Article pubs.acs.org/molecularpharmaceutics
Molecular Implications of Drug−Polymer Solubility in Understanding the Destabilization of Solid Dispersions by Milling Ziyi Yang,†,§ Kathrin Nollenberger,‡ Jessica Albers,‡ and Sheng Qi*,† †
School of Pharmacy, University of East Anglia, Norwich, Norfolk NR4 7TJ, U.K. Evonik Industries AG, Kirschenallee, Darmstadt, Germany § School of Pharmacy, University College London, 29-39 Brunswick Square, London WC1N 1AX, U.K. ‡
S Supporting Information *
ABSTRACT: The solubility of drugs in polymer matrixes has been recognized as one of the key factors governing the physical stability of solid dispersions. This study has explored the implications of drug solubility on the destabilization that occurs on milling, which is often used as an additional process for hot melt extruded (HME) solid dispersions. The theoretical drug solubility in the polymer was first predicted using various theoretical and experimental approaches. The destabilization effects of high-energy mechanical milling on the solid dispersions with drug loadings below and above the predicted solubility were then investigated using a range of thermal, microscopic, and spectroscopic techniques. Four model drug−polymer combinations were studied. The HME formulations with drug loading below the predicted solid solubility (undersaturated and true molecular dispersion) showed good stability against milling. In contrast, milling destabilized supersaturated HME dispersions via increasing molecular mobility and creating phase-separated, amorphous, drug-rich domains. However, these additional amorphous drug-rich domains created by milling show good stability under ambient conditions, though crystallization can be accelerated by additional heating. These results highlighted that the processing method used to prepare the solid dispersions may play a role in facilitating the stabilization of amorphous drug in supersaturated solid dispersions. The degree of supersaturation of the drug in the polymer showed significant impact on the destabilization behavior of milling on solid dispersions. An improved understanding of the destabilization behavior of solid dispersions upon milling can provide new insights into the processing related apparent solubility of drugs in polymers. KEYWORDS: milling, amorphous molecular dispersions, solubility prediction, crystallization, hot melt extrusion miscibility and solubility.1,2,6 Most of these reported predictive studies are based on Flory−Huggins theory, which was originally developed for polymer blend systems.7,8 The most commonly used prediction methods include calculation of solubility parameters, melting point depression method, and the recently reported thermal-based melting enthalpy method and milling method.1−3,6,9 However, each of these methods requires assumptions to be made, which often limit their wider application. More importantly, the predictions performed using these methods are based on the behavior of the physical mixtures of the drug and polymer rather than the formed solid dispersion. An increasing number of the studies have demonstrated that the processing method, such as hot melt extrusion, film casting, spray drying, and electron-spinning, can have profound effects on the kinetic physical stability of the resultant solid dispersion
1. INTRODUCTION Despite the promising in vitro and in vivo performance of many solid dispersion based formulations, the physical stabilities of solid dispersions are still not well understood. The physical stability of pharmaceutical solid dispersions over their shelf life is a complex interplay of the thermodynamic, kinetic, and molecular (such as inter- and intramolecular interactions) properties of the system. High miscibility and solubility of drug in polymer and molecular interactions, such as hydrogen bonding and acid−base interaction between the drug and polymer at a molecular level, favors the good long-term physical stability of the resultant solid dispersions. Further research on how to better predict and control the physical stability of solid dispersions is required to enable wider application and better commercialization opportunities of these formulations.1−3 The considerable research efforts that have identified the direct implications of drug−polymer miscibility for the static-state physical stability (stability on aging) of drug−polymer solid dispersions have been well recognized.4,5 Many theoretical approaches have been explored to predict drug−polymer © XXXX American Chemical Society
Received: March 17, 2014 Revised: May 12, 2014 Accepted: June 4, 2014
A
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
In this study, we first predicted the drug−polymer solubility of the model systems using established theoretical methods including, calculating solubility parameters, melting point depression, and melting enthalpy methods.1,2,6,9 In order to rapidly induce the destabilization effect, high-energy mechanical milling at room temperature was used in this study. The destabilization effects of milling on HME solid dispersions were examined in relation to the degree of supersaturation of the drug in the dispersions (based on the predicted theoretical solubilities). Felodipine−EUDRAGIT E PO (referred as EPO in the name of blends in this article) was used as a model system to study the correlation between drug−polymer solubility and the destabilization behavior of milling on hot melt extruded solid dispersions. The relationship was hypothesized based on the felodipine−EUDRAGIT E PO data and further validated using another three drug−polymer combinations, celecoxib−EUDRAGIT E PO, felodipine−PVPVA, and felodipine−E PO−PVPVA blends.
formulations.10−15 This may suggest that the apparent solubility of drug in a solid dispersion varies with the processing method used. Most of the predictive methods mentioned earlier are not able to account for the influence of manufacturing processing on the kinetic stability of drug−polymer dispersions.1,6 There has been little effort devoted to the development of appropriate methods to accurately and meaningfully measure the apparent solubility of a drug in a polymer processed by a particular preparation method. This is a result of a lack of knowledge of the relationship between the solubility of a drug in a polymer and the impact of processing on the physical stability of solid dispersions. The aim of this article is to report the implications of theoretically predicted static-state solubility on the destabilization behavior of mechanical milling on solid dispersions. Milling is a common downstream process for hot melt extruded products for particle size reduction.31−33 However, the destabilization effects of milling on drug−polymer solid dispersions is still poorly understood. By increasing the energy input of the milling process, the outcome of milling can be converted from particle size reduction to physical state transformation of the material. High-energy milling is one of the common methods used to induce solid-state transformations between polymorphic forms and from the crystalline to the amorphous state.16−23 Milling can gradually increase the surface defects of a crystalline material and leads to solid-state amorphization via surface defect-induced local melting.24 Many milling parameters have been studied for converting crystalline material into their amorphous form.20,22,25 One important parameter is the in situ milling temperature, as often milling can generate heat and elevate the temperature in the milling chamber. Under cryogenic conditions, milling often can rapidly reduce crystalline lattice arrangement and eventually convert the crystalline material into its amorphous form.26−30 If the milling process is operated at temperatures above and, even in some cases, close to the glass transition of the amorphous material, the amorphized portion of the surface may be reconverted back into its crystalline form.22,27,29,30 Therefore, milling at the temperatures around the Tg of the material can induce rapid destabilization of the amorphous portion of the material. A few studies have investigated the destabilization effect of milling on pure amorphous drug.21 For example, Desprez and Decamps reported that depending on the milling parameters, milling could result in either crystallization of amorphous indomethacin or modifications of the configuration of the glassy state.21 For an amorphous solid dispersion, if the drug is molecularly dispersed in the polymer at concentrations below the solubility of drug in polymer at a fixed temperature, this system is, in theory, thermodynamically stable. At drug concentrations higher than the solubility value, the system is thermodynamically unstable but may be kinetically stable over the typical shelf life of solid pharmaceutical products. Depending on the amount of energy input, milling may mechanically activate the instability of these systems by increasing local molecular mobility (via a localized temperature increase) and promote the aggregation of drug molecules to form amorphous domains, which can develop into nuclei and induce further crystallization. This may be manifested in the form of accelerating phase separation and recrystallization of excess amounts of drug in supersaturated HME solid dispersions. In the case of highly supersaturated solid dispersions, amorphous drug domains may already be present, and milling at temperatures around or above the Tg may promote and accelerate drug crystallization and maximize the destabilization.
2. MATERIALS AND METHODS 2.1. Materials. Felodipine, celecoxib, and EUDRAGIT E PO were kindly donated by Evonik Rohm Co. KG, Darmstadt, Germany. Kollidon VA 64 (PVPVA) was kindly donated by BASF, Ludwigshafen, Germany. 2.2. Hot Melt Extrusion. Hot melt extrusion was performed using a Thermo Scientific HAAKE MiniLab II (Thermo Scientific, U.K.) with corotating twin screws. Crystalline felodipine and EUDRAGIT E PO powder with ratios of 10:90, 30:70, 50:50, and 70:30 (w/w) were premixed in mortar and pestle before melt extrusion. The operation temperature was set at 150 °C with a dwell time in the extruder of 5 min. The rotating speed of the screws was 100 rpm. These operation parameters were optimized in a preliminary pilot study to ensure complete melting without thermal degradation of the crystalline drug and through mixing between the drug and polymer in the extruder. A round shape die with diameter of 2 mm was attached to the extruder. Other model drug−polymer systems including felodipine−PVPVA and felodipine−EPO−PVPVA were prepared with the same processing parameters. For celecoxib−EPO samples, the operation temperature in HME was set at 170 °C. 2.3. Ball Milling. Extrudates with all drug loadings were milled using a Retsch MM400 Ball Miller (RETSCH, Haan, Germany). Felodipine−EUDRAGIT E PO extrudates were milled for 5, 10, 30, and 60 min at the frequency of 30 Hz by a single 7 mm stainless steel ball. The high milling frequency used in this study was selected to maximize any destabilization effect. In the study on the effect of thermal treatment on the postmilled samples, a 10 min milling duration was used for all samples. The particle size of the milled powders used for testing was controlled by a sieve between 63 and 106 μm. The strand form extrudates and milled extrudates were stored under 75% RH at room temperature for up to 6 months and were characterized by modulated temperature differential scanning calorimetry (MTDSC), powder X-ray diffraction (PXRD), attenuated total reflectance-Fourier transform infrared spectroscopy (ATRFTIR), and scanning electron microscopy (SEM) regularly. 2.4. Modulated Temperature Differential Scanning Calorimetry (MTDSC). The modulated differential scanning calorimetry (MTDSC) study was performed using a Q-2000 MTDSC (TA Instruments, Newcastle, USA). Temperature calibrations were carried out using standard materials including octadecane, indium, and tin. For modulated mode, aluminum oxide sapphire was used for heat capacity calibration. The B
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 1. (a) Qi and Belton model prediction result of the miscibility between felodipine and EUDRAGIT E PO, and (b) PXRD results of the freshly milled extrudates with drug loadings from 10 to 70%.
Table 1. Drug−Polymer Solubility Prediction Using Theoretical Approaches and Experimentally Measured Apparent Drug−Polymer Solubility for Solid Dispersions Prepared by HME
APIs
Tg of amorphous drug (°C)
felodipine 46
celecoxib
58
polymer EUDRAGIT PVPVA polymer blendd EUDRAGIT
measured Tg of HME extrudates with 70% drug loading (°C)
solubility by solubility parameter (w/w)a
solubility by melting point depression (w/w)b
solubility by melting enthalpy method (w/w)c
apparent solubility based on real-time stability data (w/w)
37.0 56.7 47.6
25.0% 35.0% 30.0%
30.7% 41.7% 36.2%
20.0% 30.0% 25%
∼69% >70% >70%
60.0
35.0%
41.0%
20.0%
∼57%
Condition of 298.15 K. bCondition of temperature close to melting points of individual drug; pressure under the flow of 50 mL/min nitrogen. c Condition of temperature close to melting points of individual drug. dAll solubility predictions for the polymer blend based systems were the calculated values of the average of the individual solublities predicted in EUDRAGIT E PO and PVPVA. a
modulation parameters used were ±0.318 °C/60 s with a 2 °C/ min underlying heating rate. N2 gas was purged through the DSC cell at a rate of 50 mL/min. TA standard crimped pans were used for all measurements. For each sample, triplicate measurements were performed (n = 3). 2.5. Powder X-ray Diffraction (PXRD). PXRD tests were performed at room temperature with a Thermol-ARL Xtra diffractometer (Thermo scientific, U.K.). Milled samples were placed on a zero background sample holder and incorporated onto a spinner stage. Cu Kα1 was used as the X-ray source (voltage, 45 kV; current, 40 mA). It was mounted with the wavelength of 1.5405 Å. The angular range (3−80° 2θ) was scanned with a step size of 0.01° and time per step of 0.5 s. 2.6. Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR). Infrared spectroscopy was carried out on the IFS 66/S FTIR spectrometer (Bruker Optics Ltd., Coventy, U.K.) fitted with a Golden Gate ATR accessory with heated top plate (Orpington, U.K.). The crystal is a single reflection diamond element. The spectral resolution was 2 cm−1, and 32 scans were taken for each sample. The HME strand form samples and milled extrudates were tested by ATRFTIR spectroscopy. For variable temperature ATR-FTIR spectroscopy (VT-ATR-FTIR), samples were heated from 30 to 170 °C at a ramp rate of 2 °C/min. 2.7. Scanning Electron Microscopy (SEM). The strand form samples and milled extrudates were sputter coated with Au/ Pd. The images of the surface morphology of the samples were taken using a Phillips XL20 SEM (Phillips Electron Optics, Netherlands).
3. RESULTS 3.1. Solubility Prediction Using Drug−Polymer Physical Mixes. The solubilities of two model drugs in the model polymers were first predicted using the conventional solubility parameter and melting point depression methods. The theoretical background of these methods have been described in detail elsewhere; thus, readers are directed to the relevant literature.1,2,6,34 The detailed calculation procedures of these two methods on the model systems are described in the Supporting Information. The recently reported thermal enthalpy method by Qi and Belton was also used to calculate solubility.2 This method is based on deconstructing the melting enthalpy of the drug in the physical mixtures of the drug and polymer into enthalpy contributions of melting and dissolution of drug in polymer during heating.2 An example prediction plot is shown in Figure 1a using felodipine and EUDRAGIT E PO as the model system. It can be seen that this method can not only predict the solubility of drug in polymer but also the solubility of polymer in molten drug. The predicted results are summarized in Table 1. It can be seen that the predicted solubility for both drugs varies with the prediction method used. The predictions given by the melting enthalpy method gave the lowest values for the model drugs in comparison to the other two methods. Taking into account predictions by all three methods, felodipine shows lower solubility in EUDRAGIT E PO than celecoxib and has higher solubility in PVPVA and the polymer blend than EUDRAGIT E PO (Table 1). C
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 2. SEM pictures of freshly prepared and milled HME felodipine−EUDRAGIT E PO extrudates and 6 month-aged milled HME felodipine− EUDRAGIT E PO extrudates under 75% RH at room temperature.
3.2. Destabilization Effects of Milling on Hot Melt Extruded Solid Dispersions. 3.2.1. Effect of Drug Loading. To minimize the static state physical instability on aging, having a drug loading lower than the solid solubility of the drug in the polymer has been suggested as a favorable strategy to form thermodynamically stable dispersions.1,2 In order to study the destabilization behavior of milling, undersaturated HME felodipine−EUDRAGIT E PO solid dispersions (with 10% and 30% drug loading), which should be more stable and supersaturated dispersions (with 50% and 70% drug loading), which are more likely to phase separate and exhibit physical instability, were used as the starting materials for the milling study. As seen in Figure 1b, freshly prepared extrudates that have been milled for 5 min show amorphous PXRD patterns for all formulations (with drug loadings from 10 to 70% w/w). This indicates that no significant amount of felodipine recrystallized during the milling. SEM results show that the freshly prepared hot melt extruded strands with 50 and 70% drug loadings had a few crystal-like particles on the surface of the extrudates (Figure 2a). In comparison, hardly any particle features can be seen on the surfaces of the freshly milled samples for all formulations (Figure 2b). This may indicate that a moderate level of milling can reduce the crystallinity of formulations of freshly prepared samples with high drug loadings. The milled samples were further characterized using MTDSC. The HME solid dispersions (with 10−50% drug loading) before and after milling showed no detectable melting/crystallization transitions during heating and cooling (data not shown) indicating that milling had little impact on the stability of these samples. The 6-month static-state physical stability data of these samples also showed no significant drug crystallization indicating the apparent drug solubility in the polymer after hot melt extrusion is at or above 50%. In fact, less than 1% of crystalline
drug was detected in the unmilled extrudates with 70% drug loading. This indicates that in the freshly prepared extrudates, a 69% drug load was stabilized either as a molecular dispersion or as phase-separated amorphous drug. This is significantly higher than the felodipine in EUDRAGIT E PO solubility predicted by the conventional method (Table 1). This may be a result of two possible reasons: HME processing facilitated the mixing and stabilization of supersaturated solid dispersions (the supersaturation relates to the processing related apparent solubility of drug in polymer in this study), and the theoretical prediction is irrelevant on the time scale of the test. Therefore, real-time physical stability tests were carried out and are discussed in section 3.3. The effect of milling on the formulations with high drug loading (70%) is more profound. As seen in Figure 3a, a minor melting peak, correlating to approximately 0.5% (w/w) crystalline felodipine, can be detected in the HME extrudates with 70% drug loading before milling. However, after milling a clear recrystallization exotherm (upon heating between 70 and 110 °C) and a melting transition of felodipine form I at 137 °C with a much greater enthalpy value can be detected. The melting points of pure felodipine form I and form II are 141−146 °C and 131−135 °C, respectively.35 The lower melting point of the crystalline felodipine detected in the milled extrudates than that of pure felodipine form I is a result of melting point depression caused by the polymer in the solid dispersions.36 The recrystallization exotherm indicates that milling induces the formation of phase-separated amorphous drug-rich domains, which may act as nuclei for crystallization on heating. This result is a clear evidence of the rapid destabilization of milling on the supersaturated HME solid dispersions, which is not observed in the undersaturated (10−30%) and saturated solid dispersions (50%). D
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 3. (a) Total heat flow signals of the MTDSC results of the HME extrudates with 70% drug loading milled for different periods; (b) recrystallization enthalpy of 70% HME felodipine−EUDRAGIT E PO systems milled for different time periods; (c) NH group from felodipine in 70% (w/w) HME felodipine−EUDRAGIT E PO systems milled for different time periods (from top to bottom: crystalline felodipine, milled for 60 min, milled for 30 min, milled for 10 min, milled for 5 min, and amorphous felodipine); (d) comparison of the total heat flow signals of the MTDSC results of the HME extrudates with 70% drug loading milled for 1 and 2 h.
from 3336 cm−1 (pure amorphous felodipine) to 3343 cm−1 in the spectra of freshly prepared strand form extrudates before milling. The shift is likely to be associated with the weak intermolecular interactions, such as hydrogen bonding, between felodipine and EUDRAGIT E PO.2,37 After 5 min of milling, the single NH peak at 3343 cm−1 split into two peaks. They are the peaks at 3370 cm−1 (a small shoulder peak) representing the presence of crystalline felodipine and at 3336 cm−1 associated with the presence of phase-separated amorphous felodipine-rich domains. The peak intensity of the 3370 cm−1 peak shows a subtle increase with increasing the milling time to 10 min; whereas the peak intensity of the 3336 cm−1 peak decreases. No significant changes can be observed in the spectra with increasing the milling time above 10 min, which agrees well with the DSC data. However, with increasing the milling time to 2 h, a profound change is evident in the DSC results of the milled extrudates with 70% drug loading. As seen in Figure 3d, a double crystallization peak with an earlier onset (at approximately 55 °C) and a melting with higher enthalpy values can be observed in comparison to the DSC result of the extrudates milled for 1 h. The origin of the double exothermic peaks is not clear, but they are possibly related to the crystallization of different polymorphic forms of the felodipine. This result clearly indicates that an extended milling time can lead to solid dispersions having a greater degree of instability.
3.2.2. Effect of Milling Time. In order to further examine the effect of milling on the physical stability of the samples, the freshly prepared HME extrudates with 70% drug loading were milled for different periods of time. It can be seen in Figure 3a that as little as 5 min of milling lead to a significant increase in the recrystallization exotherm indicating the presence of considerable amounts of amorphous drug-rich domains generated during milling. The recrystallization enthalpy should be quantitatively proportional to the amount of phase-separated amorphous felodipine-rich domains in the milled extrudates. The enthalpy values of the recrystallization peaks at 70−110 °C reduce with increasing the milling time to 10 min and reach a plateau with milling times up to 60 min (Figure 3b). The total amount of crystalline felodipine on heating (calculated by the total melting enthalpy values) in the milled samples shows a significant increase from 0.5% w/w (unmilled strand form extrudates) to 21.5% (out of 70% total drug loading) after 5 min of milling. With increasing the milling time from 10 to 60 min, no further significant increases in crystallinity were detected, as seen in Figure 3b. This suggests that the amount of crystalline felodipine already present in the HME solid dispersions after milling (recrystallized during milling) also reached a constant level with a milling time between 10 and 60 min. The presence of phase-separated amorphous felodipine-rich domains were further confirmed by ATR-FTIR spectroscopy. As seen in Figure 3c, the NH peak of amorphous felodipine shifted E
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 4. (a) Three-dimensional and (b) 2D spectra of the NH region of freshly prepared 70% HME felodipine−EUDRAGIT E PO extrudates tested as strand form heated up from 30 to 150 °C at the heating rate of 2 °C/min.
Figure 5. (a) Three-dimensional and (b) 2D spectra of the NH region of freshly prepared and milled 70% HME felodipine−EUDRAGIT E PO extrudates heated up from 30 to 150 °C at the heating rate of 2 °C/min.
felodipine) after milling at 30 °C. The separation of the two peaks is less resolved at lower temperatures (below 110 °C) indicating a low level of phase separation. Between 70 and 110 °C, the peak intensity of the 3336 cm−1 peak increased rapidly indicating an increased amount of phase-separated amorphous felodipine-rich domains in the milled sample. This increase is absent in the unmilled sample, thus the destabilization of milling is responsible for this rapid phase separation. This phase separation that occurred at elevated temperatures may also contribute to the observed exothermic peak in the MTDSC results of the freshly milled sample (Figure 3a). A sharp decrease in the peak intensity of the 3336 cm−1 peak can be observed between 110 to 130 °C. The sequence of the changes in the peak intensities of the two peaks representing the amorphous and crystalline felodipine confirmed the presence of phase-separated amorphous felodipine can be increased by heating milled extrudates. This phase-separated amorphous felodipine transforms to crystalline drug on further heating. This confirms that under thermal stress amorphous phase separation and recrystallization of these amorphous domains can be accelerated in milled supersaturated HME solid dispersions.
3.2.3. Effect of Postmilling Thermal Treatment. The significant recrystallization exotherm seen in Figure 3a indicates the high thermal instability of milled extrudates upon heating. In order to further study this physical instability of milling destabilized HME solid dispersions, VT-ATR-FTIR spectroscopy was used for thermal-treatment and real-time monitoring of the freshly milled HME solid dispersions (all dispersions studied in this section were milled for 10 min). In order to compare with the corresponding MTDSC results, the same heating rate as the MTDSC experiments, 2 °C/min, was used. As seen in Figure 4, the freshly prepared HME solid dispersion with 70% drug loading (tested as intact strands) did not show separate crystalline and amorphous felodipine peaks in the NH region. The shift of 3343 cm−1 NH peak to 3358 cm−1 when heated above 130 °C may be associated with the melting of extrudates. However, there is no clear evidence of phase separation. In comparison to the unmilled strand form sample, more complex data profiles were obtained from the freshly milled samples with 70% drug loading. As seen in Figure 5, the single NH 3343 cm−1 peak splits into two peaks at 3370 (crystalline felodipine) and 3336 cm−1 (phase-separated amorphous F
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 6. (a) Total heat flow signals of the MTDSC results of the milled HME felodipine−EUDRAGIT E PO extrudates with 70% drug loading aged for 6 months under 75% RH at room temperature; (b) comparison of the MTDSC results of strand form and milled felodipine−EUDRAGIT E PO extrudates with 50% and 70% (w/w) drug loadings after 6 months aging under 75% RH at room temperature; (c) the calculated total crystallization of felodipine from the melting enthalpy of the MTDSC results of the samples.
3.3. Long-Term Destabilization Effects of Milling on HME Solid Dispersions. The long-term impact of the destabilization effect of milling on the HME solid dispersions was tested by real-time stability studies. As discussed previously,
after as little as 5 min of milling, a significant amount of phaseseparated amorphous felodipine-rich domains can be generated in the supersaturated HME solid dispersions with 70% drug loading. After 6 months aging at room temperature/75% RH, the G
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 7. Partial ATR-FTIR spectra (NH regions of felodipine) of the milled felodipine−EUDRAGIT E PO extrudates with (a) 50% and (b) 70% drug loadings on aging under 75% RH/room temperature.
extrudates with 70% drug loading, the 3370 cm−1 peak associated with the presence of crystalline felodipine shows a clear presence after 1 week of aging, and the peak intensity increases rapidly for samples aged up to 2 months. After 2 months of aging, little change is seen in the peak intensity of the 3370 cm−1 peak with increasing the aging period to 6 months. This confirmed that no further significant drug crystallization occurs with increasing the aging time above 2 months. This agrees well with the DSC and SEM results. The felodipine−EUDRAGIT E PO data indicates that milling directly leads to an increased amount of phaseseparated amorphous drug in the solid dispersion with 70% drug loading (with high degree of saturation in comparison to the predicted drug−polymer solubility of approximately 25−30%) and increased formation of phase-separated drug-domains on heating for the solid dispersions with 50% drug loading (with lower degree of saturation in comparison to the 70% drug loading). This leads to the hypothesis that milling can accelerate amorphous phase separation in supersaturated HME solid dispersions. However, the crystallization tendency of the phase-separated amorphous drug-rich phases may be related to the intrinsic crystallization behavior of the pure amorphous drug and the stabilization capability of the polymer. This was further tested using different solid dispersions and is discussed in the next section. 3.4. Validation of the Milling Destabilization Hypothesis. The destabilization effect of milling was further tested on other model systems (celecoxib−EUDRAGIT E PO, felodipine−PVPVA, and felodipine−EPO−PVPVA blend) in order to validate the hypothesis presented in the previous section. These systems have distinct differences in their physical properties, such as the physical stability of the drug alone, Tg, and hygroscopicity of the polymers and the complexity of the matrix system (polymer blend instead of a single polymer system). For these three model systems, no destabilization effect by milling was observed in the extrudates with drug loading below the predicted solubilities of drug in the polymer (data not shown). Celecoxib has a slightly higher Tg than felodipine (Table 1) and the predicted solubility values in EUDRAGIT E PO using the three theoretical based methods mentioned previously were higher than the solubility values of felodipine in the polymer. Therefore, the expected physical stability and apparent solubility of the drug in polymer should be higher than felodipine− EUDRAGIT E PO extrudates. However, the freshly prepared
growth of crystal-like particles was observed on the surfaces of the milled extrudates (Figure 2d). However, it was noted that the size and density of the crystals that appeared on the surface of the milled extrudates were similar to those observed at the surfaces of unmilled and aged extrudates (Figure 2c). The MTDSC results of the milled and 6 month-aged extrudates show significantly higher recrystallization exotherm and melting enthalpy values in comparison to the unmilled and aged extrudates (Figure 6a). The recrystallization is absent in the aged and unmilled extrudates. The higher melting enthalpy values of the milled and aged samples are likely to be a result of the drug that recrystallized during the heating of the DSC experiments (between 70 and 110 °C). Using crystallization upon heating as a measure, Figure 6b summarizes the aging progress of milled 70% felodipine− EUDRGAIT E PO extrudates. The crystallization exotherm shifts to a lower temperature with a reduced enthalpy value after being aged for longer than 1 month. This reduction in the crystallization exothem indicates a certain amount of amorphous felodipine may have crystallized on aging in the milled sample. Between 2 and 6 months aging, little change is observed in the crystallization exotherm of the samples indicating the amount of phase-separated amorphous felodipine-rich domains reached a plateau. This is perhaps better presented in Figure 6c in which the total amount of drug crystallization on heating is plotted against the aging period. These results confirm that these populations of amorphous drug-rich domains are stable under ambient conditions for 6 months and only recrystallize upon heating. The long-term destabilization effect was also studied using ATR-FTIR. Figure 7a shows the ATR-FTIR spectra of the milled extrudates with 50% drug loading and aged at room temperature/75% RH for up to 6 months. There is little evidence of significant amounts of amorphous phase separation in the freshly milled extrudates. The NH peak of felodipine is at 3353 cm−1 in the spectra of the unmilled extrudates. However, after 1 month of aging of the milled extrudates with 50% drug loading, this peak shifted to 3363 cm−1 and became asymmetric in shape. The asymmetric shape of the NH peak is likely to be a result of the coexistence of an amorphous peak with low intensity and a crystalline peak. This peak shifted to 3370 cm−1 after 2 months of aging indicating the crystallization of felodipine form I at the surface of the extrudates. As seen in Figure 7b, for the milled H
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 8. (a) Total heat flow signals of the MTDSC results of fresh strand form and milled 70% (w/w) celecoxib−EUDRAGIT E PO melt extrudates on aging under 75% RH at room temperature; (b) comparison of the relaxation endotherm at Tg of the solid dispersions of unmilled, milled, and aged celecoxib−EUDRAGIT E PO extrudates; (c) calculated crystallization of milled 70% HME celecoxib−EUDRAGIT E PO extrudates aged for different time period.
E PO extrudates measured using MTDSC (Figure 8). The total amount of crystallized celecoxib on heating is calculated as 11.2% w/w in comparison to approximately 0.5% for felodipine− EUDRAGIT E PO extrudates. This may indicate that judgments
HME celecoxib−EUDRAGIT E PO extrudates (without milling) with 70% drug loading showed a much higher tendency for drug crystallization on heating (as indicated by the recrystallization exotherm) than the corresponding felodipine−EUDRAGIT I
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 9. (a) Total heat flow signals of the MTDSC results of 70% (w/w) melt extrudates before and after milling; (b) reversing heat flow signals of 70% (w/w) felodipine−PVPVA melt extrudates before and after milling; (c) comparison of the relaxation endotherm at Tg of the solid dispersions of felodipine in different polymers; solid line represents the signals of the samples before milling, short-dashed line represents the signals of the samples after 30 min milling, and long-dashed line represents the signals of the samples after 60 min milling; (d) comparison of the effect of milling time on felodipine crystallization on heating of the 70% felodipine-loaded HME extrudates with different polymer matrixes.
made solely based on the Tg of the drug and theoretical prediction of drug in polymer solubility are not reliable for processed solid dispersions. It can be seen that after milling the total crystalline drug content on heating increased to 28%. The real-time stability of the milled and unmilled celecoxib− EUDRAGIT E PO extrudates showed little change over 6 months. As seen in Figure 8b, the higher tendency of drug crystallization on heating in the milled samples may be attributed to the increased molecular mobility or phase separation of amorphous drug-rich domains in the milled extrudates. The hypothesis was further tested on felodipine−PVPVA extrudates. With a higher Tg and a stronger hydrogen bonding affinity with the drug in the extrudates than EUDRAGIT E PO,36 PVPVA was expect to have a higher miscibility and solubility with felodipine. However, the theoretical based prediction method gives approximately 40% w/w solubility, which is similar to the drug solubility in EUDRGAIT E PO. As seen in Figure 9a, milling leads to a lower level of drug recrystallization on heating (4% w/ w) in comparison to felodipine−EUDRAGIT E PO extrudates (23.5% w/w).36 With increasing the milling time to 60 min, no significant increase in the amount of crystalline drug was generated. This result again challenges the relevance of the theoretical prediction to the physical stability of solid dispersions. The reversing heat flow signals of these samples showed clear indications of phase-separated amorphous drug-rich domains. As
seen in Figure 9b, after 30 min of milling, a small melting peak of crystalline felodipine and a higher Tg were detected in comparison to the unmilled sample. With further increasing the milling time to 60 min, a double Tg can be seen in Figure 9b. This indicates the presence of drug-rich and polymer-rich domains, which can be related to the low Tg (57.8 °C) and high Tg (67.4 °C), respectively. Further examining the Tg region, it can be seen that the relaxation endotherm, which is associated with the molecular mobility of the system, was increased after milling (Figure 9c). This confirms that increased molecular mobility in the solid dispersions was induced by milling. Finally the hypothesis was tested using a a ternary solid dispersion with felodipine loaded in a polymer blend of EUDRAGIT E PO and PVPVA. Polymer blend based solid dispersions have been shown to provide enhanced in vitro stability and drug release performance in comparison to single polymer matrixes.36 The polymer blend extrudates loaded with 70% felodipine have an intermediate Tg of 47.6 °C between felodipine−EUDRAGIT E PO and felodipine−PVPVA extrudates.36 The relaxation endotherm at the Tg of the polymer blend samples also shows an increase after milling (Figure 9c). The DSC results of the freshly milled polymer blend extrudates reveal an intermediate level of stability against milling between felodipine−EUDRAGIT E PO and felodipine−PVPVA extrudates (Figure 9a,d). Similar to the results of milled felodipine− J
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Figure 10. Illustration of the destabilization process of milling on supersaturated HME solid dispersions.
used to produce the solid dispersion are not taken into consideration. The prediction of the milling-induced higher molecular mobilities for both model drugs is supported by the increased relaxation enthalpy at the T g and the observed drug crystallization on heating (Figures 8 and 9). Pure amorphous felodipine has good physical stability and does not recrystallize on heating (data not shown), but milled extrudates showed drug crystallization on heating. Pure amorphous celecoxib recrystallizes on heating in comparison to the milled extrudates that contain celecoxib. As illustrated in Figure 10, phase separation leads to the presence of a greater number of amorphous drug-rich domains than occurs with the unmilled solid dispersions. These phase-separated amorphous drug-rich domains exhibited a high crystallization tendency on heating but are kinetically stable over the tested period in the HME solid dispersion as suggested by the real-time stability data. The extent of this thermal instability of the amorphous, drug-rich domains is likely to depend on the intrinsic stability of the amorphous drug and the stabilization capacity of the polymeric carrier. Overall, the destabilization effect showed a strong dependence on the degree of saturation of drug in the polymer. The milling time and physical characteristics of the drug and the carrier polymer can also affect the destabilization. For example, amorphous celecoxib has lower thermal stability than amorphous felodipine. Despite the higher predicted solubility of celecoxib in EUDRAGIT E PO, supersaturated felodipine−EUDRAGIT E PO extrudates showed better stability against milling. 4.2. New Insights into the Drug−Polymer Solubility of Processed Solid Dispersions. To date, the existing predictive methods to determine the solubility of drug in polymer for the purpose of predicting the physical stability of a solid dispersion formulation have been mainly based on theoretical approaches.1,6 These approaches largely rely on the response of heated physical mixtures of crystalline drug and the polymer, and the major drawback of their use is that they do not take into
EUDRAGIT E PO extrudates, the presence of a crystallization exotherm indicates that milling induced the formation of amorphous drug-rich domains, which rapidly recrystallized on heating. The enthalpy of the drug crystallization exotherm on heating is lower than that of the milled felodipine−EUDRAGIT E PO extrudates but higher than that of the milled felodipine− PVPVA extrudates (Figure 9a). The DSC detected drug crystallization on heating is approximately 9.6% w/w, which does not change significantly with increasing the milling time.
4. DISCUSSION 4.1. Destabilization by Milling on Processed Solid Dispersions. The results of this study have demonstrated that the destabilization behavior of milling on the processed solid dispersions is highly drug loading dependent. For solid dispersions with drug loadings below the predicted saturated solubility limit, mechanical milling appears to have no impact on the physical stability of the dispersion. This can be explained by the highly stable molecular dispersion formed between the drug and polymer. For supersaturated molecular dispersions, as little as 5 min of milling can destabilize the system. The destabilization is manifested through the increased molecular mobility of the amorphous drug and the rapid phase separation of the excess drug that exists above the solid solubility of the drug in the polymer as drug-rich amorphous domains. However, these amorphous domains are stable at ambient conditions as demonstrated by the real-time physical stability studies. This indicates that HME solid dispersions can stabilize a much higher amount of amorphous drug than would be predicted using solubility values calculated using the theoretical approaches. This is likely to be because the three predictive methods rely either solely on chemical structures of the drug and polymer or experimental examinations of the nonprocessed physical mixes of the model systems. Therefore, the intrinsic physical stability of the amorphous drug and any potential impact of the processing K
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
Notes
consideration any potential influence of manufacturing process on the apparent solid solubility of drug in polymer. In this study, solid dispersions prepared by HME were used as example systems, and the real-time physical stability results of milled extrudates suggested that the apparent solubility of a drug in a polymer processed by HME can be much higher than that provided by theoretical prediction (melting point depression and melting enthalpy methods) calculated from data of the physical mixes of the drug and polymer. This is likely to be a result of not taking into account the additional solubilization impacts of the high energy input and intimate mixing provided during the HME process.2,37 Although this process-related apparent solubility may be higher than the thermodynamic equilibrium solubility, the stability data has confirmed that it is likely to be a more relevant solubility value for the typical pharmaceutical shelf life. Furthermore, there are increasing amounts of published data on the variation in the stability of the same drug−polymer combination solid dispersions prepared by different methods.10,11,13,14 Therefore, it seems reasonable to suggest that the apparent solubility of a model drug in a polymer can vary if the same solid dispersions are prepared by different methods such as organic solvent film casting or spray drying.
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS Z.Y. would like to thank Evonik for the financial support for the period of his Ph.D. The authors also would like to acknowledge the contribution of members of the Interreg IV-A project funded by the European Union. All authors would like to thank Prof. Peter Belton’s constructive discussions during the preparation of the manuscript.
■
(1) Marsac, P.; Shamblin, S.; Taylor, L. Theoretical and Practical Approaches for Prediction of Drug−Polymer Miscibility and Solubility. Pharm. Res. 2006, 23 (10), 2417−2426. (2) Qi, S.; Belton, P.; Nollenberger, K.; Clayden, N.; Reading, M.; Craig, D. M. Characterisation and Prediction of Phase Separation in Hot-Melt Extruded Solid Dispersions: A Thermal, Microscopic and NMR Relaxometry Study. Pharm. Res. 2010, 27 (9), 1869−1883. (3) Mahieu, A.; Willart, J.-F.; Dudognon, E.; Danède, F.; Descamps, M. A New Protocol To Determine the Solubility of Drugs into Polymer Matrixes. Mol. Pharmaceutics 2012, 10 (2), 560−566. (4) Yang, J.; Grey, K.; Doney, J. An Improved Kinetics Approach to Describe the Physical Stability of Amorphous Solid Dispersions. Int. J. Pharm. 2010, 384 (1−2), 24−31. (5) Djuris, J.; Nikolakakis, I.; Ibric, S.; Djuric, Z.; Kachrimanis, K. Preparation of Carbamazepine−Soluplus® Solid Dispersions by HotMelt Extrusion, and Prediction of Drug−Polymer Miscibility by Thermodynamic Model Fitting. Eur. J. Pharm. Biopharm. 2013, 84 (1), 228−237. (6) Marsac, P.; Li, T.; Taylor, L. Estimation of Drug−Polymer Miscibility and Solubility in Amorphous Solid Dispersions Using Experimentally Determined Interaction Parameters. Pharm. Res. 2009, 26 (1), 139−151. (7) Flory, P. J. Priciples of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953. (8) Nishi, T.; Wang, T. T. Melting Point Depression and Kinetic Effects of Cooling on Crystallization in Poly(vinylidene fluoride)− Poly(methyl methacrylate) Mixtures. Macromolecules 1975, 8 (6), 909− 915. (9) Greenhalgh, D. J.; Williams, A. C.; Timmins, P.; York, P. Solubility Parameters as Predictors of Miscibility in Solid Dispersions. J. Pharm. Sci. 1999, 88 (11), 1182−1190. (10) Janssens, S.; Zeure, A.; Paudel, A.; Humbeeck, J.; Rombaut, P.; Mooter, G. Influence of Preparation Methods on Solid State Supersaturation of Amorphous Solid Dispersions: A Case Study with Itraconazole and Eudragit E100. Pharm. Res. 2010, 27 (5), 775−785. (11) Weuts, I.; Van Dycke, F.; Voorspoels, J.; De Cort, S.; Stokbroekx, S.; Leemans, R.; Brewster, M. E.; Xu, D.; Segmuller, B.; Turner, Y. T. A.; Roberts, C. J.; Davies, M. C.; Qi, S.; Craig, D. Q. M.; Reading, M. Physicochemical Properties of the Amorphous Drug, Cast Films, and Spray Dried Powders to Predict Formulation Probability of Success for Solid Dispersions: Etravirine. J. Pharm. Sci. 2011, 100 (1), 260−274. (12) Brettmann, B. K.; Myerson, A. S.; Trout, B. L. Solid-State Nuclear Magnetic Resonance Study of the Physical Stability of Electrospun Drug and Polymer Solid Solutions. J. Pharm. Sci. 2012, 101 (6), 2185−2193. (13) Caron, V.; Tajber, L.; Corrigan, O. I.; Healy, A. M. A Comparison of Spray Drying and Milling in the Production of Amorphous Dispersions of Sulfathiazole/Polyvinylpyrrolidone and Sulfadimidine/ Polyvinylpyrrolidone. Mol. Pharmaceutics 2011, 8 (2), 532−542. (14) Karmwar, P.; Graeser, K.; Gordon, K. C.; Strachan, C. J.; Rades, T. Investigation of Properties and Recrystallisation Behaviour of Amorphous Indomethacin Samples Prepared by Different Methods. Int. J. Pharm. 2011, 417 (1−2), 94−100. (15) Surana, R.; Pyne, A.; Suryanarayanan, R. Effect of Preparation Method on Physical Properties of Amorphous Trehalose. Pharm. Res. 2004, 21 (7), 1167−1176.
5. CONCLUSIONS This study has reported the destabilization effect of high-energy mechanical milling on processed solid dispersions and its correlation with drug−polymer solubility. The destabilization effect was only observed in the samples with drug loadings much higher than the predicted drug solubility in the polymer using theoretical approaches. For these supersaturated solid dispersions, milling led to increased molecular mobility and subsequently the formation of an increased amount of amorphous drug-rich domains in comparison to unmilled samples. Interestingly these milling-induced amorphous drugrich domains can be stable in the solid dispersions and only recrystallize on heating. Although milling time and drug− polymer physical properties also have effects on the destabilization, it is clear that drug−polymer solubility is the most important parameter. Although it might be useful to keep drug loading below the predicted solubility value in the interests of producing a stable system, the results of this study have demonstrated that in many cases the theoretical prediction is underestimating the apparent solubility by not taking into account the solubilization impact of processing. Processed solid dispersions can have a significantly higher apparent solubility than the predicted value. The real-time physical stability results of this study challenged the kinetic relevance of the solubility value measured by theoretical methods to the physical stability of processed solid dispersions. Therefore, the development of a valid method to predict the apparent solubility of drug in polymer in processed solid dispersions is crucial for more accurate prediction of the physical stability of solid dispersions.
■
ASSOCIATED CONTENT
S Supporting Information *
Drug−polymer solubility predictions using theoretical approaches. This material is available free of charge via the Internet at http://pubs.acs.org.
■
REFERENCES
AUTHOR INFORMATION
Corresponding Author
*(S.Q.) E-mail:
[email protected]. L
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
Molecular Pharmaceutics
Article
(16) Otsuka, M.; Matsumoto, T.; Kaneniwa, N. Effect of Environmental Temperature on Polymorphic Solid-State Transformation of Indomethacin during Grinding. Chem. Pharm. Bull. 1986, 34 (4), 1784− 1793. (17) Willart, J. F.; De Gusseme, A.; Hemon, S.; Odou, G.; Danede, F.; Descamps, M. Direct Crystal to Glass Transformation of Trehalose Induced by Ball Milling. Solid State Commun. 2001, 119 (8−9), 501− 505. (18) Willart, J. F.; Caron, V.; Lefort, R.; Danède, F.; Prévost, D.; Descamps, M. Athermal Character of the Solid State Amorphization of Lactose Induced by Ball Milling. Solid State Commun. 2004, 132 (10), 693−696. (19) Willart, J.-F.; Lefebvre, J.; Danède, F.; Comini, S.; Looten, P.; Descamps, M. Polymorphic Transformation of the Γ-Form of D-Sorbitol upon Milling: Structural and Nanostructural Analyses. Solid State Commun. 2005, 135 (8), 519−524. (20) Chikhalia, V.; Forbes, R. T.; Storey, R. A.; Ticehurst, M. The Effect of Crystal Morphology and Mill Type on Milling Induced Crystal Disorder. Eur. J. Pharm. Sci. 2006, 27 (1), 19−26. (21) Desprez, S.; Descamps, M. Transformations of Glassy Indomethacin Induced by Ball-Milling. J. Non-Cryst. Solids 2006, 352 (42−49), 4480−4485. (22) Descamps, M.; Willart, J. F.; Dudognon, E.; Caron, V. Transformation of Pharmaceutical Compounds upon Milling and Comilling: The Role of Tg. J. Pharm. Sci. 2007, 96 (5), 1398−1407. (23) Willart, J.-F.; Carpentier, L.; Danède, F.; Descamps, M. SolidState Vitrification of Crystalline Griseofulvin by Mechanical Milling. J. Pharm. Sci. 2012, 101 (4), 1570−1577. (24) Fecht, H. J. Defect-Induced Melting and Solid-State Amorphization. Nature 1992, 356 (6365), 3. (25) Willart, J. F.; Descamps, M. Solid State Amorphization of Pharmaceuticals. Mol. Pharmaceutics 2008, 5 (6), 905−920. (26) Crowley, K. J.; Zografi, G. Cryogenic Grinding of Indomethacin Polymorphs and Solvates: Assessment of Amorphous Phase Formation and Amorphous Phase Physical Stability. J. Pharm. Sci. 2002, 91 (2), 492−507. (27) Chieng, N.; Rades, T.; Saville, D. Formation and Physical Stability of the Amorphous Phase of Ranitidine Hydrochloride Polymorphs Prepared by Cryo-Milling. Eur. J. Pharm. Biopharm. 2008, 68 (3), 771− 780. (28) Feng, T.; Pinal, R.; Carvajal, M. T. Process Induced Disorder in Crystalline Materials: Differentiating Defective Crystals from the Amorphous Form of Griseofulvin. J. Pharm. Sci. 2008, 97 (8), 3207− 3221. (29) Feng, T.; Bates, S.; Carvajal, M. T. Toward understanding the Evolution of Griseofulvin Crystal Structure to a Mesophase after Cryogenic Milling. Int. J. Pharm. 2009, 367 (1−2), 16−19. (30) Qi, S.; Weuts, I.; De Cort, S.; Stokbroekx, S.; Leemans, R.; Reading, M.; Belton, P.; Craig, D. Q. M. An investigation into the crystallisation behaviour of an amorphous cryomilled pharmaceutical material above and below the glass transition temperature. J. Pharm. Sci. 2010, 99 (1), 196−208. (31) Miller, D. A.; McConville, J. T.; Yang, W.; Williams, R. O.; McGinity, J. W. Hot-Melt Extrusion for Enhanced Delivery of Drug Particles. J. Pharm. Sci. 2007, 96 (2), 361−376. (32) Lakshman, J. P.; Cao, Y.; Kowalski, J.; Serajuddin, A. T. M. Application of Melt Extrusion in the Development of a Physically and Chemically Stable High-Energy Amorphous Solid Dispersion of a Poorly Water-Soluble Drug. Mol. Pharmaceutics 2008, 5 (6), 994−1002. (33) Albers, J.; Alles, R.; Matthée, K.; Knop, K.; Nahrup, J. S.; Kleinebudde, P. Mechanism of drug release from polymethacrylatebased extrudates and milled strands prepared by hot-melt extrusion. Eur. J. Pharm. Biopharm. 2009, 71 (2), 387−394. (34) Yang Z. Y.; et al. A Theoretical and Experimental Analysis of Solid Solubility of Felodipine in Eudragit E and Its Implications for the Prediction of the Physical Stability of Hot Melt Extruded Solid Dispersions. AAPS Conf. October 23−27, 2011; Washington, D.C. Poster W4308.
(35) Rollinger, J. M.; Burger, A. Polymorphism of Racemic Felodipine and the Unusual Series of Solid Solutions in the Binary System of Its Enantiomers. J. Pharm. Sci. 2001, 90 (7), 949−959. (36) Yang, Z.; Nollenberger, K.; Albers, J.; Craig, D.; Qi, S. Microstructure of an Immiscible Polymer Blend and Its Stabilization Effect on Amorphous Solid Dispersions. Mol. Pharmaceutics 2013, 10 (7), 2767−2780. (37) Qi, S.; Belton, P.; Nollenberger, K.; Gryczke, A.; Craig, D. M. Compositional Analysis of Low Quantities of Phase Separation in HotMelt-Extruded Solid Dispersions: A Combined Atomic Force Microscopy, Photothermal Fourier-Transform Infrared Microspectroscopy, and Localised Thermal Analysis Approach. Pharm. Res. 2011, 28 (9), 2311−2326.
M
dx.doi.org/10.1021/mp500205c | Mol. Pharmaceutics XXXX, XXX, XXX−XXX
