The drug crystallinity and possible solid-state structrure of MNZ present in the microparticles were investigated using a combination of XRD, DSC, Raman spectroscopy and FT-IR spectroscopy.
The XRD spectrum of P microparticles, comprising solely PEG (Table I), was similar to that of the raw PEG (P*), indicating that any effect of the spray congealing process on PEG was transitional. Upon melting, PEG molecules were less organised, but upon solidification when the temperature was lowered, the molecules rearranged themselves to form helices that aligned in an orderly fashion, indicating that PEG crystallised easily (21). The PM microparticles consisting of MNZ showed lower MNZ peak intensity in comparison with the corresponding physical mixture (PM*). There were no new peaks observed, indicating the absence of MNZ polymorphic transition. This highlights the ability of the spray congealing process in reducing the drug crystallinity, thereby potentially enhancing dissolution of the drug. The reduction in drug crystallinity was also affirmed in other studies using olanzapine, felodipine, avobenzone and salbutamol sulphate (22–25). However, there were also studies which showed that the spray congealing process had little impact on drug crystallinity of carbamazepine, praziquantel, glimepiride, verapamil or theophylline (15,16,26–28). Various factors can affect the crystallinity of the drug such as the viscosity of the carrier used and the cooling rate of the spray congealing process. Viscosity of a formulation can retard the diffusion of the molecules towards the crystal interface for crystal growth to occur, thereby preventing the rearrangement of drug molecules and resulting in lower drug crystallinity (5). Upon rapid cooling in the spray congealer, the drug molecules, having little time to rearrange themselves, would largely orientate in a haphazard manner. Thus, most of the drug would become amorphous particles or exist as a molecular dispersion, leading to a reduction in the drug crystallinity (29,30). Hence, the crystallinity of a drug may be dependent on the process method besides the formulation employed.
The MNZ peak intensity at 12.5° for PMH5 (417), PMH10 (458) and PMH15 (594) microparticles was significantly lower (p < 0.05) than that of the PM (901) microparticles with equivalent amount of MNZ but without HPMC. Compared to the corresponding physical mixture, the percentage reduction in MNZ crystallinity in microparticles of PM, PMH5, PMH10 and PMH15 was 75.3, 85.0, 83.5 and 78.6%, respectively. Collectively, the results showed that the spray congealing process reduced the MNZ crystallinity to a much greater extent than HPMC. The reduction in MNZ crystallinity by HPMC was dependent on the concentration of HPMC. Several other studies have shown the ability of HPMC to decrease the crystallinity of the drug such as albendazole, felodipine, nicotinamide, acetaminophen, carbamazepine and trans-resveratrol by incorporating into the drug crystal lattice, having molecular interaction with the drug molecules or adsorbing onto the surface of the drug, thereby preventing crystal growth (9–11,31–33). However, the addition of HPMC into the formulation does not necessarily lead to a reduction in drug crystallinity as seen in other studies on omeprazole and prednisolone (34,35).
The XRD data were further supported by the results of the DSC studies. Small MNZ endothermic peaks were observed for the physical mixtures PMH10* and PMH15*, with enthalpy of 4.46 and 4.60 J/g, respectively. The higher concentration of HPMC in the formulation prevented complete dissolution of the drug in PEG. The corresponding PMH10 and PMH15 microparticles also showed very small MNZ endothermic peaks, but their enthalpy of 3.21 and 3.64 J/g, respectively, were significantly lower (p < 0.05). This significant reduction in enthalpy of fusion in the microparticles compared to the corresponding physical mixture indicated decrease in MNZ crystallinity (36,37).
The drug can be present as small crystalline particulates, amorphous particulates or it may be molecularly dispersed within the PEG matrix (38,39). Therefore, Raman spectroscopy was used to further elucidate the solid state structure of the drug and the homogeneity of the drug distribution in the microparticles. The characteristic peak of the NO2 (N–O stretching) of MNZ at 1528 cm−1 was observed in the spectra of microparticles with MNZ, indicating the presence of MNZ on the microparticle surface. This suggests that the drug was not totally embedded in the core of the microparticles when spray-congealed and that the concentration of HPMC up to 15% was not sufficient to fully coat the microparticle surface. Materials of crystalline structure are characterised by distinct and well-defined peaks in the Raman spectrum. The presence of drug peaks in the XRD and Raman spectra and endothermic peak in DSC curves indicate the presence of crystalline MNZ in the microparticles. If the percentage of the drug is high, it will form small crystals within the microparticles (40). In light of this, solubility of MNZ in PEG was determined and found to be 10.67% (w/w) at 80°C. The theoretical percentage weight of the drug with respect to PEG in PM (11.11%), PMH5 (11.76%), PMH10 (12.50%) and PMH15 (13.33%) was higher than the solubility limit. This aptly explains the presence of small crystals in those microparticles, accounting for peaks in the Raman spectra. Taking into account that there was a decrease in drug crystallinity, the remaining portion of MNZ within the microparticles could exist in the amorphous form and/or dispersed molecularly in PEG.
Drug–matrix interactions, such as hydrogen bonding, can be used to demonstrate the existence of drug as a molecular dispersion in the matrix (39). Both Raman spectroscopy and FT-IR spectroscopy were used to investigate the possible interactions of MNZ with PEG or HPMC. Collectively, the MNZ-PEG interaction observed in PM indicates the existence of MNZ as a molecular dispersion in the PEG matrix. Formation of hydrogen bonds between HPMC and MNZ further hindered recrystallisation of the molecularly dispersed drug as indicated by the reduction in MNZ crystallinity in the earlier XRD and DSC studies (Figs. 2 and 5).
Mean percentage reduction in MNZ crystallinity of freshly spray-congealed (week 0) microparticles without HPMC (control) was approximately 75%. This demonstrates that the spray congealing process was able to alter the solid state of the drug, leading to a reduction in drug crystallinity. In addition to MNZ–PEG interaction, the rapid cooling and solidification of the molten droplets also hindered the drug molecules from rearrangement into its crystalline form. Spray-congealed microparticles were also stable over 3 months (week 12) when stored at 25°C and 30% RH as the percentage reduction in drug crystallinity at different time intervals did not change significantly. Metronidazole largely remained in the non-crystalline form after spray congealing.
Generally, all formulations containing HPMC exhibited a greater reduction in drug crystallinity. As mentioned previously, the MNZ–PEG interaction enabled the drug to exist as a molecular dispersion, reducing MNZ crystallinity of the spray-congealed microparticles. The drug could also exist as amorphous particles. HPMC particles could serve as a nucleating site, allowing MNZ molecules to deposit on the surface of the particle and orientate in a random and haphazard manner, leading to formation of amorphous particles. Formulations A-5 and A-10% showed the greatest reduction in drug crystallinity (82.9 and 81.9%, respectively) compared to all other formulations. It could be postulated that smaller HPMC particles (33 μm) delayed nucleation and crystal growth of MNZ more than larger particles. At the same concentration, HPMC with a smaller particle size would be present in a larger number leading to a higher probability of the MNZ molecules adsorbing onto the surface of the particles. Hence, fewer crystals were formed, resulting in lower drug crystallinity. Larger HPMC particles of 59 μm and above decreased the crystallinity of MNZ to a similar extent. Therefore, there was a critical HPMC particle size, beyond which changes in HPMC particle size had little effect on MNZ crystallinity.
No significant differences in the percentage reduction in drug crystallinity were observed between 5 and 10% HPMC concentration of the same mesh size, leading to the inference that particle size played a greater role in reduction of drug crystallinity than the HPMC concentration studied.
Even though MNZ is poorly water soluble, the particles dissolved rapidly due to the small particle size of the drug and the sink condition of the dissolution medium. Therefore, marked increase in dissolution rate of MNZ from the spray-congealed microparticles would not be expected.
It was clearly seen that the PM, PMH5 and PMH10 microparticles showed comparable dissolution rates which were slightly greater than that of the pure MNZ. This shows that spray congealing is able to enhance the dissolution of MNZ through reduction in MNZ crystallinity to a certain extent. HPMC is commonly used in preparations to modify drug release. Addition of 5 or 10% w/w HPMC (PMH5 and PMH10) further reduced the MNZ crystallinity by 3 to 10%, but this did not further enhance the release of MNZ.
Burst effect of drug was observed for all microparticles, indicating that drug particles were present on the outer layer of the microparticle, which dissolved rapidly upon contact with the dissolution medium. The presence of MNZ on the surface of the microparticles was previously indicated by Raman spectroscopy. The results showed that the barrier to drug release did not form spontaneously upon hydration. The HPMC particles required time to hydrate and swell, to form the barrier. Interestingly, PMH15 was found to have a slower dissolution than PM, PMH5 and PMH10. It was possible that the concentration of HPMC (15% w/w) in PMH15 was sufficient to form a layer of gel that acted as a barrier to drug release. Upon hydration, the individual HPMC particles swelled to form clusters. These clusters would eventually merge, forming a barrier surrounding the microparticle, thereby modifying drug release.
When low additive concentration was used, formulations with sodium oleate/citric acid or MgSt had the fastest drug release. The coat layer formed by sodium oleate/citric and the hydrophobic MgSt could have retarded the wetting of HPMC particles by water. Without the wetting and swelling of HPMC particles, no barrier could be formed. Formulations containing PVP K90, MC or SiO2 all showed faster drug release than PMH15 while the formulation with DCP showed the opposite. Being hydrophilic, PVP K90 attracted water into the PEG matrix and the amount present was probably insufficient to exact the desired binding and gelling effect. Similarly, the hydrophilic SiO2 and MC attracted water in the PEG matrix, causing it to disintegrate before the HPMC particles could hydrate to form a barrier. On the other hand, the insoluble DCP reduced the hydration of the PEG matrix to an appropriate degree to prevent its disintegration, which allowed the HPMC particles to have sufficient time to hydrate, swell and form the barrier.
At 5%, MgSt resulted in faster drug release than 1%. This could be attributed to the greater hydrophobic effect of a higher concentration of MgSt. Thus, the hydrophobicity of an additive has to be carefully considered as it would affect the hydration of HPMC and consequently the drug release rate. At 5%, the hydrophilic SiO2 attracted more water into the PEG matrix than 1%, causing the rapid disintegration of the PEG matrix and resulted in even faster drug release. On the other hand, MC produced a greater gelling effect at higher concentration, leading to slower drug release than 1% MC. More interestingly, the addition of DCP at 5% markedly reduced the release of drug, with 87.7% MNZ released at 60 min. The hydration and disintegration of the PEG matrix was markedly retarded by the higher concentration of DCP, allowing more HPMC to swell and form a more effective barrier.
From the above results, hydrophilic non-gelling additives could enhance the hydration rate of the microparticles, but they were ineffective for decreasing the gellation time of HPMC to form a barrier because of the rapid disintegration of the PEG matrix. The latter could be achieved by employing a high concentration of DCP. DCP reduced the hydration rate and retarded the disintegration of the PEG matrix, thus allowing sufficient time for HPMC to swell and form a barrier.