the placebo PCL-NPs presented mean hydrodynamic diameter of 228 ± 8 nm, while the AZM-loaded formulations (AZM-PCL-NPs) presented mean hydrodynamic diameter around 195 ± 12 nm. The corresponding polydispersity index (PdI) were 0.099 ± 0.047 and 0.137 ± 0.032, respectively, indicating homogenous monodisperse systems. These values are similar to previous ones reported in the literature for PCL nanoparticles prepared by the emulsion/solvent diffusion-evaporation method, corresponding to structures of 180–250 nm [26]. The small size and narrow size distribution can be related to the NP preparation method, such as emulsification-diffusion-evaporation. The ratio between components of the organic phase, high shear speed (i.e., 14,000 rpm) promotes the breakage of droplets into smaller ones, resulting in a narrow size distribution [27]. The use of surfactant copolymer (i.e., Pluronic F-127) stabilizes the interface between the organic and aqueous phase, preventing droplet coalescence during emulsification process. Maintaining the aqueous phase in excess can favor the efficient diffusion of the solvent into the aqueous phase, facilitating the formation of homogenous NPs. Finally, conducting the diffusion and evaporation processes at low temperature (i.e., 50°C) allows the formation of more uniform NPs [28].

This approach also allowed us to achieve a high AZM encapsulation efficiency (%EE) of 83%. The adequate molar proportion of AZM, PCL and DPPC (0.06:0.19:0.75) in the AZM-PCL-NP formulation favored the formation of homogeneous and stable droplets, avoiding particle aggregation and promoting better drug retention inside the NPs, resulting in a high EE%. The adequate ratio of AZM:PCL provided the necessary polymeric matrix to encapsulate AZM, while DPPC stabilized the particle structure without compromising encapsulation. Besides, Pluronic F127 used as a polymeric surfactant also stabilized the emulsification and diffusion process, preventing particle coalescence and resulting in homogeneous particles, favoring high %EE [29, 30]. A high drug encapsulation efficiency of 84% for paclitaxel was also achieved using PLGA-PCL microparticles [31].

DSC analysis was performed on the raw materials (AZM and excipients) as well the AZM-PCL-MPs to confirm the successful loading of the drug into the formulation (Figure 3). According to previous reports, the melting temperature (T_m) of PCL was demonstrated to be 60°C–71°C [34, 35]. In this study, the endothermic peak of PCL was observed at 66°C. The T_m value of Pluronic F127 obtained in this work, 57°C, is similar to that reported by Karolewicz et al. [36]. For DPPC, no thermal event was observed in the range of 25°C–300°C, which has already been verified by Eedara et al. [37]. On the other hand, the lactose LH206 presented the T_m at 221°C. Listiohadi et al. [38] reported another thermodynamic event of exothermic nature for lactose in the range of 150°C–175°C, which suggests a possible recrystallization of the amorphous phase. We observed this transition around 176°C. AZM presented a T_m value of 126.4°C, which is also closer to the value reported by Maswadeh et al. [39] at 125°C. Furthermore, another thermodynamic event can be observed around 260°C, which can be attributed to the AZM degradation temperature (Td).

For the AZM-PCL-MP (Figure 3), the endothermic peak shifted to a lower temperature of 57°C, corresponding to the T_m of PCL and Pluronic F127 and indicating the miscibility of both polymers in the formulation. Endothermic events at 132°C and 212°C can be attributed to lactose monohydrate, present in higher proportion in the formulation. The endothermic peak corresponding to the Tm of AZM (126°C) was not observed, indicating that the AZM is dispersed in the polymeric nanoparticle matrix [33]. Contributions from the NP formation process may result in the absence of the AZM melting peak, such as the interactions between AZM and the PCL matrix, which stabilize the drug in the amorphous form [40]. The spray-drying process also promotes rapid drying of NPs, preventing recrystallization of AZM. Lactose, which acts as a drying aid and can improve the distribution of AZM in the matrix [41]. Therefore, the resulting formulation presents a molecular or solid dispersion of AZM, which explains the absence of the AZM melting peak in the DSC analysis.

The TGA curves (Figure 4A) represent the thermal degradation of the components (AZM and excipients) and the spray-dried AZM-PCL-MPs. The PCL thermogram showed two events of degradation at 325°C and 410°C. Pluronic F127 thermograms showed one-step process of thermal decomposition at 389°C, corroborating with previous results [33]. AZM presented thermal degradation at 260°C, reaffirming results of DSC analysis [37]. The TGA curve for LH206 showed a minimum loss of mass (ca. 0.4%) between 40°C and 130°C which corresponds to the loss of water at the surface, while a second mass loss (ca. 5.1%) around 153°C, indicating a loss of hydration water [42]. The TGA curve for spray-dried AZM-PCL-MPs remained constant up to 120°C where loss of water was observed. A second event occurs at 212°C, leading to an abrupt drop in the mass (ca. 40%) and suggesting the beginning of the decomposition. Finally, a decomposition of more than 70% is observed at 400°C.

The Derivative Thermogravimetry (DTG) curves of raw materials (AZM and excipients) and spray-dried AZM-PCL-MPs are shown in (Figure 4B). The thermal stability of PCL and Pluronic F127 were observed in the thermal degradation pattern of the spray-dried formulation AZM-PCL-MP indicating that both polymers retained their original thermal behavior in the microparticles. PCL and Pluronic F127 showed a clear degradation peak temperature at 410°C and 388°C, respectively, confirmed by Ramírez-Agudelo et al. [43]. AZM thermogram showed a thermal degradation at 260°C, in agreement with values reported by Maswadeh et al. [37]. According to the DTG results, the thermal degradation of AZM-PCL-MP could be divided into three stages. The first stage is associated to the intermolecular water loss around 136°C. The second stage at 220°C is attributed at the degradation of LH206, which is in higher proportion in the AZM-PCL-MP formulation. The third stage is assigned to the decomposition of the AZM-PCL-MP formulation at 400°C. Data obtained by DSC not only corroborate with the TGA, but also present first order phase transitions (melting temperature), indicating the high degree of purity of the microparticles produced.

X-ray powder diffraction was used to evaluate the solid-state properties of both raw materials (AZM and excipients) and spray-dried AZM-PCL-MPs (Figure 5). Data on the crystal lattice parameters of AZM have already been described in the literature [44]. AZM at room temperature is an orthorhombic structure, and belongs to the space group P21, a = 14.735 Å, b = 16.844 Å, c = 17.810 Å, α = 90°, β = 90°, γ = 90°, describing the crystallinity of AZM.

In this work, the AZM diffractogram showed an intense peak at 10.06° (2θ) and other sharp peaks at 16.62°, 17.80°, 18.88° and 19.96° (2θ), highlighting the crystallinity of the drug as studied by [5], where similar peaks were obtained for AZM encapsulated by poly-lactic acid (PLA). PCL and Pluronic F127 exhibited the characteristics of a semicrystalline polymer and copolymer, respectively, with two broad peaks at 21.36°; 23.78° and 19.28°; 23.46° (2θ), respectively. These peaks were also found for other research using PCL and chitosan nanofibers with encapsulated AZM [40]. Lactose LH206 appeared as a crystalline material, with an intense peak at 20.12° 2θ and the presence of some peaks with lower intensity at 12.68°, 19.28° and 21.38° (2θ). On the other hand, in the DPPC, a diffraction pattern without peaks and in a disorganized form can be observed, demonstrating the amorphous structure of this raw material.

For PCL-MPs and AZM-PCL-MPs, they showed a peak with greater intensity at 20° and 20.1° (2θ), and other peaks with lower intensity were observed at 12.5° and 19.2° (2θ), which are attributed to lactose LH206. Furthermore, both powder formulations (PCL-MPs and AZM-PCL-MPs) exhibited diffraction patterns similar to those obtained for PCL (21.48° 2θ) and Pluronic F127 (23.1° 2θ) raw materials. However, no peaks were observed for AZM (10.06° 2θ), indicating that AZM is preferably dispersed in the polymeric nanoparticle matrix [33].

According to (Figure 6A, the spectrum of Pluronic F127, showed the main stretching of aliphatic bands (C-H) around 2,882 cm⁻¹ and symmetric C-O-C stretching at 1,104 cm⁻¹, already reported in literature [45]. Lactose LH206 showed the stretching vibration of the hydroxyl group (OH) at 3,667 cm⁻¹ and the characteristic bands at 2,989 cm⁻¹ and 2,898 cm⁻¹ attributed to the asymmetric and symmetric stretching vibrations of the CH₂, respectively [46]. The FTIR spectrum of DPPC presented the principal band between 3,600 cm⁻¹ and 3,700 cm⁻¹, which indicates the vibration of the OH group. Asymmetric and symmetric vibration at 2,931 cm⁻¹ and 2,852 cm⁻¹ were observed and attributed to the CH₂ group. The band at 1736 cm⁻¹ clearly represents the carbonyl stretching. The stretching vibrations of the phosphate group were observed in the range of 1,230 cm⁻¹ and 1,190 cm⁻¹. In addition, the band at 1,050 cm⁻¹ refers to the vibrations of the CO-PO₂ group vibrations [47].

In Figure 6B, the spectrum of AZM showed a band in the range of 3,559 cm⁻¹ and 3,488 cm⁻¹, attributed to the vibration of N-H group [37]. Furthermore, The CH₂ groups were also confirmed by the asymmetric and symmetric vibrations bands at 2,974 cm⁻¹ and 2,898 cm⁻¹, respectively [48]. The band at 1,721 cm⁻¹ confirmed the vibration of the carbonyl group. Other bands between 1,123 cm⁻¹ and 1,250 cm⁻¹ suggested the absorption associated with axial stretching of the C-O functional group [49].

The FTIR spectrum of PCL (Figure 6B) has a characteristic band at 2,950 cm⁻¹ indicating the asymmetric CH₂ stretching and symmetric CH₂ stretching at 2,880 cm⁻¹. The band at 1,723 cm⁻¹ represents the carbonyl stretching and the 1,108 cm⁻¹ was attributed to the symmetric C-O-C stretching [50].

Comparing the FTIR spectra of the raw materials and the AZM-PCL-MP formulation (Figures 6A, B), we can see that the characteristic band of AZM was present in the formulation spectra. For AZM-PCL-MPs (Figure 6B) spectra, not only the carbonyl group at 1,725 cm⁻¹ but also the asymmetric and symmetric stretching vibrations of CH₂ at 2,931 cm⁻¹ and 2,852 cm⁻¹ were observed. The band at 1,100 cm⁻¹corresponds to vibrations of C-O-C and were expected according to the raw materials. In addition, the displacement of the bands corresponding to the AZM for AZM-PCL-MPs indicated the formation of microparticles in which the drug was molecularly dispersed [5].

The particle size distribution of PCL-MP and AZM-PCL-MP formulations are showed in Figure 7. For the deep lung deposition of inhalable dry powders, diameters in the range of 1–5 µm are required [51]. In this study, 90% (d₉₀) of the AZM-PCL-MPs presented a size of 4.06 µm, or smaller which may result in successful deposition in the deeper airways. This particle size distribution can be related to the lower concentration of PCL used during the emulsification process, which consequently results in smaller microparticles when dried by spray-drying process. When the PCL concentration is reduced in the emulsification step, the viscosity of the organic phase decreases, leading to a smaller droplet size due to more efficient dispersion and breakup during emulsification step. This effect is consistent with observations made by Balmayor et al., who reported that lower PCL concentrations lead to a decrease in the size of microparticles formed during the emulsification process. In the spray-drying process, rapid solvent evaporation solidifies the droplets, fixing the microparticle size [52]. Since PCL has a relatively low glass transition temperature and is flexible, it supports the formation of small, spherical particles when processed under optimal conditions [53].

The particle size distribution of powder formulation is summarized in Table 2. As can be seen, volume median diameter ( d 0.5 ) and volume mean diameter ( d 4.3 ) varied from 1.95 µm to 1.96 µm and 2.47 µm to 2.83 µm, respectively. Span values of 1.70–1.83 indicate homogenous size distributions, which is important for accurate dosing [54].

The morphology of AZM, lactose LH206 and microparticles was investigated by SEM (Figure 8A) exhibited a slightly rough surface, in an aggregated disposition, with D > 10 µm. Spherical microparticles smaller than 5 µm and presenting a rough and highly porous surface were observed in lactose after spray-drying (Figure 8B).

Notably, the obtained AZM-PCL-MPs and PCL-MPs present a nearly spherical shape < 5 µm (Figures 8C, D). The microparticles presented shapeless surface, as expected for a complex mixture including DPPC phospholipids, polymers (PCL and Pluronic F127), AZM and lactose LH206. Other authors showed uneven particles and aggregation when adding different types of drying aids such as mannitol or leucine into polymeric formulations [32]. Rough surfaces were predominantly observed in our spray-dried formulations, which is preferred for pulmonary delivery since it tends to increase the aerosolization efficiency [49].

Overall, the particles obtained in this work showed a proper powder dispersion, which can result in an optimal inhalation profile and, therefore, higher drug delivery into the deeper airways [55].