Indomethacin: New Polymorphs of an Old Drug - Molecular

Sep 11, 2013 - Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University (Parkville Campus), Parkville, ...
2 downloads 5 Views 1MB Size
Subscriber access provided by University of South Dakota

Article

Indomethacin: New polymorphs of an old drug Sachin A. Surwase, Johan P Boetker, Dorothy Saville, Ben J. Boyd, Keith C Gordon, Leena Peltonen, and Clare J. Strachan Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/mp400299a • Publication Date (Web): 11 Sep 2013 Downloaded from http://pubs.acs.org on October 1, 2013

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Molecular Pharmaceutics is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Indomethacin: New polymorphs of an old drug Sachin A. Surwase1, Johan P. Boetker2, Dorothy Saville1, Ben J. Boyd3, Keith C. Gordon4, Leena 5

5

Peltonen , Clare J. Strachan * 1

School of Pharmacy, University of Otago, Dunedin, New Zealand

2

Department of Pharmacy, Faculty of Health and Medical Sciences, University of Copenhagen, Denmark

3

Drug Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash University (Parkville Campus), Parkville, Victoria 3052, Australia

4

Department of Chemistry and MacDiarmid Institute for Advanced Materials and Nanotechnology, University of Otago, Dunedin, New Zealand 5

*

Division of Pharmaceutical Technology, Faculty of Pharmacy, University of Helsinki, Finland

Corresponding author: Division of Pharmaceutical Technology, Faculty of Pharmacy, P.O. Box 56

(Viikinkaari 5E), FI-00014 University of Helsinki, Finland. Tel: +358 9 191 59736. Fax: +358 9 191 59144. E-mail: [email protected].

1 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

Graphical Abstract

2 ACS Paragon Plus Environment

Page 3 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

ABSTRACT This study reports the appearance and characterization of multiple new polymorphic forms of indomethacin. Considering the interest in amorphous suspensions for toxicology studies of poorly water soluble drugs, we sought to investigate the crystallization behaviour of amorphous indomethacin in aqueous suspension. Specifically, the effect of pH and temperature on crystallization behaviour was studied. Quench cooled amorphous powder was added to buffered media at different pH values (1.2, 4.5, and 6.8) at 5 and 25°C. Both the solid and solution were analyzed at different time points up to 24 h. ATR-FTIRspectroscopy (with principal component analysis) was used to study solid-phase transformations and UV-spectroscopy used to probe solution concentration. The crystallization onset time decreased and rate of crystallization increased with increasing pH and temperature. Diverse polymorphic forms were observed, with three new forms being identified; these were named ε, ζ and η. At 25°C, the amorphous form recrystallized directly to the α form, except at pH 6.8, where it initially converted briefly into the ε form. At 5°C, all three new polymorphic forms were observed sequentially in the order ε, ζ and then η, with the number of these forms observed increasing sequentially with decreasing pH. The three new forms exhibited distinct XPRD, DSC, and FTIR and Raman spectroscopy profiles. The solution concentration profiles suggest that the relative physical stabilities of the polymorphs at 5°C from lowest to highest is ε < ζ < η < α. The appearance of new polymorphs in this study, suggests that amorphous suspensions are worth considering when performing polymorphic screening studies.

Keywords: Amorphous, Indomethacin, New Polymorphs, Aqueous Suspensions, Toxicolgy Study, Solvent mediated crystallization

3 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1.

Page 4 of 22

INTRODUCTION With the advent of target-based drug discovery and high-throughput technology, early

screening of molecules has become more focused towards synthesizing compounds against specific targets.1, 2 The size and complexity of the molecules identified has increased, which, in turn, has led to more poorly water soluble drugs entering the drug development pipeline. At present, it is estimated that more than 70% of the new chemical entities (NCEs) synthesized have issues with their aqueous solubility and hence oral bioavailability.3-5 For these drugs, solid state modification may be considered to increase the apparent solubility. The amorphous form, which exhibits a higher apparent solubility and associated dissolution rate compared to any crystalline material, can be considered the optimal solid state form in this respect.6-9 However, being thermodynamically unstable, it has a tendency to crystallise during storage. This, in turn, leads to a loss of the desired solubility advantage. Therefore, a detailed understanding of the physical stability and crystallization behaviour of drugs in the amorphous form is needed to develop strategies to prevent crystallization during storage.10 For preclinical toxicology studies, suspensions are the recommended formulations for oral drug delivery. The amorphous form can theoretically be formulated as a suspension, so long as the solid phase remains amorphous until (and during) administration. In suspension, crystallization of the amorphous form can take place not only by direct conversion of the solid, but also via solution.11, 12 This makes earlier crystallization more likely, with the nature of the solvent also having an effect on the crystallization behaviour. Furthermore, the direct solid conversion may also be accelerated compared to that in storage in dry conditions, due to an increase in molecular mobility with solvent absorption.13

While stability of the

amorphous form in solid dosage forms during storage has been explored in depth, less attention has been given to its stability in solvent environments.14 Recently, the physical stability of amorphous drugs during dissolution testing has been investigated12, 14, 15, but, to the best of our knowledge, the physical stability of amorphous suspension formulations has not yet been investigated. The objective of this study was to investigate the physical stability of amorphous form of the model drug, indomethacin, in aqueous suspension. The effects of two variables, pH and temperature, were investigated. Indomethacin was selected as the model drug since it is ionizable and the crystallization behaviour of its amorphous form during storage (in a gaseous environment) and dissolution testing is comparatively well studied.16-18 Since amorphous suspensions are not typically used during polymorphic screening studies, we were also interested in the possible generation of new polymorphs in the suspensions as a function of pH and temperature. 4 ACS Paragon Plus Environment

Page 5 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Indomethacin is polymorphic; five different polymorphic modifications have been reported, namely α, β, γ and δ, as well as an unnamed crystal form.19,

20

The γ form is the

thermodynamically stable form and the α form is the most commonly observed metastable form.16, 21 Amorphous indomethacin commonly crystallizes to either the metastable α and stable γ forms, depending on amorphous form preparation and storage conditions. For example, at temperatures below the glass transition (Tg) and at low humidity, crystallization to the γ form is observed. At high humidities and temperatures above the Tg and in aqueous environments, the metastable α form appears.12

, 14-16

The δ form has been reported to

crystallise for example from indomethacin-poly(vinyl pyrrolidone) solid dispersions at 94% relative humidity and from a methanolic solution of indomethacin on drying.22, 23 The β form and the unnamed form have only been reported in single publications, and were crystallized from dioxane and cyclodextrin solutions, respectively.19, 20

2. MATERIALS AND METHODS 2.1 Materials Indomethacin (Figure 1) (>98% purity, γ form) was purchased from Chemie Brunschwig AG (Basel, Switzerland). Ethanol (analytical grade) and phosphorus pentoxide were obtained from Merck (Darmstadt, Germany). O

H3C O

OH CH3

N O Cl

Figure 1. Structure of indomethacin

2.2 Preparation of α, γ and δ form The γ form of indomethacin was used as received. The α form of indomethacin was prepared by addition of Milli-Q water (antisolvent) to a saturated solution of the indomethacin obtained in ethanol at 80°C. The precipitated crystals were removed by filtration and then dried under vacuum at room temperature.21 For δ form generation, first, indomethacin methanolate was prepared by air drying the methanolic solution of indomethacin and then desolvation was carried out under vacuum at 30°C for 10 days.22 All the solid-state forms obtained were then characterized by differential scanning calorimetry (DSC), attenuated total

5 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 22

reflection Fourier transform infrared (ATR-FTIR) spectroscopy, Raman spectroscopy and Xray powder diffraction (XRPD).

2.3 Preparation of amorphous form Amorphous indomethacin was prepared by heating the γ form to 165°C and quench cooling with liquid nitrogen. Amorphous material was then milled using an oscillatory ball mill (Mixer Mill MM301, Retsch GmbH & Co., Haan, Germany) at 30 Hz for 30 seconds for use in the study.

2.4 Equilibrium solubility The equilibrium solubility of indomethacin (α form) was studied at 5 and 25°C by the conventional shake flask method. Excess drug was added to 10 ml of standard 0.1 M aqueous media, HCl (pH 1.2), and phosphate buffer (pH 4.5 and 6.8)] and the resulting suspension was shaken at 250 rpm (Ratek, Orbital Mixer Incubator, Biolab Scientific Ltd., New Zealand) for 48 h. The concentration of dissolved drug in the different media was determined by UV absorbance (CARY Varian, Clayton South, Victoria, Australia) at a wavelength of 320 nm. Tests were carried out in triplicate. The excess solid remaining after 48 h was analyzed by ATR-FTIR spectroscopy to determine if any solid-state transformation had occurred.

2.5 Preparation and characterization of suspensions To prepare the suspensions, amorphous indomethacin was added (20 mg/ml) to aqueous media in 20 ml scintillation vials. The vials were capped and placed in double jacketed thermostatted beakers mounted on a magnetic stirrer plate. The suspensions were then stirred at 250 rpm with the help of small magnetic beads during the time course of the experiments. The crystallization behaviour and dissolved concentration were studied at different pH values (1.2, 4.5, and 6.8) and temperatures (5 and 25°C). Both the excess solid and concentration in the solution were analyzed at different time points (5, 15, 30, 60, 120, 240, 360, 480 and 1440 min (24 h)). At each time point 400 µl of suspension was pipetted from the center of the vials. Samples were immediately centrifuged for 30 sec at 14000 rpm. The clear supernatant obtained was diluted with pH 6.8 buffer, vortexed and analyzed by using UV at wavelength of 320 nm. The remaining solid was analyzed by ATR-FTIR spectroscopy to study the phase transformations. Suspensions were both prepared and analyzed in triplicate.

2.6 HPLC Wet solids obtained (section 2.5) were air dried at 5°C and then were analyzed using a HPLC system (Shimadzu SP-6A) at a wavelength of 320 nm. Chromatography was 6 ACS Paragon Plus Environment

Page 7 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

performed at ambient temperature using a Luna 5 µ C18 (2) column (250mm length, 4.6 mm diameter, 5 μ particle size; Phenomenex, Terrence, CA, USA). The mobile phase consisting of 75% methanol and 25% of 0.2% phosphoric acid was used at a flow rate of 1.0 ml/min with the injection volume of 20 μl. For analysis, dried samples along with γ and amorphous forms of indomethacin were prepared separately at 50 µg/ml concentration in mobile phase.

2.7 IR and Raman spectroscopy The wet solid obtained after centrifugation (section 2.5) was used for analysis. Samples were pressed firmly between the ATR crystal and probe using the clamp in order to remove excess water. FTIR spectroscopy was performed on a Bruker Vertex 70 Fourier transform infrared spectrometer (Bruker Optik, Ettlingen, Germany) using an ATR accessory with a single reflection diamond crystal (MIRacle, Pike Technologies, Madison, WI, USA). The spectrometer was equipped with a KBr beam splitter, an MIR source and a RT-DLaTGS detector. Samples were measured over a wavenumber range from 650 to 4000 cm−1 and the final spectrum was the mean of 64 scans. The interferograms were apodized with the Blackman-Harris 3-term function and subjected to Fourier transformation yielding spectra with a resolution of 4 cm−1. The ATR spectra were converted to absorbance spectra using OPUS software (v. 5.0, Bruker Optik, Ettlingen, Germany). The FT-Raman instrument consisted of a Bruker FRA 106/S FT-Raman accessory (Bruker Optik, Ettlingen, Germany) with a Coherent Compass 1064-500N laser (Coherent Inc., Santa Clara, USA) attached to a Bruker Equinox 55 FT interferometer, and a D 418T liquid nitrogen cooled Ge diode detector. The solid obtained after centrifugation was packed into an aluminum cup and the analysis was carried out at room temperature using a laser wavelength of 1064 nm (Nd: YAG laser). Spectra were the average of 128 scans, taken at 4 cm−1 resolution with a laser power of 120 mW.

2.8 XRPD XRPD patterns were recorded using a PANalytical X’Pert PROMPD system (PW3040/60, Philips, The Netherlands) using Cu Kα radiation with λ = 1.542 Å and a divergence slit of 1°. Dried samples (section 2.6) were loaded on to an aluminum sample holder. Samples were then scanned at 40 kV and 30 mA from 5° to 35°2θ using a scanning speed of 0.1285°min-1 and a step size of 0.0084°. The diffraction patterns were generated using X’Pert High Score version 2.2.0 (Philips, The Netherlands). Also for some samples XRPD analysis was performed over 2.5 hours using an X’Pert PRO θ/θ X-ray diffractometer (PANalytical B.V., Almelo, The Netherlands) and Ni filtered Cu Kα1 radiation (λ=1.5406Å). Samples were measured in Bragg Brentano reflection mode in the 2θ–range 2 to 35° using a PIXel detector (PANalytical B.V., Almelo, The Netherlands); 7 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 22

step size 0.0030° 2θ and a scan step time of 180 s. The operating current and voltage were 40 mA and 45 kV, respectively. Data were collected using X’Pert Data Collector (PANalytical B.V., Almelo, The Netherlands). XRPD diffractograms were conducted at 5°C using an Anton Paar CHC chamber (Anton Paar GmbH, Graz, Austria) mounted on the goniometer. The temperature was controlled using a TCU 110 temperature controller (Anton Paar GmbH, Graz, Austria).

2.9 DSC A TA instrument (V8.2 Build 268, New Castle, USA) DSC Q100 was used to record the thermograms after temperature and enthalpy calibration using indium. Dried samples (see section 2.6) (1-3 mg) were crimped in an aluminum pan and heated at a rate of 10°Cmin-1 from 0 to 180ºC, under a nitrogen gas flow of 50 ml min-1. Thermal events observed were determined using TA Universal Analysis software (version 4.0C).

2.10

Multivariate data analysis

Principal component analysis (PCA) was performed on the ATR-FTIR spectra, for better understanding of the spectral and hence the structural variation between the samples at different time points. The region between 1000 and 1400 cm-1 was chosen to reduce the interference of water in the infrared spectrum. In order to remove intensity differences unrelated to the sample composition, standard normal variant (SNV) transformation was performed on the spectra before PCA. PCA was performed on the selected spectral ranges after spectral preprocessing and mean centering by using Unscrambler software version 10.01 (CAMO Software AS, Oslo, Norway).

3. RESULTS AND DISCUSSION 3.1 Physicochemical changes in suspensions during storage To study the effect of pH and temperature on the crystallization of amorphous indomethacin in an aqueous suspension, experiments were performed at pH values of 1.2, 4.5 and 6.8, each at both 5 and 25°C. ATR-FTIR spectroscopy along with PCA was utilized to identify onset of crystallization and subsequent phase transformations. Further, concentration-time profiles were generated to understand the effect of crystallization on the apparent solubility of the drug. The PCA score plots of the ATR-FTIR spectra of the reference and stored forms at pH 1.2 (at both 5 and 25°C) are shown in Figure 2a. The α, γ and freshly prepared amorphous forms of indomethacin were used as references, and are resolved mainly by the second principal component (PC), with the α form positively correlated with PC2, the γ form 8 ACS Paragon Plus Environment

Page 9 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

negatively correlated, and the amorphous form somewhat intermediate to the two crystalline forms. At 25°C, the scores of the solid component in the suspension started moving towards the α form over time, indicating that the amorphous form was crystallizing into this form. After 5 min, the scores had progressed almost halfway in a direction directly towards the α form. After 24 h (1440 min) the scores were close to the α form, indicating it had predominantly crystallized to this form. A decrease in temperature to 5°C resulted in an increase in the onset of crystallization time. At this temperature, initial samples clustered near the unprocessed amorphous form in the scores plot, with the 5 and 15 min samples superimposed on each other. After 30 min, the PC2 score values started to decrease, but in a different direction to that of the γ form, and reached a minimum at 60 min. The scores then changed direction again, moving towards the origin at 120 min and then into negative PC1, but neutral PC2, space, where the spectra clustered from 360 min until the end of the experiment. These samples appear to form three separate clusters, as circled in Figure 2a. The loadings plots of the spectra were used to probe the spectral differences leading to the clustering of the samples observed in the scores plot (see Supporting Information). Large differences were observed in the region from 1600 cm-1 to 1750 cm-1 (PC1 and PC2). The bands in this region are due to carbonyl (benzoyl and carboxylic acid) stretching vibrations, which are associated with hydrogen bonding variation between different solid state forms.24, 25

Representative spectra of the three clusters observed at 60, 120 and 360 min (together with the α, γ and amorphous form) are shown in Figure 2b. The spectra at these time points are distinct from the three reference forms, especially between 1600 and 1750 cm-1, with unique peak positions for all three spectra. The spectrum at 60 min resembles that of an unnamed form reported once previously.20 To the best of our knowledge, the other two spectra do not resemble any reported forms. We have designated the forms that appeared at 60, 120 and 480 min as the ε, ζ, and η forms, respectively and these are further characterized later in this study. At 25°C and pH 1.2, the solubility of the α form was previously determined to be 2.4 µg/ml (Table 1). In the amorphous suspension, supersaturation with respect to the α form (8.3 fold) was generated with the dissolved drug concentration peaking within 5 min, followed by a rapid drop to a concentration equal to the equilibrium solubility of the α form within the first 15 min of the experiment (Figure 2c).

9 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

Figure 2. Crystallization behaviour and solution concentration of amorphous indomethacin suspension in pH 1.2 at 5 and 25°C; (a) PCA scores plot of the IR spectra of samples at different time points (5, 15, 30, 60, 120, 240, 360, 480, and 1440 min), (b) IR spectra of reference forms and spectra from clusters at 5°C in scores plot, (c) solution concentration-time profile. Each bar represents mean ± SD of 3 independent experiments. Amorph = amorphous.

10 ACS Paragon Plus Environment

Page 11 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Table 1. Equilibrium solubilities of the α form at different pH values and temperatures Solubility (µg/ml) pH 5°C 25°C

1.2

1.9 ± 0.3

2.4 ± 0.2

4.5 6.8

9.3 ± 1.6 680.5 ± 9.3

11.6 ± 0.4 777.3 ± 11.1

At 5°C, supersaturation with respect to the α form was slightly higher at 5 min (10.6 fold) and the concentration drop at this temperature was slower than at 25°C. Interestingly, there appears to be a plateau in the concentration drop between 60 and 120 min. As can be seen from the scores plot (Figure 2a), this is the time where the ε form converted into the ζ form. Subsequently the concentration dropped further, coinciding with the appearance of the η form at the expense of the ζ form. After 24 hours, the solution concentration remained slightly supersaturated with respect to the α form, which suggests that the η form is more soluble than the α form, and therefore less stable. It is evident from the initial drop in dissolved drug concentration after 5 min, that crystallization at both temperatures is completely or largely solvent mediated. Direct solidsolid transformation of the amorphous material in the aqueous suspensions may also be occurring to some extent. It has been reported that every 1% increase in water content lowers the Tg of amorphous indomethacin by 10°C.16 An aqueous environment will therefore increase the molecular mobility and promote crystallization also in the solid phase. The contribution of the solid-solid transformation to the overall crystallization would be more likely to be significant at 25°C compared to 5°C, and this may contribute to the higher rate of crystallization observed at 25°C. The step-like decrease in the dissolved drug concentration associated with the multiple polymorphic transformations at 5°C is consistent with progressively more thermodynamically stable polymorphs being formed. The transformation to progressively less soluble and more stable forms is an example following Ostwald’s rule of stages.26 At pH 4.5, the reference forms (α, γ and amorphous) could also be resolved in the PCA scores plot, and changes in the suspensions at 5 and 25°C can be followed (Figure 3a). At 25°C, the score values proceeded directly from the amorphous form (close to the origin) towards the α form (positive PC1 values), revealing similar behaviour to that at pH 1.2. However, while the degree of crystallization appeared similar for the first 5 min at both pH values, after this the crystallization appears more rapid at pH 4.5. This faster rate of crystallization is likely to be due to the higher solubility of indomethacin at the higher pH (Table 1). At 5°C, the scores moved to more negative PC1 and PC2 values after 60 min indicating crystallization, and remained in this region until 240 min. At 360 min and

11 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

afterwards, the scores clustered in the upper left quadrant and remained in this region for the remainder of the experiment.

Figure 3. Crystallization behaviour and solution concentration of amorphous indomethacin suspension in pH 4.5 at 5 and 25°C; (a) PCA scores plot of the IR spectra of samples at different time points (5, 15, 30, 60, 120, 240, 360, 480, and 1440 min), (b) IR spectra of reference forms and spectra from clusters at 5°C in scores plot, (c) solution concentration-time profile. Each bar represents mean ± SD of 3 independent experiments. Amorph = amorphous.

12 ACS Paragon Plus Environment

Page 13 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

As was observed at pH 1.2, the loadings revealed the main spectral variation between 1600 and 1750 cm-1 (see Supporting Information). The spectra at 60 min and 360 min from the 5°C samples (Figure 3b), matched those of the ε and ζ forms observed at pH 1.2 respectively. These forms were observed in the same order at both pH values, but unlike at pH 1.2, at pH 4.5 the ζ form did not convert to any other form, even at 24 h. The ε form was also present for longer at the higher pH. At pH 4.5, a drop in dissolved drug concentration associated with crystallization was also observed at both temperatures (Figure 3c). At 25°C, similar to the concentration time profile obtained at pH 1.2, supersaturation with respect to the α form (in this case 4.5 fold) was generated within 5 min and then the concentration dropped rapidly after this, and reached the equilibrium solubility of the α form within 30 min. At 5°C, a higher degree of supersaturation (5.5 fold) was obtained, followed by a slower drop in concentration associated with crystallization from solution (Figure 3c). Similar to at pH 1.2 and 4.5, changes at pH 6.8 in the suspensions at both temperatures were observed in the scores plot (Figure 4a). At 25°C, the score values at 5 min had already proceeded negatively towards the origin, away from the α form, due to the appearance of the ε form at this time point. After 5 min, the scores rapidly changed in the direction of the α form and clustered close to it after 60 min, indicating the complete crystallization. Again, the higher solubility of indomethacin at this pH than the other pH values has likely resulted in the faster rate of crystallization. At 5°C, the scores values at 60 min were similar to those at 5 min for 25°C, indicating the crystallization to the ε form had begun. After 60 min, the scores had progressively more negative PC1 values, away from the reference forms, until 1440 min (24 h) when it formed a cluster in the lower left quadrant near the PC1 axis. This was due to gelling of the material. At 24 h the score values moved towards the α form. At this pH too, the loadings revealed the main spectral variation was between 1600 and 1750 cm-1 (see Supporting Information). The spectra at 5 min and 60 min from the 25 and 5°C, samples, respectively (Figure 4b), matched those of the ε form observed at other pH values. A drop in dissolved drug concentration at this pH also was associated with crystallization (Figure 4c). At 25°C, like other pH values, supersaturation (4.3 fold) with respect to the α form was generated within 5 min, followed by a rapid concentration drop to the equilibrium solubility value of the α form (Table 1) at this pH after 30 min. At 5°C, supersaturation (6.0 fold) with respect to the α form was generated initially and the concentration dropped more slowly than at 25°C. Interestingly, the system gelled after 60 min. This prevented further concentration measurements (Figure 4c). 13 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 22

Figure 4. Crystallization behaviour and solution concentration of amorphous indomethacin suspension in pH 6.8 at 5 and 25°C; (a) PCA scores plot of the IR spectra of samples at different time points (5, 15, 30, 60, 120, 240, 360, 480, 1440 and 1440 min), (b) IR spectra of reference forms and spectra from clusters at 5°C in scores plot, (c) solution concentration-time profile. At 5°C, the samples gelled after 60 min, preventing concentration determination. Each bar represents mean ± SD of 3 independent experiments. Amorph = amorphous.

14 ACS Paragon Plus Environment

Page 15 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

In summary, three new polymorphs of the indomethacin appeared during crystallization of the amorphous form in aqueous suspension. At 5°C, the ε form was the first to appear at all pH values and this form was also observed briefly at 25°C and pH 6.8. This form converted first to the ζ form and then to the η form at pH 1.2. An increase in the pH resulted in a decrease in the number of the polymorphic forms observed. At pH 4.5, the ε form converted to the ζ form only, whereas, at pH 6.8 it converted in to the α form. The solution concentration profiles suggest that the relative stability of the polymorphs observed at 5°C from lowest to highest is ε < ζ < η < α. Interestingly, at 5°C, the α form was not observed for the duration of experiment except at pH 6.8.

3.2 Physicochemical characterization of the solid state forms HPLC analysis of the amorphous and all observed polymorphs resulted in a peak with a retention time of 21.4 min and peak areas within 1.96% variation for all samples. This, together with the absence of other peaks in all samples, suggests degradation was negligible and all solids analyzed were chemically indistinguishable from the starting material. This was further supported by analyzing the physical stability of the ζ and η forms of indomethacin in suspension at 25°C, where they both transformed into the α form. Since all the new solid state forms were observed at pH 1.2 (where the drug was almost completely unionized and no possible counterion was available in this medium) the possibility of salt formation was excluded. Nevertheless, the HPLC results for all the forms at all pH values and temperatures also confirm that the new forms obtained were not salts of indomethacin. There was no evidence of dehydration endotherms in the DSC thermograms of the new forms (see Figure 7) and also thermogravimetric analysis did not show (data not shown) any significant weight loss, confirming that none of the new forms were hydrates. In addition to the widely reported α and γ forms (whose crystal structures have been solved)21,

24, 27, 28

, the β and δ forms have also been reported19,

22

, as well as the form

reported by Lin, which we believe to be the same as the ε form identified in the present study20. We attempted to prepare all these solid state forms for comparison. The characterization of the α, γ and δ forms are also presented below. We were unable to prepare the β form as reported by Borka.19 The ζ and η forms identified during the study were further characterized by XRPD (Figure 5). The ε form was not sufficiently stable to allow its diffractogram to be recorded. The XRPD patterns were then compared with the patterns of the α and γ forms reported in the Cambridge Structural Database (ref codes INDMET02 and INDMET respectively) and that of the δ form generated in the laboratory. Amorphous indomethacin featured a halo in the

15 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

diffractogram. The unreported forms, i.e. ζ and η forms exhibited several unique diffraction peaks that distinguished them from the known crystal forms (Table 2). Table 2. Unique XRPD peaks and thermal properties of indomethacin solid state forms XRPD DSC Solid state form

Unique diffraction peak positions (°2θ)

Onset temperature (°C) Tg (°C)

Exothermic event

43.3 ± 0.9

101.2 ± 2.2

Endothermic event

Amorphous

NA

153.8 ± 0.4,158.1 ± 1.1

α

7.0, 8.5, 11.6, 12.0, 14.0 8.0

NA

ND

154.1 ± 0.6

γ

10.2, 11.8, 17.0, 19.9, 21.9

NA

ND

159.2 ± 0.1

δ

9.6, 10.5, 11.3, 13.0, 14.9

NA

134.1 ± 0.2

ζ

6.5, 11.0, 11.8, 12.8, 14.4, 16.4

NA

ND

η

9.1, 9.3, 12.2, 18.2, 20.5

NA

114.1 ± 0.6

129.2 ± 0.4, 158.1 ± 0.2 142.2 ± 0.4 157.0 ± 0.3 154.2 ± 0.2

NA: Not applicable ND: Not detected

These differences in the XRPD patterns indicate different arrangements of the indomethacin molecules in the crystal lattice of each form. However, the crystal structures of the new forms need to be solved for a detailed understanding about these differences.

Figure 5. Diffractograms of different solid state forms of indomethacin. Amorph = amorphous.

The IR and Raman spectra of the solid state forms between 1000 and 1800 cm -1 are shown in Figure 6. The IR and Raman spectra of the α, γ, and amorphous samples, and the IR spectrum of the δ form, have been documented.21, 23, 25, 29, 30 The spectrum of the ε form was consistent with that reported by Lin20. The β form is reported to have peaks at 1692, 1675 and 1606 cm-1.20 None of the forms obtained in this study had peaks at these positions.

16 ACS Paragon Plus Environment

Page 17 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

Figure 6. IR (a) and Raman (b) spectra of different solid state forms of indomethacin. Amorph =amorphous.

Peak assignments have been made based on both experimental evidence29 and previously published theoretical predictions.25 The peak positions associated with benzoyl and acid C=O stretching of the α, γ and amorphous forms, together with the tentative peak assignments for the other solid state forms, are given in Table 3. The benzoyl carbonyl (Hbond acceptor) and carboxylic acid (H-bond acceptor and donor) are the only two groups capable of participation in H-bonding in the indomethacin molecule, and interpretation of the C=O peaks in these groups can help to provide insight into the H-bonding arrangements in the different solid state forms whose crystal structures have not been solved. To help with vibrational assignment of these peaks, theoretical IR and Raman spectra of monomer, dimer (corresponding to the dimer structure in the γ form, with cyclic H-bonding between the carboxylic acid groups) and trimer structures (corresponding to the asymmetric unit of the α form, with H-bonding between carboxylic acid groups of first two molecules and

17 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 22

H-bonding between benzoyl carbonyl group of second molecule and carboxylic acid group of third molecule) have previously been determined.25, 30 -1

Table 3. Calculated and experimental wavenumbers (cm ) for C=O stretching modes in different solid state forms Assignment Calculated a

monomer

Calculated

Calculated trimerb

α

γ

εd

δ

ζ

η

Amorphous

1678

a

dimer

1644 Benzoyl C=O stretching

1705 (unbound)

1692

1657 (H-bonded to 1704 (unbound)

carboxylic acid group),

1649, 1680

1705 (unbound)

(IR), 1698 (Raman)

1675 (IR),1690 c

Raman)

(IR,

1635

1669

Raman),

(IR),

(IR),

(IR)

1679

1642

1681

(IR,

(Raman)

(Raman)

1728

1707 (IR),

(IR)

1737 (IR)

Raman)

Acid O-C=O stretch

1735 (H-bonded 1791

to second

(unbound) carboxylic acid, out of phase)

1734 (H-bonded to second carboxylic acid, out of phase vibration), 1757 (H- bonded to benzoyl C=O group)

1691

1694

1699

(IR and

1714

(IR),

1711

Raman) c,

(IR)

1690

(IR)

1735 (IR)

(Raman)c

(shoulder), 1724 (both IR only)

a

From Strachan et al 2007; dimer with H-bonding between carboxylic acid groups (corresponding to dimer structure in γ form)

b

From Heinz et al 2008; trimer with H-bonding between carboxylic acid groups of first two molecules and H bonding between benzoyl carbonyl

c

Possibly Raman active modes involving coupling of benzoyl and acid C=O groups

d

Only the IR spectrum of the ε form recorded; insufficient stability prevented the Raman spectrum being recorded

group of second molecule and carboxylic acid group of third molecule (corresponding to asymmetric unit of α form)

The calculations

support

experimental evidence that the

predicted vibrational

wavenumbers for both the benzoyl and acid C=O stretching are affected by H-bonding, with the modes shifting to lower wavenumbers upon H-bonding. As a result, the trimer (and α form) vibrations involving both the benzoyl and acid C=O groups appeared at two wavenumbers in the IR spectra, since both groups exist in both H-bonded and unbound arrangements. In the simpler dimer structure (and γ form), only one peak is observed in the IR spectrum for each moiety. These interpretations may be extended to the δ, ε, ζ and η forms to gain some insight into the nature of H-bonding in these crystal structures. For the benzoyl C=O stretch, a mode occurs at 1692 cm-1 (IR) for the γ form (non H-bonded) and 1649 cm -1 (H-bonded) and 1680 cm -1 (non H-bonded) for the α form. For the δ and ε forms, benzoyl C=O stretching occurs at 1675 cm-1 (IR) and 1669 cm-1 respectively, suggesting that H-bonding of this group is absent or weak in these solid state forms. In the η form, the appearance of the band at 1635 cm-1 (IR), suggests strong H-bonding of the benzoyl C=O, while the benzoyl C=O group in the ζ form is both H-bonded (1644 cm-1) and not or weakly H-bonded (1679 cm-1). Evidence of variation in H-bonding is also provided by the antisymmetric acid C=O stretching mode. Peaks appear at 1714 cm-1 for the γ form (with cyclic COOH bonding in the dimer), and at 1691 cm-1 (H-bonding to benzoyl C=O, possibly with some vibrational coupling, also evidenced by a band at 1692 in the Raman spectrum) and 1735 cm-1 for the α form (cyclic COOH bonding). The δ, ε, ζ, and η forms exhibit peaks at various wavenumbers. 18 ACS Paragon Plus Environment

Page 19 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

The ζ form exhibits two bands (1694 cm-1 and 1724 cm-1). Together with the benzoyl C=O stretching evidence, this suggests at least two H-bonding interactions in the crystal, one between two acid groups, and the other between the benzoyl carbonyl and acid group. The single acid C=O stretching band observed at 1728 cm-1 for the η form, combined with evidence of strong H-bonding involving the benzoyl C=O group, suggests that acid-benzoyl carbonyl interactions may be the predominant form of H-bonding interaction in this polymorph. Some additional evidence for a lack of cyclic COOH dimerization in the η form is provided by the position of an OH stretching band in the Raman spectrum, which appears between 2929 and 2931 cm -1 for the all forms except the η form, where it appears at 2943 cm -1 (the spectrum of the ε form was not recorded). The thermal behaviour of all solid state forms in this study, except for the ε form, was determined with DSC (Figure 7). The ε form was too unstable for analysis and transformed into the α form during drying. The α, β, γ, and δ forms have been described by Borka19, with the α form having a melting point onset of 154 to 155°C, β at 148°C, γ at 161°C and δ at 129°C.19,

23

Out of those, the α and γ forms have been isolated in pure form and

characterized in depth.20, 31, 32 We have found no other reports of the β form, and have also not been able to prepare it ourselves according to the method described by Borka.19 None of the forms characterized in this study had a melting onset of 148°C, corresponding to the β form. Thermograms obtained for α, γ, δ and amorphous samples were similar to those reported in the literature.19, 23, 31 The thermal properties of indomethacin polymorphs are summarized in Table 2. Single sharp melting endotherms were observed at 154 and 159°C for the α and γ forms respectively (all temperatures for melting and recrystallization events in the present study are onset temperatures), whereas the δ form melted at 129°C followed by recrystallization at 134°C and melting at 158°C. The amorphous form exhibited a Tg at 43°C (mid-point) followed by recrystallization at 101°C and two melting endotherms at 154 and 159°C. DSC thermograms of the ζ and η forms showed endo- and exothermic events at different temperatures. The ζ form featured two endothermic events at 142 and 156°C. The η form showed an exothermic event at 114°C followed by endothermic event at 154°C. The exothermic event was further probed by heating the η form up to 130°C, cooling to 0°C and again reheating up to 180°C. The absence of an exothermic event during reheating indicates a solid-solid phase transition occurred, with the subsequent melting endotherm suggesting that it transformed to the α form. The two endothermic events in the thermogram of ζ form were not sufficiently resolved to probe the solid state form between the two endothermic events by this method. 19 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 22

Figure 7. Thermograms of different solid state forms of indomethacin. Amorph = amorphous.

It is interesting to note that despite indomethacin being such a widely studied drug with respect to its solid state properties, new polymorphs were discovered. However, amorphous suspensions are not normally employed during polymorph screening. During manufacturing and storage, the forms reported in this study might be most likely to appear if the amorphous form were exposed to a low pH aqueous environment below room temperature. However, the presence of different excipients may also affect the polymorphic forms observed, and therefore the forms might potentially appear in other processing or storage environments. In any case, if unexpected solid state forms were observed, the current study may help in identifying the forms.

4. CONCLUSION The crystallization behaviour of amorphous indomethacin in aqueous suspension was studied as a function of pH and temperature. The pH and temperature of the suspensions affected the crystallization onset and rate, with lower pH and temperature leading to longer onset and slower crystallization. The resulting polymorphic forms were diverse. At 25°C, the amorphous form recrystallized directly to the α form, except at pH 6.8, where it converted into the ε form prior to the α form. Overall, at 5°C, three new polymorphic forms, named ε, ζ and η, were observed, depending on the pH values. The total number of these forms observed increased with decreasing pH, with the order of appearance always being: ε, ζ then η. The three new forms exhibited distinct DSC, XRPD, FTIR and Raman spectroscopy profiles. The solution concentration profiles suggest that the relative stability of the polymorphs at 5°C from lowest to highest is ε < ζ < η < α. By showing that as yet unknown polymorphs can be obtained from amorphous suspensions even for a compound as 20 ACS Paragon Plus Environment

Page 21 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Molecular Pharmaceutics

extensively studied as indomethacin, this study demonstrates the important role amorphous suspension studies may play in exhaustive polymorph screens.

5. ACKNOWLEDGEMENTS The authors thank Dr Jacco van der Streek and Prof. Thomas Rades (University of Copenhagen) and Prof. Andrew Bond (University of Southern Denmark) for helpful suggestions. SS is funded by a University of Otago Doctoral Scholarship.

6. SUPPORTING INFORMATION Corresponding loadings plot of all the scores plots in this paper are available. This material is available free of charge via the Internet at http://pubs.acs.org.

7. REFERENCES 1. Csermely, P.; Korcsmaros, T.; Kiss, H. J.; London, G.; Nussinov, R. Structure and dynamics of molecular networks: A novel paradigm of drug discovery: A comprehensive review. Pharmacol. Ther. 2013, 138, (3), 333-408. 2. Stegemann, S.; Leveiller, F.; Franchi, D.; de Jong, H.; Lindén, H. When poor solubility becomes an issue: From early stage to proof of concept. Eur. J. Pharm. Sci. 2007, 31, (5), 249-261. 3. Di, L.; Fish, P. V.; Mano, T. Bridging solubility between drug discovery and development. Drug Discov. Today 2012, 17, (9–10), 486-495. 4. Turner, J. R., Drug Discovery. In New Drug Development, Springer New York: 2010; pp 21-34. 5. Li, P.; Zhao, L. Developing early formulations: Practice and perspective. Int. J. Pharm. 2007, 341, (1–2), 1-19. 6. Baird, J. A.; Van Eerdenbrugh, B.; Taylor, L. S. A classification system to assess the crystallization tendency of organic molecules from undercooled melts. J. Pharm. Sci. 2010, 99, (9), 3787-3806. 7. Engers, D.; Teng, J.; Jimenez-Novoa, J.; Gent, P.; Hossack, S.; Campbell, C.; Thomson, J.; Ivanisevic, I.; Templeton, A.; Byrn, S.; Newman, A. A solid-state approach to enable early development compounds: Selection and animal bioavailability studies of an itraconazole amorphous solid dispersion. J. Pharm. Sci. 2010, 99, (9), 3901-3922. 8. Neervannan, S. Preclinical formulations for discovery and toxicology: physicochemical challenges. Expt. Opn. Drug Metabol. Tox. 2006, 2, (5), 715-731. 9. Hancock, B. C.; Zografi, G. Characteristics and significance of the amorphous state in pharmaceutical systems. J. Pharm. Sci. 1997, 86, (1), 1-12. 10. Laitinen, R.; Lobmann, K.; Strachan, C. J.; Grohganz, H.; Rades, T. Emerging trends in the stabilization of amorphous drugs. Int. J. Pharm. 2012, In press, corrected proof. 11. Bhugra, C.; Pikal, M. J. Role of thermodynamic, molecular, and kinetic factors in crystallization from the amorphous state. J. Pharm. Sci. 2008, 97, (4), 1329-1349. 12. Savolainen, M.; Kogermann, K.; Heinz, A.; Aaltonen, J.; Peltonen, L.; Strachan, C.; Yliruusi, J. Better understanding of dissolution behaviour of amorphous drugs by in situ solid-state analysis using Raman spectroscopy. Eur. J. Pharm. Biopharm. 2009, 71, (1), 71-79. 13. Wu, T.; Sun, Y.; Li, N.; de Villiers, M. M.; Yu, L. Inhibiting surface crystallization of amorphous Indomethacin by nanocoating. Langmuir 2007, 23, (9), 5148-5153.

21 ACS Paragon Plus Environment

Molecular Pharmaceutics

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 22

14.Alonzo, D.; Zhang, G.; Zhou, D.; Gao, Y.; Taylor, L. Understanding the behavior of amorphous pharmaceutical systems during dissolution. Pharm. Res. 2010, 27, (4), 608-618. 15. Greco, K.; Bogner, R. Crystallization of amorphous Indomethacin during dissolution: effect of processing and annealing. Mol. Pharm. 2010, 7, (5), 1406-1418. 16.Andronis, V.; Yoshioka, M.; Zografi, G. Effects of sorbed water on the crystallization of indomethacin from the amorphous state. J. Pharm. Sci. 1997, 86, (3), 346-351. 17.Andronis, V.; Zografi, G. Crystal nucleation and growth of indomethacin polymorphs from the amorphous state. J. Non-Cryst. Solids 2000, 271, (3), 236-248. 18. Hancock, B. C.; Shamblin, S. L.; Zografi, G. Molecular mobility of amorphous pharmaceutical solids below their glass transition temperatures. Pharm. Res. 1995, 12, (6), 799-806. 19. Borka, L. The polymorphism of indomethacin. New modifications, their melting behaviour and solubility. Acta Pharm. Suec. 1974, 11, 295-303. 20. Lin, S. Y. Isolation and solid-state characteristics of a new crystal form of indomethacin. J. Pharm. Sci. 1992, 81, (6), 572-576. 21. Savolainen, M.; Heinz, A.; Strachan, C.; Gordon, K. C.; Yliruusi, J.; Rades, T.; Sandler, N. Screening for differences in the amorphous state of indomethacin using multivariate visualization. Eur. J. Pharm. Sci. 2007, 30, (2), 113-123. 22. 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. 23. Rumondor, A. C. F.; Marsac, P. J.; Stanford, L. A.; Taylor, L. S. Phase behavior of poly(vinylpyrrolidone) containing amorphous solid dispersions in the presence of moisture. Mol. Pharm. 2009, 6, (5), 1492-1505. 24. Heinz, A.; Savolainen, M.; Rades, T.; Strachan, C. J. Quantifying ternary mixtures of different solid-state forms of indomethacin by Raman and near-infrared spectroscopy. Eur. J. Pharm. Sci. 2007, 32, (3), 182-192. 25. Strachan, C. J.; Rades, T.; Gordon, K. C. A theoretical and spectroscopic study of γ-crystalline and amorphous indometacin. J. Pharm. Pharmacol. 2007, 59, (2), 261-269. 26. Threlfall, T. Structural and thermodynamic explanations of Ostwald's rule. Org. Process Res. Dev. 2003, 7, (6), 1017-1027. 27.Aceves-Hernandez, J. M.; Nicolás-Vázquez, I.; Aceves, F. J.; Hinojosa-Torres, J.; Paz, M.; Castaño, V. M. Indomethacin polymorphs: Experimental and conformational analysis. J. Pharm. Sci. 2009, 98, (7), 2448-2463. 28.Aubrey-Medendorp, C.; Swadley, M.; Li, T. The polymorphism of indomethacin: An analysis by density functional theory calculations. Pharm. Res. 2008, 25, (4), 953-959. 29. Taylor, L. S.; Zografi, G. Spectroscopic characterization of interactions between PVP and Indomethacin in amorphous molecular dispersions. Pharm. Res. 1997, 14, (12), 1691-1698. 30. Heinz, A.; Strachan, C.; Gordon, K.; Rades, T., Insight into different solid-state forms of indomethacin using vibrational spectroscopy, multivariate analysis, and quantum chemical modeling. In American Association of Pharmaceutical Scientists, Georgia World Congress Center, Atlanta, 2008; pp AAPS2008-001461. 31. Priemel, P. A.; Grohganz, H.; Gordon, K. C.; Rades, T.; Strachan, C. J. The impact of surfaceand nano-crystallisation on the detected amorphous content and the dissolution behaviour of amorphous indomethacin. Eur. J. Pharm. Biopharm. 2012, 82, (1), 187-193. 32. Nicolaï, B.; Céolin, R.; Rietveld, I. Polymorphism and solvation of indomethacin. J Therm Anal Calorim 2010, 102, (1), 211-216.

22 ACS Paragon Plus Environment