Keywords: Evaporative precipitation of nanosuspension; nanoparticles; dissolution; glibenclamide; hesperetin; silymarin
Today many of the drug entities chosen for further development are extremely hydrophobic, displaying low or negligible water solubility . Poorly water-soluble drugs tend to be eliminated from the gastrointestinal tract before they get an opportunity to fully dissolve and be absorbed into the blood circulation. This results in low bioavailability and poor dose proportionality, which greatly hinders their clinical translations  reasoned that as about 65% of the human body is made up of water, a drug must have certain water solubility and thus, possess an acceptable bioavailability level. Therefore, poor water solubility of many drugs is one of the major obstacles in the development of highly potent pharmaceuticals. In the present study, we have used three poorly water soluble drugs, namely: silymarin, hesperetin and glibenclamide.
Silymarin and hesperetin are polyphenols. Silymarin, extracted from the seeds of milk thistle (Silibum marianum Gaertn), is a mixture of flavonolignans including silybin, silychristin, isosilybin, silydianin, taxifolin, and various derivatives of these components . Among the isomers, silybin is the major and most active component and responsible for its pharmacological activity. [4,5] highlighted the role of silymarin as a hepatoprotector to treat liver injuries and chronic hepatitis. Its hepatoprotective effect is due to its anti-oxidant and anti-inflammatory properties as demonstrated by [6,7]. However, the main problem with silymarin is its poor oral bioavailability (23–47%) which is attributable to its poor solubility  and low permeability as well as degradation in the gastrointestinal tract . The solubility of silymarin in distilled water was reported to be 58 μg/ml at 25 °C and can be considered as a practically insoluble drug . Hence, silymarin is required in a large dose to achieve therapeutic plasma levels. Hesperetin is a kind of flavonoid that exists ubiquitously in plants, fruits and flowers [11-13] have shown that hesperetin is a powerful radical scavenger that promotes cellular antioxidant defence-related enzyme activity. Besides acting as an antioxidant , hesperetin is also a potential anti-inflammatory  and anticancer agent . In spite of this wide spectrum of pharmacological properties, hesperetin has a low bioavailability due to its poor aqueous solubility and slow dissolution rate in the aqueous gastro-intestinal fluids, which greatly restricts its use in therapy according to . The solubility of hesperetin as reported by  is around 15 μg/ml at 25 °C. Formulations that improve the solubility of this drug may, therefore, provide improved therapeutic options for patients to achieve higher oral bioavailability. Glibenclamide, on the other hand, is a second-generation sulfonylurea and is a widely employed oral hypoglycemic drug for the treatment of non-insulin-dependent diabetic patients (Type 2 diabetes). It causes hypoglycemia by stimulating release of insulin from pancreatic β cells and by increasing the sensitivity of peripheral tissue to insulin . Clinical trials for glibenclamide have shown its therapeutic effects to control glucose level . Besides being antidiabetic, glibenclamide has been shown to exhibit in-vivo antiplatelet potential as well and thus, suggested to protect against thrombosis development . Glibenclamide is known to have a poor bioavailability, which is attributed to poor dissolution in the body and hence it is not fully absorbed by the body. Although therapeutic effects of glibenclamide being used to treat diabetic patients show significant decrease of blood glucose , the drug remains highly insoluble in water (about 2-4 μg/ ml). Therefore improvements in the solubility and dissolution of the drug will greatly increase the absorption by the human body.
Introducing an upgraded or advanced formulation significantly lowers the time, risk and money expended in drug development, in contrast to developing completely new drugs. According to , a classical formulation approach for the poorly soluble drugs is to prepare smaller drug particles.  Have agreed and stated that the intention for reducing the particle size is to enlarge the surface area of the drug particles available for dissolution. At present, several new drug candidates possess very poor solubility and micronization does not bring those drugs to achieve satisfactory enhancement in the bioavailability. Subsequently, the choice is to go from micronization to nanonization to produce drug nanoparticles. The very small particle size results in a large surface area (Figure 1) and thus in an increased dissolution rate. As predicted by the  model of dissolution wherein the surface area of the drug is directly proportional to its rate of dissolution, drug particles in the nanometer size range will dissolve much more rapidly than a conventional formulation. Besides faster drug dissolution, nanoparticles also exhibit some interesting surface properties due to their very small size as emphasized by . Particles having a size below 100 nm can evade mechanical filtration from the blood stream and accumulation in liver, spleen and kidneys, resulting in longer blood circulation times. Nanoparticles are able to deliver active pharmaceutical ingredients across a number of biological barriers, i.e., the blood-brain barrier , different types of mucosa  and epithelia , and cell membranes for transfection applications. Nanoparticles also show excellent adhesion to biological surfaces such as the epithelial gut wall, which is an advantage for sustained drug delivery. Positive effects of nanoparticles can be obtained for a wide range of drugs having diverse structures, molecular weights, steric groups and solubilities. Moreover, employing drug nanoparticles presents an added benefit of elevated mass per volume loading, which is important when a high dosage is needed. Such cases often fail with approaches involving molecular complexation (for example, cyclodextrins) because a high molar ratio of complexing excipient must be used. Fast dissolution of nanoparticles inspires us to fabricate a drug into nanoparticles so the drug will dissolve quickly.
In this study, the method called evaporative precipitation of nanosuspension (EPN) was used to fabricate drug nanoparticles. It is a new method which relies on high supersaturation and homogeneous nucleation to obtain drug nanoparticles. In this process, after adding hexane as an antisolvent to the drug solution, a uniform nanosuspension was obtained, which was independent of the agitation speed and flow rate. Filtering such a nanosuspension did not yield any particles as they were too small. Therefore, the EPN process was devised to extract the drug particles from the nanosuspension. The lower boiling point and heat of vaporization of hexane made it possible to quickly evaporate the drug nanosuspension to obtain the drug nanoparticles. The two process parameters studied were drug concentration in the solution and the solvent to antisolvent (SAS) ratio. Although the EPN method requires use of organic solvents, in our previous study we have shown that the amount of the residual solvents (ethanol and hexane) in the samples prepared by EPN was below the acceptable level for residual solvents in pharmaceuticals as determined by FDA for the safety of the patients . For example, in the quercetin samples prepared the content of hexane was below 125 ppm and that of ethanol was below 20 ppm as determined by gas chromatography. According to FDA, hexane is a Class 2 solvent, whose content should be limited (≤ 290 ppm) and ethanol is a Class 3 solvent with a minimum content of 5,000 ppm. EPN represents an efficient method to obtain uniform drug nanoparticles with a significantly higher rate of dissolution in-vitro and hence, is used in the present study to fabricate drug nanoparticles. The present study focuses on fabrication of nanoparticles of poorly water soluble drugs: silymarin, hesperetin and glibenclamide by the EPN method.
Silymarin (SLM), hesperetin (HSP) and glibenclamide (GLB) were purchased from Sigma Aldrich Singapore. Organic solvents such as ethanol (99.9%), acetone, methanol and hexane were obtained from Merck Singapore. Deionized (DI) water (Milli-Q, Millipore Singapore) was used.
Pure drug powder was dissolved in a solvent to prepare a 5-15 mg/ml drug solution. SLM was dissolved in acetone, HSP in ethanol and GLB in methanol respectively. Later the antisolvent (hexane) was added so as to get a nanosuspension. The ratio of solvent to antisolvent was varied from 1:10 to 1:20. For this process, the condition is that the solvent and the antisolvent should be miscible. The solvents used such as ethanol and acetone, were miscible with hexane. For methanol, although it is immiscible with hexane at 1:1 ratio but in our experiments using methanol to hexane ratio of 1:10 to 1:20 they formed homogeneous solutions, as has been reported earlier by .
Later, the drug nanosuspension prepared in a round bottom flask was attached to a rotary evaporator. The flask was immersed partly in water bath at 40°C and set to a rotation of 90 rpm. The evaporation occurred at 300 mbar. The solid in the flask after evaporation was vacuum dried overnight at 40°C and the solid particles in the flask were collected for further analysis. Parameters such as drug concentration in solution and solvent to antisolvent (SAS) ratio were optimized.
Scanning Electron Microscopy (SEM): The morphology of samples was observed using a scanning electron microscope (JSM- 6390LA-SEM, Jeol Co., Tokyo, Japan). The powder samples were spread on a SEM stud and sputtered with gold before the SEM observations. The analysis of the particle size was performed using the UTHSCSA ImageTool program. Five or six high resolution and high magnification SEM pictures representative of the sample were used to find the particle size. The software was used to measure the average size of the particles seen in each of the SEM picture.
Differential Scanning Calorimetry (DSC): Differential scanning calorimetric (DSC) measurements were carried out using a TA DSC 200 thermal analyzer in a temperature range of 25 – 250°C at a heating rate of 10°C / min in nitrogen gas. About 5 to 10 mg of the sample mass was sealed in a closed aluminium pan for the DSC scan and an empty aluminum pan was used as the reference pan. The melting point and heat of fusion were calculated using the Universal Analysis (TA Q Series Advantage) software.
Dissolution rate: The in vitro dissolution of the samples was determined using the paddle method (USP apparatus II) (Verkin Dissolution Tester DIS 8000) in 900 mL of DI water. The paddle rotation was set at 100 rpm and the temperature was maintained at 37 ± 0.5°C. The dry powder samples weighing about 50 mg, 10 mg and 5 mg for SLM, HSP and GLB respectively, were added to the dissolution vessel. About 1 mL of the dissolution media was collected at regular intervals and filtered for further analysis. For each sample the dissolution test was repeated for 3 times and the dissolution data were then averaged.
UV spectroscopy: The concentrations of drugs were analyzed using a UV spectrophotometer (UV-2102, Shanghai Instrument Ltd, China). SLM, HSP and GLB were analyzed at the detection wavelength of 286 nm, 323 nm and 299 nm, respectively.
Mathematical modeling of release kinetics
The Noyes -Whitney equation provides a general guideline as to how the dissolution rate of an insoluble drug might be improved by reducing the particle size of a solid drug. The dissolution rate equation based on mass is expressed as follows:
Where m is the dissolution amount of a drug at time t, dm/dt is the dissolution rate, MS is the dissolution amount at infinite time, and t is the dissolution time. Integrating Equation1 with the initial condition of m = 0 for t = 0, then Equation 2 is obtained.
Diving both sides by MS, and we get Equation 3:
Here, the dissolution rate constant K is defined as AD/h, where A is the surface area available for dissolution, D is the diffusion coefficient of the drug, and h is the thickness of the diffusion boundary layer adjacent to the surface of the dissolving drug.
As seen from Figure 2 and Table 1, original SLM powder exhibited the irregular shape and much bigger size (15.97 μm) than the SLM nanoparticles prepared by the EPN method. It can be observed from Figure 2 that the SLM particles prepared by EPN were uniform and small (350 nm to 670 nm). The process parameters for all the samples prepared are listed in Table 1 along with the resulting particle sizes. The sample 3 exhibited the smallest particle size of 350.2 nm. Decreasing the drug concentration and increasing the SAS ratio resulted in smaller particle size. This observation can be interpreted by the formation of nanoparticles through homogeneous nucleation. A higher initial concentration and larger solvent to antisolvent ratio may lead to greater supersaturation, which favors the formation of a large number of nuclei . A larger number of nuclei mean smaller sized nuclei. Although a high initial concentration helps in greater supersaturation, it also has the opposite effect of causing aggregation of particles. In our case, the aggregation effect dominates and hence, a lower drug concentration produces smaller particles. Once nuclei are formed, growth occurs simultaneously. For the subsequent growth a large solvent to antisolvent ratio increases the diffusion distance for growing species and consequently molecular diffusion becomes the limiting step for nuclei growth as also observed in our previous studies [32,33].
|Samples||Drug Conc.||SAS||Particle Size (nm)||ΔHf (J/g)||Dissolution (%)||K(h-1)||R2|
|15,970||175.2||21.9 ± 2.2||0.058||0.84|
|1||5||1:10||670.8||99.3||78.2 ± 4.1||0.345||0.85|
|2||5||1:15||590.7||95.6||80.5 ± 3.9||0.358||0.84|
|3||5||1:20||350.2||74.8||95.2 ± 3.2||0.646||0.900|
|4||10||1:20||465.0||81.2||93.6 ± 3.8||0.604||0.93|
|5||15||1:20||530.4||88.5||90.3 ± 3.3||0.518||0.94|
|34,000||302.0||4.72 ± 1.0||0.011||0.71|
|1||5||1:10||890.3||240.6||60.0 ± 3.0||0.218||0.74|
|2||5||1:15||720.6||190.1||75.6 ± 3.1||0.325||0.71|
|3||5||1:20||450.2||147.9||90.0 ± 3.2||0.516||0.79|
|4||10||1:20||545.4||140.1||85.9 ± 3.0||0.443||0.75|
|5||15||1:20||600.7||137.7||84.0 ± 2.9||0.417||0.73|
|111,200||287.0||10.9 ± 2.0||0.027||0.85|
|1||5||1:10||226.0||165.2||83.6 ± 4.2||0.462||0.96|
|2||5||1:15||185.2||110.9||90.0 ± 4.5||0.605||0.95|
|3||5||1:20||120.0||84.6||99.6 ± 4.0||1.345||0.991.007|
|4||10||1:20||140.3||93.7||98.1 ± 2.2||1.007||0.99|
|5||15||1:20||150.1||99.1||96.3 ± 3.2||0.845||0.98|
Table 1: Preparation parameters namely drug concentration (mg/ml) and solvent to antisolvent ratio (SAS), and properties like mean particle size, melting enthalpy (ΔHf) and the dissolution parameters: percent dissolution (at 4 hours), dissolution rate constant K(h-1) and the correlation coefficient R2 of the original drugs (SLM, HSP and GLB) and their nanoparticles prepared by EPN.
In order to understand the effect of the fabrication method on the thermal properties of SLM, DSC measurements were conducted. The DSC thermograms (Figure 3) and the melting enthalpy ΔHf values are given in Table 1. The original SLM had a melting peak at 137 °C and a melting enthalpy of 175 J/g, whereas the EPN prepared SLM nanoparticles had the melting peaks around ~134-135 °C and the melting enthalpy was lower than that of the original SLM. From these results it can be concluded that the crystallinity of the SLM particles prepared by the EPN method was reduced compared to the original SLM powder.
The dissolution tests were also performed for the original SLM and the EPN prepared samples and their dissolution profiles (Figure 4). The dissolution of the EPN prepared samples was much faster than the original SLM. Only about 21.9% of the original SLM dissolved within 4 hours as compared to 78% - 95% of dissolution for the EPN prepared samples. The sample 3 with the smallest particle size showed the greatest increase (~5 folds) in the dissolution rate of SLM compared to the original SLM powder. This improvement in the dissolution rate is attributed to the reduction in the particle size of SLM particles prepared by the EPN method.
The dissolution rate constant K for the original drug and the nanoparticles prepared by EPN was obtained according to Equation 3. The calculated dissolution rate constant K, along with the correlation coefficient R2, are listed in Table 1. The R2 value of more than 0.8 indicated good correlation between the Noyes-Whitney equation and our measured dissolution data. The value of dissolution rate constant K increased with the decrease in the particle size. According to the Noyes- Whitney equation, the dissolution rate of a drug can be increased by reducing the particle size to increase the particle’s surface area available for dissolution. An increase in dissolution velocity and saturation solubility can also be achieved by changing the crystalline state of the material (e.g. from crystalline to amorphous or partially amorphous) . Proposed that an amorphous or metastable form will dissolve at a faster rate because of its higher internal energy and greater molecular motion, as compared to crystalline materials. Our DSC study revealed that the crystallinity of SLM was reduced after the preparation by the EPN method, which also contributed to a faster rate of dissolution for the SLM nanoparticles.
The preparation conditions and properties of the original HSP powder and the HSP samples prepared by EPN are listed in Table 1. It can be observed from Figure 5 that as the SAS ratio increased from 1:10 (sample 1) to 1:20 (sample 3), the particl e size decreased from 890 nm to 450 nm, and it increased to 600 nm as drug concentration increased from 5 mg/ml (sample 3) to 15 mg/ml (sample 5). The DSC thermograms of the original HSP and its particles prepared by the EPN (Figure 6) and the melting enthalpy (ΔHf) obtained from the DSC study are summarized in Table 1. The original HSP used in this study had a sharp melting endothermic peak at 230°C and a melting enthalpy of 302 J/g. It can also be observed that the EPN prepared HSP nanoparticles had a melting point of about 230°C, very similar to the original HSP, but the melting enthalpy was lower than that of the original HSP. These results suggested that the crystallinity of HSP was decreased after it was prepared by the EPN method.
Next, the dissolution properties of the HSP samples were studied by the dissolution test. As seen from the dissolution profiles of the samples in Figure 7, only about 4.7% of the original HSP dissolved within 4 hours as compared to 60 to 90% of dissolution for the EPN prepared HSP samples. This represents a 12 to 20 times increase in the percent dissolution of HSP. In accordance with the Noyes-Whitney equation, the dissolution rate constant K increased with the decrease in the particle size as seen from Table 1. The best dissolution was shown by sample 3, which was attributed to the maximum reduction in particle size which led to an increase in the exposed surface area available for dissolution. The DSC study revealed that the crystallinity of EPN prepared HSP nanoparticles was lower than that of the original HSP, which also contributed to the faster rate of dissolution for the HSP nanoparticles prepared.
The effects of drug concentration and solvent to antisolvent (SAS) ratio on the fabrication of GLB nanoparticles by the EPN process were also examined as described in Table 1. The SEM microphotographs of the EPN prepared GLB particles under variable conditions (Figure 8). It is observed from Figure 8 that the EPN prepared GLB particles were in the size range of 120 – 226 nm. The EPN prepared drug nanoparticles were more uniform and the uniformity was more prominent at a higher solvent to antisolvent ratio. The size of GLB nanoparticles tended to decrease at lower drug concentration and higher solvent to antisolvent ratio. In order to understand the effect of the EPN process on the thermal properties of GLB, DSC measurement was conducted and the melting heat (ΔHf) values obtained from the DSC study are summarized in Table 1. The heat of fusion of the original GLB powder was higher than that of the EPN prepared GLB particles. These results suggest that the crystallinity of GLB particles was decreased when the particles were prepared by the EPN method. Table 1 shows the experimental data for the dissolution of the original GLB powder and EPN prepared GLB particles, and the dissolution profiles for the samples (Figures 9 and 10). The dissolution of original GLB powder in water was very low so that after 4 hours only about 10.9% of the original GLB powder was dissolved. For the EPN prepared samples the dissolution rate was better. The best dissolution was observed for the EPN prepared sample 3, with 99.6% drug dissolved after 4 hours. This is an approximately 9 times substantial increase in the percent dissolution. As seen from Table 1, the R2 value of more than 0.9 indicated good correlation between the Noyes-Whitney equation and our measured dissolution data, and the dissolution rate constant K increased with the decrease in the particle size. The particle size of original GLB powder was reduced from 111.2 μm to a size in a range of approximately 120-226 nm by the EPN process along with the reduction in the crystallinity proved by the lower ΔHf values shown in Table 1. Both these factors resulted in an increased percent dissolution for the GLB nanopaticles by 7.7 to 9 times compared to the original GLB powder.
Overall, the EPN method developed is very simple but has the limitation of using organic solvents. Although the EPN method requires use of organic solvents but it should be noted that the amount of the residual solvents (ethanol and hexane) in the samples prepared by EPN was below the acceptable level for residual solvents in pharmaceuticals as determined by FDA for the safety of the patients. If the solvents used during the process can be recycled and reused economically, it will greatly increase the benefit of the method. Therefore, EPN represents an effective method to obtain uniform drug nanoparticles with a significantly higher rate of dissolution in-vitro. The scaling up would require a low-pressure set up for quick evaporation of the solvents. Large surface area of the set up and smaller batches of the nanosuspension would ensure faster evaporation and recovery of the nanoparticles. In comparison to the existing techniques for production of drug nanoparticles, the EPN method has the advantage of preparing drug nanoparticles without using any toxic surfactants or stabilisers.
The drug nanoparticles prepared can be used in the capsules or tablet form for oral drug delivery. The next step in our study is to evaluate the stability of these nanoparticles in the formulation. The particle size and zeta potential will be monitored over time for further studies.
This study demonstrated that the evaporative precipitation of nanosuspension (EPN) method is able to prepare nanoparticles of several poorly water soluble drugs such as silymarin, hesperetin and glibenclamide. The method developed is simple and cost effective. The process parameters, such as drug concentration and solvent to antisolvent volume ratio were investigated and optimized to produce the smallest drug nanoparticles. The drug nanoparticles prepared by the EPN method exhibited lower crystallinity compared to the original drug powder as revealed by the DSC analysis. The percent dissolution of drug particles depended on particle size and crystallinity. The EPN prepared particles of the drugs presented the considerable reduction in particle size and consequently the significant enhancement in the dissolution rate compared to the original drug powders. To conclude, EPN represents an efficient method to obtain uniform drug nanoparticles with a significantly higher rate of dissolution than original drug powders. The higher dissolution in-vitro can translate into an increased bioavailability in-vivo. The drug nanoparticles produced by this systematic method would have high potential for delivery in much smaller doses compared with commercial preparation containing the normal form of the drug.