Synthesis of Microwave Functionalized, Nanostructured Polylactic Co-Glycolic Acid (nfPLGA) for Incorporation into Hydrophobic Dexamethasone to Enhance Dissolution

The low solubility and slow dissolution of hydrophobic drugs is a major challenge for the pharmaceutical industry. In this paper, we present the synthesis of surface-functionalized poly(lactic-co-glycolic acid) (PLGA) nanoparticles for incorporation into corticosteroid dexamethasone to improve its in vitro dissolution profile. The PLGA crystals were mixed with a strong acid mixture, and their microwave-assisted reaction led to a high degree of oxidation. The resulting nanostructured, functionalized PLGA (nfPLGA), was quite water-dispersible compared to the original PLGA, which was non-dispersible. SEM-EDS analysis showed 53% surface oxygen concentration in the nfPLGA compared to the original PLGA, which had only 25%. The nfPLGA was incorporated into dexamethasone (DXM) crystals via antisolvent precipitation. Based on SEM, RAMAN, XRD, TGA and DSC measurements, the nfPLGA-incorporated composites retained their original crystal structures and polymorphs. The solubility of DXM after nfPLGA incorporation (DXM–nfPLGA) increased from 6.21 mg/L to as high as 87.1 mg/L and formed a relatively stable suspension with a zeta potential of −44.3 mV. Octanol–water partitioning also showed a similar trend as the logP reduced from 1.96 for pure DXM to 0.24 for DXM–nfPLGA. In vitro dissolution testing showed 14.0 times higher aqueous dissolution of DXM–nfPLGA compared to pure DXM. The time for 50% (T50) and 80% (T80) of gastro medium dissolution decreased significantly for the nfPLGA composites; T50 reduced from 57.0 to 18.0 min and T80 reduced from unachievable to 35.0 min. Overall, the PLGA, which is an FDA-approved, bioabsorbable polymer, can be used to enhance the dissolution of hydrophobic pharmaceuticals and this can lead to higher efficacy and lower required dosage.


Introduction
Poor solubility and low bioavailability of active pharmaceutical ingredients (API) have hindered drug development, and pose many challenges for the pharmaceutical industry [1]. It is estimated that about 40% of market-approved and 90% of the development pipeline API have low aqueous solubility [2]. Such hydrophobic low-solubility APIs are classified as Biopharmaceutical Classification System (BCS) Class II and Class IV drugs, mostly weakly acidic or basic [3]. Pharmacokinetics and pharmacodynamic parameters such as drug distribution, therapeutic activity, metabolism and absorption are strongly dependent on their solubility [4]. Different approaches for solubility enhancement including particle size reduction, amorphous solid dispersions, microencapsulation, complexation, micelles, microemulsions formation, solid-state alternation, soft gel encapsulation, crystal engineering and lipid-based technologies have been used to deliver hydrophobic molecules [5][6][7][8][9]. However, they have their limitations such as alteration in the polymorph, miscibility, addition of undesirable additives and complex processing [10].
tions of dexamethasone to enhance its efficacy, and reduce pharmacokinetic cytotoxicity. Research has shown that an anti-inflammatory dexamethasone encapsulated into a biological cell-coated membrane has enhanced therapeutic efficacy with extended in vivo delivery [54]. Liposome-encapsulated dexamethasone is another such formulation with promising results [55]. However, the major limiting factor for hydrophobic dexamethasone is its intrinsic poor solubility and dissolution properties. Developing the right nanoparticle and functionality to enhance drug solubility will greatly aid in the effectiveness of treatment using dexamethasone.
It is anticipated that PLGA, which is not water-soluble by itself, can be surfacefunctionalized to have a hydrophilic surface. Nanoparticles of this functionalized form can potentially be incorporated onto the surface of a hydrophobic drug crystal, which then can be conduits for bringing water in contact with the drug crystal for faster dissolution. The objective of this research was to develop water-insoluble, surface-hydrophilized, nanostructured PLGA (referred to as nf PLGA) that can be incorporated into API crystals to synthesize drug-nf PLGA composites with enhanced dissolution properties. Another objective was to carry out microwave synthesis of nf PLGA, which is a fast and eco-friendly process. Corticosteroid dexamethasone (DXM), which is a highly hydrophobic BCS-IV drug with low water solubility (0.089 mg/mL), was used to form the soluble composites.

Synthesis of nfPLGA
The acid oxidation of PLGA was carried out in a multimode CEM microwave reactor (MARS-5, Matthews, NC, USA). The PLGA was then mixed with a 3:1 H 2 SO 4 and HNO 3 mixture. Next, the acid mixed poly(lactic-co-glycolic acid) or PLGA samples were placed into the microwave chamber and reacted at an applied power of 800 W (maximum 1600 W ± 15%) and frequency of 2450 MHz (wavelength λ of 0.1223642685714 m) at 60 • C with the use of IEC method. After 1.0 h of microwave induced reaction, the samples were vacuum filtered and washed to obtain the functionalized particles. After drying the f PLGA particles were mixed in Milli-Q water for dispersion and sonicated using high-power (110 V, 20 kHz) probe sonicator (Ultrasonic processor FS-900 N) for one hour to produce nano f PLGA or nf PLGA.

Synthesis of DXM-nfPLGA
An antisolvent precipitation technique was used to synthesize the DXM-nf PLGA composites. This was a modification of a process described before [61]. Acetone solvent was used to dissolve dexamethasone (DXM) drug. A clear solution of nf PLGA was also made in acetone. This was added to the dexamethasone solution dropwise and sonicated for 10.0 min. Next, the drug composite solution was placed into a cold ice bath and the antisolvent Milli-Q water was added dropwise. A white and milky suspension of the drug-polymer composite was observed during the precipitation process. The precipitate was then filtered and dried in a vacuum oven (Isotemp vacuum oven, model 280A, Fisher Scientific) up to 48 h to reach constant weight.

Characterization of nfPLGA and DXM-nfPLGA Composites
The nf PLGA and DXM-nf PLGA were characterized using several analytical techniques. The synthesized particles hydrodynamic properties were analyzed through Dynamic Light Scattering or DLS (Malvern Nano ZS 90, Model: ZEN 3690, Worcestershire, UK). The functionalized poly(lactic-co-glycolic acid) molecular properties such as weight average molecular weight (M w ) and number average molecular weight (M n ) were identified by using Gel Permeation Chromatography or GPC with the Waters Breeze GPC System w/Autosampler at Rutgers Newark NJ. Scanning electron microscopy (SEM) using a JEOL JSM 7900F microscope (JEOL, Tokyo, Japan) was used to image the crystals after carbon coating with an EMS Quorum instrument. The SEM was operated at 1.0 kV at a working distance of 10.0 mm. Additionally, surface elemental composition of nf PLGA particles were determined by the Energy Dispersive Spectroscopy (EDS) connected to a SEM instrument. Thermogravimetric analysis (TGA) (PerkinElmer 8000) was used to study nfPLGA incorporation by heating the samples from 30 to 700 • C under a 20 mL/min nitrogen flow at 10 • C/minute. Differential Scanning calorimetry (PerkinElmer DSC 6000) measurements were used to determine the melting point. Raman spectroscopy and microscopy (DXRxi Raman Microscope, Thermo Fisher Scientific, USA) were carried out using a 532 nm laser and gratings and filters. X-Ray diffraction (PANalytical EMPYREAN XRD, Malvern, UK) was performed using a Cu Kα radiation source where determined the crystal structure intensity for 5-70 • 2-theta ranges. Additionally, transmission mode Fourier transform infrared (FTIR) spectroscopy analysis was carried out using IR Affinity-1, Agilent Cary 670 Benchtop FTIR instrument (Agilent Technologies, Santa Clara, CA, USA).
The contact angle measurements were performed by recording an image of a water droplet (which acted as the probe) on the solid polymer particles. These images were then analyzed using an online protractor to determine the contact angle [62]. In addition, octanol-water partitioning was studied by placing a drug sample onto a 1:1 ratio of water (aqueous phase) and octanol (organic phase). The mixture was stirred for an hour and allowed to reach equilibrium. The partitioned samples were extracted from the water and octanol phases by ultracentrifugation, and the concentration in each phase was determined using a UV-vis spectrometer to compute the partition coefficient or logP [63].
Dissolution measurements were made using United States Pharmacopeia or USP-42 paddle-II method. A Symphony 7100 Distek instrument (North Brunswick, NJ 08902, USA) was used for this. The pH was set at 1.4 to simulate stomach conditions. The DXM-nf PLGA samples were added to a dissolution bath containing 900.0 mL 0.1 N HCl to simulate the pH and dissolution was carried out at 37.5 ± 0.5 • C, rotation speed was set at 75.0 rpm. About 2.0 mL of dissoluted aliquots were transferred from the dissolution medium using needle syringe at 1, 20, 30, 50, 80, 120, 150, 180, and 240 min intervals, then filtered using 0.2 µm syringe filter and analyzed concentration by ultraviolet-visible (UV-vis) measurements. Agilent 8453 (Santa Clara, CA, 95051, USA) model UV-vis spectrophotometer was used for measuring dexamethasone (DXM) absorption at a wavelength at 243.0 nm. Finally, the saturation solubility of the synthesized DXM-nfPLGA composites were determined by stirring the sample in water for 48 h at a room temperature (25 • C) and at pH 7.0.

Synthesis of nfPLGA
The microwave functionalization process altered the PLGA particle properties quite dramatically. This is evident from the photographs in Figure 1. Table 1 presents some of the physicochemical properties of the synthesized nf PLGA as compared to the original PLGA. The experimental study found the microwave acid functionalization led to nanosizing and extensive oxidation on the polymer surface. The SEM-EDS analysis showed that the oxygen content increased from 24.76% to 53.07%, implying extensive surface oxidation. Some of the partial ester linkages were broken, which led to more carboxylation and hydroxylation in the synthesized product.

Synthesis of nfPLGA
The microwave functionalization process altered the PLGA particle properties quite dramatically. This is evident from the photographs in Figure 1. Table 1 presents some of the physicochemical properties of the synthesized nfPLGA as compared to the original PLGA. The experimental study found the microwave acid functionalization led to nanosizing and extensive oxidation on the polymer surface. The SEM-EDS analysis showed that the oxygen content increased from 24.76% to 53.07%, implying extensive surface oxidation. Some of the partial ester linkages were broken, which led to more carboxylation and hydroxylation in the synthesized product. Based on Figure 3b and the physical properties presented in Table 1, the nfPLGA nanoparticles were relatively water-dispersible (as high as 4.0 mg/mL) whereas the pure PLGA micron crystals were highly hydrophobic and non-dispersible. The water contact angle (°) was measured by placing a drop of water onto a pile of particles which showed that pure PLGA had a contact angle of 82°, while the nfPLGA had a low contact angle of 36°. This clearly demonstrated that the nfPLGA was significantly more hydrophilic in nature. The differential scanning calorimetry (DSC) analysis showed that nfPLGA had a melting point of 331.78 ℃ and glass transition (Tg) of 46.14 ℃ which were slightly lower than the original PLGA, which implied that crystallinity was unaltered. Furthermore, gel permeation chromatography (or GPC) analysis found that nfPLGA had a lower weight average molecular weight of Mw = 38.3 kDa and number average molecular weight of Mn = 18.4 (a.u). Based on the dynamic light scattering (DLS) analysis, the hydrodynamic diameter of nfPLGA particles in water was between 100 and 200 nm with an average (mean) of 161.0 nm with a polydispersity index (PDI) of 0.185. The dispersibility and size distribution of the nfPLGA were suitable for nfPLGA-drug composite formation.  Based on Figure 3b and the physical properties presented in Table 1, the nf PLGA nanoparticles were relatively water-dispersible (as high as 4.0 mg/mL) whereas the pure PLGA micron crystals were highly hydrophobic and non-dispersible. The water contact angle ( • ) was measured by placing a drop of water onto a pile of particles which showed that pure PLGA had a contact angle of 82 • , while the nf PLGA had a low contact angle of 36 • . This clearly demonstrated that the nf PLGA was significantly more hydrophilic in nature. The differential scanning calorimetry (DSC) analysis showed that nf PLGA had a melting point of 331.78°C and glass transition (T g ) of 46.14°C which were slightly lower than the original PLGA, which implied that crystallinity was unaltered. Furthermore, gel permeation chromatography (or GPC) analysis found that nf PLGA had a lower weight average molecular weight of Mw = 38.3 kDa and number average molecular weight of Mn = 18.4 (a.u). Based on the dynamic light scattering (DLS) analysis, the hydrodynamic diameter of nf PLGA particles in water was between 100 and 200 nm with an average (mean) of 161.0 nm with a polydispersity index (PDI) of 0.185. The dispersibility and size distribution of the nf PLGA were suitable for nf PLGA-drug composite formation.
The powder x-ray diffraction (XRD) analysis of PLGA and nf PLGA presented in Figure 2a show similar crystalline peak intensity and demonstrate that crystallinity did not change during microwave functionalization. The RAMAN data presented in Figure 2b

Characteristics of DXM-nfPLGA Composites
The scanning electron microscopy (SEM) image of pure DXM and DXM-nfPLGA composites are presented in Figure 3. Additionally, the SEM images of PLGA and nfPLGA are shown in Figure 3a,b.These images show that the crystal structure of the drug remained unchanged and the nfPLGA was successfully incorporated. Figure 3d,e show the presence of nfPLGA in a uniform distribution on the surface of the drug crystal, and these were expected to provide the hydrophilic linkages to the aqueous medium, leading to higher dispersibility and solubility. c)

Characteristics of DXM-nfPLGA Composites
The scanning electron microscopy (SEM) image of pure DXM and DXM-nf PLGA composites are presented in Figure 3. Additionally, the SEM images of PLGA and nf PLGA are shown in Figure 3a,b.These images show that the crystal structure of the drug remained unchanged and the nf PLGA was successfully incorporated. Figure 3d,e show the presence of nf PLGA in a uniform distribution on the surface of the drug crystal, and these were expected to provide the hydrophilic linkages to the aqueous medium, leading to higher dispersibility and solubility.

Characteristics of DXM-nfPLGA Composites
The scanning electron microscopy (SEM) image of pure DXM and DXM-nfPLGA composites are presented in Figure 3. Additionally, the SEM images of PLGA and nfPLGA are shown in Figure 3a,b.These images show that the crystal structure of the drug remained unchanged and the nfPLGA was successfully incorporated. Figure 3d,e show the presence of nfPLGA in a uniform distribution on the surface of the drug crystal, and these were expected to provide the hydrophilic linkages to the aqueous medium, leading to higher dispersibility and solubility.    Figure 4 presents the solubility (mmol/L) and octanol-water partition coefficients of DXM-nfPLGA. The saturation solubility of the formulated drug composites at pH 7.0 showed that that it changed from 0.13 mmol/L for the original drug to 1.89 mmol/L for DXM-nfPLGA. At the same time, the zeta potential, which is used to define colloidal stability, changed from −34.8 for the original drug to −44.3 mV 27 for the DXM-nf PLGA. Moreover, the physiological stability of DXM-nfPLGA-1.50 composite was assessed by dispersing particles in different buffer solutions at pH values of 4.0, 6.0, 7.0, and 10.0. No significant changes were observed in the average particle size or mean hydrodynamic diameter at different time intervals of 0, 1, 2, 4, 6, 24 h. This is shown in Figure 4b. This indicated high stability at all pHs. Octanol-water partitioning showed a similar effect, as logP reduced from 1.96 for pure DXM to 0.24 for DXM-nf PLGA (Figure 4a). An important consideration was whether the dexamethasone (DXM) was altered during the composite formation. Figure 5a presents the Raman spectrum of DXM and DXM-nfPLGA composites after nfPLGA incorporation where the major peak intensities observed for pure dexamethasone were at 688 cm −1 , 1448 cm −1 , 1602 cm −1 , 1658 cm −1 , 1704 cm −1 , 2908 cm −1 , and 2939 cm −1 . The spectral intensity for the DXM-nfPLGA composites shows no variation in these peaks associated with the different functional groups. X-ray diffraction (XRD) analysis (Figure 5b) showed similar crystal structures for DXM and the DXM-nfPLGA composites. This was based on the major intensity peak observed at two thetas (2θ) = 6.6°, 7.5°, 9.4°, 10.8°, 12.6°, 13.8°, 14.3°, 15.2°, 15.7°, 17°, 18.6° and so on. Therefore, it is concluded that there was no variation in polymorph. As a result, DXM in the DXM-nfPLGA composites are expected to remain biologically similar with increased solubility through the incorporation of inactive nfPLGA particles. Additionally, FTIR analysis presented in Figure 6a for the drug composites and pure DXM showed that they were chemically similar. Furthermore, DSC analysis presented in Figure 6b showed similar melting point for both the DXM and nfPLGA incorporated DXM composites. An important consideration was whether the dexamethasone (DXM) was altered during the composite formation. Figure 5a presents the Raman spectrum of DXM and DXM-nf PLGA composites after nf PLGA incorporation where the major peak intensities observed for pure dexamethasone were at 688 cm −1 , 1448 cm −1 , 1602 cm −1 , 1658 cm −1 , 1704 cm −1 , 2908 cm −1 , and 2939 cm −1 . The spectral intensity for the DXM-nf PLGA composites shows no variation in these peaks associated with the different functional groups. X-ray diffraction (XRD) analysis (Figure 5b) showed similar crystal structures for DXM and the DXM-nf PLGA composites. This was based on the major intensity peak observed at two thetas (2θ) = 6.6 • , 7.5 • , 9.4 • , 10.8 • , 12.6 • , 13.8 • , 14.3 • , 15.2 • , 15.7 • , 17 • , 18.6 • and so on. Therefore, it is concluded that there was no variation in polymorph. As a result, DXM in the DXM-nf PLGA composites are expected to remain biologically similar with increased solubility through the incorporation of inactive nf PLGA particles. Additionally, FTIR analysis presented in Figure 6a for the drug composites and pure DXM showed that they were chemically similar. Furthermore, DSC analysis presented in Figure 6b showed similar melting point for both the DXM and nfPLGA incorporated DXM composites.
Additionally, RAMAN mapping and imaging in Figure 7a,b was carried out to see the distribution of the nf PLGA on the single drug crustal surface. The distribution map was carried out with a strong peak at 1660 cm −1 , which corresponded to the carbonyl (C=O) band. The green spots in the images were attributed to the drug crystal surface, while the red was for the nf PLGA surface interaction and blue indicated the microscopic image background surfaces. The image showed a non-uniform distribution of nf PLGA on the drug crystal surface.      Thermogravimetric analysis (TGA) of nf PLGA composites as well as DXM are presented in Figure 8a. These show that the pure drug as well as the DXM-nf PLGA composites has similar decomposition profile. The concentrations of nf PLGA were also determined from the TGA data. The level of nf PLGA incorporation in the different composites were between 0.55% to 1.25%. The melting point (m.p.) data of the different composites are also presented in Table 2 found that the melting point of the original DXM was between 260 and 262 • C, and the DXM-nf PLGA composites showed similar values. This also confirms that the polymorph was not altered by nf PLGA incorporation, and the composites are thermally stable. Thermogravimetric analysis (TGA) of nfPLGA composites as well as DXM are presented in Figure 8a. These show that the pure drug as well as the DXM-nfPLGA composites has similar decomposition profile. The concentrations of nfPLGA were also determined from the TGA data. The level of nfPLGA incorporation in the different composites were between 0.55% to 1.25%. The melting point (m.p.) data of the different composites are also presented in Table 2 found that the melting point of the original DXM was between 260 and 262 °C, and the DXM-nfPLGA composites showed similar values. This also confirms that the polymorph was not altered by nfPLGA incorporation, and the composites are thermally stable.   Thermogravimetric analysis (TGA) of nfPLGA composites as well as DXM are presented in Figure 8a. These show that the pure drug as well as the DXM-nfPLGA composites has similar decomposition profile. The concentrations of nfPLGA were also determined from the TGA data. The level of nfPLGA incorporation in the different composites were between 0.55% to 1.25%. The melting point (m.p.) data of the different composites are also presented in Table 2 found that the melting point of the original DXM was between 260 and 262 °C, and the DXM-nfPLGA composites showed similar values. This also confirms that the polymorph was not altered by nfPLGA incorporation, and the composites are thermally stable.

Dissolution of DXM and DXM-nfPLGA
The in vitro dissolution experiment was based on the United States Pharmacopeia or USP-42 dissolution protocol, and the study was conducted in media that mimicked the gastric pH of 1.4. The enhanced dissolution of the drug composites was attributed to the presence of the hydrophilic nf PLGA, which led to hydrogen bonding (H-bond) with the API crystal, and eventually enhanced dissolution. Figure 9 shows the dissolution profile for dexamethasone (DXM) and the nf PLGAincorporated DXM composites. It is clear that nf PLGA led to enhanced dissolution rate and aqueous solubility. Table 2 presents the enhanced dissolution rates, as well as the time required to reach 50% (T 50 ) and 80% (T 80 ) dissolution. Pure DXM showed low solubility, and 100 % dissolution was not possible; however, that was made possible by the incorporation of nf PLGA. With the incorporation of 0.75% and 1.5% of nf PLGA, T 50 decreased from 24.0 to 18.0 min and T 80 reduced from 50.0 min to 35.0 min. Similarly, the initial dissolution rate (or DR) with nf PLGA incorporation increased from 289.7 µg/min for pure dexamethasone (DXM) to 513.02 µg/min when the nf PLGA incorporation was 1.5%.

Dissolution of DXM and DXM-nfPLGA
The in vitro dissolution experiment was based on the United States Pharmacopeia or USP-42 dissolution protocol, and the study was conducted in media that mimicked the gastric pH of 1.4. The enhanced dissolution of the drug composites was attributed to the presence of the hydrophilic nf PLGA, which led to hydrogen bonding (H-bond) with the API crystal, and eventually enhanced dissolution. Figure 9 shows the dissolution profile for dexamethasone (DXM) and the nfPLGAincorporated DXM composites. It is clear that nfPLGA led to enhanced dissolution rate and aqueous solubility. Table 2 presents the enhanced dissolution rates, as well as the time required to reach 50% (T50) and 80% (T80) dissolution. Pure DXM showed low solubility, and 100 % dissolution was not possible; however, that was made possible by the incorporation of nfPLGA. With the incorporation of 0.75% and 1.5% of nfPLGA, T50 decreased from 24.0 to 18.0 min and T80 reduced from 50.0 min to 35.0 min. Similarly, the initial dissolution rate (or DR) with nfPLGA incorporation increased from 289.7 µg/min for pure dexamethasone (DXM) to 513.02 µg/min when the nfPLGA incorporation was 1.5%.

Conclusions
The incorporation of nano-formulated hydrophilic functionalized poly(lactic-co-glycolic acid) or nf PLGA significantly enhanced the solubility and the dissolution rate of the hydrophobic drug DXM. The SEM images clearly show the presence of nf PLGA dispersed over the surface of the DXM drug crystal. Raman, FTIR, DSC and XRD data point to the fact that the presence of nf PLGA did not alter the polymorph and even the melting point remained unaltered. Increase in dissolution rate in the presence of a small amount of the hydrophilic nf PLGA was quite pronounced, and consequently the T 50 and T 80 values were significantly lower. Finally, the synthesized drug composite particles showed excellent physiological stability at different pH. Mechanistically speaking, we believe that hydrophilic channels produced by nf PLGA incorporation enhanced intermolecular interaction with water molecules, and this led to faster dissolution of the API. The use of nf PLGA provides an efficient route to drug delivery by increasing the aqueous solubility of hydrophobic molecules. It also requires a minimal amount of the biodegradable polymer, and the process can be easily scaled up. The approach is applicable to other BCS-II and BCS-IV hydrophobic compounds as well. The methodology and the enhanced dissolution are very promising for the drug delivery, and is applicable to bioavailability improvement. Future studies including in vivo measurements are expected to further demonstrate improvement in efficacy.