Can vacuum morphologies predict solubility and intrinsic dissolution rate? A case study with felodipine polymorph form IV

Accepted Manuscript Title: Can vacuum morphologies predict solubility and Intrinsic Dissolution Rate? A Case Study with Felodipine Polymorph Form IV Author: Dinesh Kumar Rajesh Thipparaboina Nalini R Shastri PII: DOI: Reference:

S1877-7503(15)00036-8 http://dx.doi.org/doi:10.1016/j.jocs.2015.03.009 JOCS 341

To appear in: Received date: Revised date: Accepted date:

29-10-2014 5-3-2015 24-3-2015

Please cite this article as: Dinesh Kumar, Rajesh Thipparaboina, Nalini R Shastri, Can vacuum morphologies predict solubility and Intrinsic Dissolution Rate? A Case Study with Felodipine Polymorph Form IV, Journal of Computational Science http://dx.doi.org/10.1016/j.jocs.2015.03.009 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Can vacuum morphologies predict solubility and Intrinsic Dissolution Rate? A Case Study with Felodipine Polymorph Form IV Dinesh Kumara, Rajesh Thipparaboinaa, Nalini R Shastria,* a

National Institute of Pharmaceutical Education & Research, Hyderabad, India

*Corresponding author. Nalini R Shastri Tel. +91-040-23423749 Fax. +91-040-23073751 E-mail: [email protected], [email protected] Address: Associate Professor, Department of Pharmaceutics, NIPER (National Institute of Pharmaceutical Education & Research), Balanagar, Hyderabad, India, Pin Code – 500037

1

Abstract The impact of Felodipine (Fel) polymorphism on vacuum morphology was studied and correlated with the structural properties like solubility and intrinsic dissolution rate (IDR). A correlation was established between solubility and IDR of three Fel polymorphs with their BFDH aspect ratio, growth morphology aspect ratio and polar/non polar ratio. The predicted solubility and IDR values for form IV by three methods were in agreement, however, morphology growth aspect ratio model showed better prediction capability due to its higher coefficient of determination. The solubility for form IV was 0.0154 mol.l-1 while the IDR was 0.246 mg.min-1.cm-2 for growth morphology aspect ratio. KEY WORDS: Habit, felodipine, solubility, IDR, polymorph 1. Introduction Different polymorphs of any given API (active pharmaceutical ingredient) can have different physicochemical properties such as melting point, solubility, dissolution rate, and oral bioavailability; which in turn may affect their adequacy in drug formulations [1, 2]. Prediction of crystal structures on the basis of molecular information [3-5] and use of a variety of theoretical methods to generate possible crystal structures [6] has thus resulted in increased interest of pharmaceutical industries toward the crystal morphology prediction. This may be due to the fact that crystal habits, especially the preferred equi-dimensional habit, tend to show considerable impact on the pharmaceutical and biopharmaceutical properties of API [7, 8]. Queries on growth mechanism and growth rate of crystals and simultaneous research work related to different crystal morphologies in various environments are increasing day by day. A detailed information on the growth mechanism of crystals usually aids in controlling purity, cost of manufacture and end-use [9, 10]. Crystal habit simulations have advanced to a state where habit prediction for drug molecules is relatively straightforward [11]. With the help of molecular simulation tools, the crystal habit prediction along with solvent and additive interactions has become feasible [11, 12]. Bravais-Friedel-Donnay-Harker (BFDH), morphology growth (MG) and equilibrium morphology (EM) models are widely employed for comparing the morphologies [1316]. 2

In order to determine the surface chemistry of a specific crystal facet, information on crystal structure and the miller indices of that particular crystal face is required [17-19]. Thus, in an effort to demonstrate the computational efficiency for optimizing the crystal morphology, various reported morphology models were analyzed for different polymorphs of a single drug molecule, felodipine (Fel). Fel is a calcium channel inhibitor, which is widely recommended for treatment of hypertension and prevention of angina pectoris. Fel belongs to class II of BCS (biopharmaceutical classification system) scheme and is practically insoluble in aqueous medium [20]. The structure, pharmaceutical and biopharmaceutical profiles of three polymorphs of Fel (form I, II and III) are established and reported in literature [2]. The crystal structure of Fel form I (marketed product) is well described and is reported to be the most stable form by R Fossheim [21]. Form II was first discussed by Srcic et al., in 1992 [22] and its structure was reported by Lou et al., in 2009 [23]. Surov et al., 2012 have reported form III and IV along with their crystal structures. Additionally, Surov et al., has also reported the solubility and IDR (Intrinsic dissolution rate) for I, II and III but were unable to calculate the solubility and IDR of form IV due to limited amount of crystals available [2]. Hence, the study was further extended to predict important pharmaceutical properties like solubility and IDR for form IV. Aspect ratio of crystal habit and distribution of functional groups exposed to the most relevant crystal faces was calculated from vacuum morphology models for three polymorphs of Fel and a correlation was established between surface structural parameters and their solubility/IDR. On basis of these correlations, the solubility and IDR of form IV was predicted. 2. Methods for computer simulation Crystal structures of Fel polymorphs were obtained from CSD (Cambridge Structural Database). Crystal dimensions were defined in terms of length, height, width as a, b, c and angles between them as α, β and γ respectively. Fel form I crystallographic information file (DONTIJ) was reported by R Fossheim [21] with the following cell parameters: Symmetry: Monoclinic P21/c, a: 12.086 Å, b: 12.077 Å, c: 13.425 Å, α: 90, β: 116.13, γ: 90. Fel form II crystallographic information file (DONTIJ01) was reported by Lou et al., in 2009 [23] with the following cell parameters: Symmetry: Monoclinic C2/c, a: 32.392 Å, b: 18.717 Å, c: 23.771 Å, α: 90, β: 91, γ: 90. Fel form III 3

crystallographic information file (864026) was reported by Surov et al., in 2012 [2] with the following cell parameters: Symmetry: Monoclinic P21/n, a: 15.1255 Å, b: 7.2302 Å, c: 17.2796 Å, α: 90, β: 110.198, γ: 90., Fel form IV crystallographic information file (864027) was also reported by Surov et al., in 2012 [2] with the following cell parameters: Symmetry: Monoclinic P21/n, a: 11.1129 Å, b: 12.5688 Å, c: 13.4969 Å, α: 90, β: 107.009, γ: 90. The crystal morphology modelling procedure was developed on basis of the reported literature [24]. Prediction and study of possible crystal morphologies was performed using a preliminary equilibration protocol, by means of the Morphology package included in the Material Studio 6.1 package of Accelrys, adopting the molecular mechanics approximation and the COMPASS (condensed phase optimized molecular potentials for atomistic simulation studies) force field. Geometry optimization was done with forcite algorithm with COMPASS force field. Face list was generated using morphology calculation which gave hkl values of important faces with dhkl (centre to plane distance) values. The morphology prediction tools consist of three different computational approaches: BFDH, MG, and EM methods. The first vacuum model used was BFDH, which generated a list of possible growth faces [19]. The second vacuum morphology model used was the attachment energy also known as MG method. The MG method assumes that the growth rate of a crystal face is proportional to its attachment energy, i.e., faces with the lowest attachment energies are the slowest growing and, therefore, have the morphological importance [25-27]. The third prediction model for vacuum morphology used was surface free energy model, which is also known as EM method. The surface energy at a temperature of 0 K, was calculated by EM model [25]. In this study, the reported solubility and IDR of polymorphs were correlated with their various structure and morphology related factors like aspect ratio, polar/non-polar, surface/volume (S/V) ratio, attachment energy and surface energy. Simple linear regression equations were obtained using three reported polymorphs data. From the obtained equations, values were plugged in to estimate the solubility and IDR of Fel polymorph IV. Hirshfeld surface analysis of intermolecular interactions for each polymorph was performed using Crystal Explorer (Version 3.0). This alternative way was employed to assess the differences among polymorphs, by comparing the intermolecular interactions a molecule makes with its neighbors. Felodipine polymorphs were comparing taking 4

into account the molecular conformation differences amongst the polymorphs, and the absence of strong hydrogen bonding that limits the utility of topological descriptions[28].

The Hirshfeld surface defines each independent molecule’s

environment within a crystal. This information was used to describe the dissolution potential of a particular crystal structure. 3. Results and discussions 3.1 Form I Crystal structure of Fel form I is shown in fig 1a. Vacuum morphology of form I was generated by BFDH model (fig. 1b), which gave 6 important facets along with their planes (hkl), centre to plane distance (dhkl) and % surface area. The BFDH method is an approximation and does not account for the any kind of energetics of the system [25]. The accuracy of method reduces inversely with the bonding strength of system [26]. Thus, the only benefit of this method was to identify important faces in the growth process [26, 29]. Table1 lists the inter-planar spacings of various low index faces of the crystal habit of form I based on the BFDH calculation. Result in table1 & fig. 1b, shows that the crystal faces consisted of (1 0 0), (0 1 1), (1 1 0), (1 1 -1) and (1 0 -2) planes, of which (0 1 1) with 37.48%, and (1 0 0) plane with 26.49% of the total facet area were the most important faces. The calculated aspect ratio by BFDH morphology for form I was 1.81 and the surface area/volume ratio (S/V) ratio was 1.123 (table 2). In the second step, morphology was determined by MG model, which shows (Table 1 & fig. 1c) that the crystal faces consist of (1 0 0), (0 1 1), (1 1 0), (1 1 -1), (1 0 -2) & (0 2 0) planes, of which (0 1 1) with 43.56%, and (1 0 0) plane with 32.95% of total facet area were the most important faces. The calculated aspect ratio for MG model of form I was 2.15 and S/V ratio was found 1.16. In the third step, morphology was determined by EM. This method determined the EM of the crystal by calculating the minimum of the surface free energy for a given volume and temperature [26, 29]. The crystal faces consisted of (1 0 0), (0 1 1), (1 1 -1), (1 0 -2), (0 2 0), (1 1 1) and (2 1 -1) planes, of which (0 1 1) with 28.95%, and (1 0 0) plane with 17.69% of total facet area were the most important faces (table1 & fig. 1d). The calculated aspect ratio for EM model of form I was 1.44 and the S/V ratio was 1.06.

5

Surface structures of all-important facets of form I given by MG model were studied (fig. 2). The most dominant facet (0 1 1) was covered by one methyl & one chloride, whereas second dominant facet (1 0 0) was covered by 2 methyl and 2 carbonyl (table 1) and the third dominant facet (1 1 -1) was covered by 1 methyl, 2 carbonyl, one chloride and 1 aromatic ring. After calculating the functional groups present on all facets, it was observed that ~46% of crystal surface of form I was covered by polar functional groups and ~54 % by non polar functional groups. Based on the literature, form I expressed solubility ~0.0148 mol.l-1 and IDR ~0.235 mg.min-1.cm-2[2]. 3.2 Form II Crystal structure of Fel form II is shown in fig 3a. Vacuum morphology of form II generated by BFDH model (table1 and fig. 3b) gave six important unique facets; (1 1 0), (2 0 0), (1 1 -1), (1 1 1), (0 0 2) and (2 0 -2). The calculated aspect ratio for BFDH morphology of form II was 1.68 and S/V ratio was 1.10 (table 3). Crystal faces of (1 1 0), (2 0 0), (1 1 -1), (1 1 1), and (0 0 2) planes were the most important faces for form II as calculated by MG model (table 1 & fig. 3c) giving a calculated aspect ratio of 1.55 and the S/V ratio of 1.12. Morphology determined by EM model gave crystal faces of (2 0 0), (1 1 -1), (1 1 1), (0 0 2), (1 1 2), and (0 0 2) planes as the most important faces (table1 & fig. 3d) with a calculated aspect ratio and S/V ratio of 1.69 and 1.07 respectively. The calculated results obtained with form II were different when compared with form I for all morphologies studied. To understand the underlying cause, surface structures exposed on important facets of form II given by MG model were studied (fig. 4). The most dominant facet (0 0 2) was covered by one methyl, 1 carbonyl & one chloride, whereas the second dominant facet (2 0 0) was covered by 3 methyl and 2 chloride. The third dominant facet (1 1 0) was covered by 5 methyl, 1 carbonyl, 2 chloride and 1 aromatic ring while the fourth dominant facet (1 1 1) was covered by 3 methyl, 1 carbonyl, 3 chloride and 1 aromatic ring, which on computing gave ~53% crystal surface coverage by polar functional groups and ~47 % by non polar functional groups. Form II expressed highest solubility (0.0156 mol.l-1) and IDR (0.247 mg.min-1.cm-2) [2]. This can be easily correlated to highest abundance of polar functional groups on its major facets. This conclusion was also supported by Hirschfled surface analysis. 6

3.3 Form III Crystal structure of Fel form III is shown in fig 5a. Vacuum morphology of form III generated by BFDH model (fig. 6a), consisted of six important facets; (1 0 -1), (1 0 1), (0 0 2), (0 1 1) (1 1 0), and (1 1 -1) (table 1 & fig. 5b) with a calculated aspect ratio (2.20) and S/V ratio (1.18) (table 3). The MG model (table1 & fig. 5c), gave important crystal faces of (1 0 -1), (1 0 1), (0 1 1), (1 1 0) and (1 1 -1) planes and calculated aspect ratio and S/V ratio for MG model of form III as 2.74 and 1.25 respectively. Important crystal faces of (1 0 -1), (1 0 1), (1 1 -1), (1 0 -3), (0 1 2) and (2 1 0) were obtained for morphology determined by EM model (table1 & fig. 5d). The calculated aspect ratio for MG model of form III was 1.49 and S/V ratio was 1.07. The most dominant facet of form III; (1 0 -1) was covered by 2 methyl, 1 chloride, the second dominant facet (1 0 1) was covered by 3 methyl and 2 carbonyl whereas the third dominant facet (1 1 -1) was covered by 2 methyl, 1 carbonyl, and 1 chloride (fig. 6). Around 44% of crystal surface of form III was covered by polar functional groups and ~56 % by non polar functional groups. Based on Surov et al., experiments, form III showed least solubility (0.0145 mol.l-1) and IDR (0.195 mg.min-1.cm-2) [2]. This data can be easily correlated to least abundance of polar functional groups on its major facets. 3.4 Form IV Crystal structure of Fel form IV is shown in fig 7a. Vacuum morphology of form IV generated by BFDH model (fig. 7b) gave five important unique facets; (1 0 -1), (0 1 1) (1 1 0), (1 1 -1) and (1 0 1) (table 1 & fig. 7b). The calculated aspect ratio and S/V ratio by BFDH morphology for form IV was 1.53 and 1.11 respectively (table 3). Crystal faces by MG model (table 1 & fig. 7c), consisted of (1 0 -1), (0 1 1) (1 1 0), (1 1 -1), (1 0 1) and (0 0 2) planes while the calculated aspect ratio and S/V ratio was 1.66 and 1.11 respectively. The EM model (table 1 & fig. 7d), showed crystal faces of (1 0 -1), (0 1 1), (1 1 -1), (1 0 1), (0 0 2), (0 2 0) and (1 1 1) planes giving a calculated aspect ratio and S/V ratio of 1.29 and 1.05 respectively. Notable difference in the aspect and S/V ratios were obtained with all forms of Fel, which was partially attributed to difference in the abundance of the surface groups on important facets. The most dominant facet of form IV given by MG model (0 1 1) was covered by 3 methyl and 1 carbonyl, whereas 7

second dominant facet (1 0 -1) was dominated by 2 chloride groups. The third dominant facet is (1 1 0) was covered by 4 methyl and 2 chloride (fig. 8). After determining the functional groups present on all other facets, it was observed that ~46% of crystal surface of form IV was covered by polar functional groups and ~54 % was covered by non polar functional groups. This led to an understanding that the relatively higher abundance of polar functional group exposed on crystal morphology of form II was responsible for its lower aspect ratio and lower S/V ratios. 3.5 Prediction of solubility and IDR of form IV The facets of a growing crystal can be considered as composed of ‘active sites’ that can have definite interactions with molecules in medium. These interactions depend upon surface functionality, which in turn decides the solubility and IDR for a crystal. In present study, the reported solubility and IDR of polymorphs were correlated with various structure and morphology related factors like aspect ratio, polar/non-polar, S/V ratio, attachment energy and surface energy. The solubility and intrinsic dissolution rate (IDR) of only three polymorphs were reported. Solubility and IDR experiments for form I, II, and III were carried out by Surov et al., 2012 by using the shake- flask method. 50% ethanol−water mixture was used as medium [2]. Based on their experiments, form II (0.0156 mol.l-1) expressed highest solubility, followed by form I (0.0148 mol.l-1) and form III (0.0145 mol.l-1) respectively. Similarly, IDR also followed the same order where form II (0.247 mg.min-1.cm-2) with highest IDR, was followed by form I (0.235 mg.min-1.cm-2) and form III (0.195 mg.min-1.cm-2). Correlation coefficient was checked between BFDH aspect ratio, MG aspect ratio, EM aspect ratio, polar/non-polar (P/NP) ratio, surface energy and attachment energy of Fel polymorphs with their respective solubility and IDR. However, good correlation was obtained only between BFDH aspect ratio, MG aspect ratio and polar/non polar ratio of Fel polymorphs with their respective solubility and IDR (table 3). The calculated BFDH aspect ratio shows the order III (2.2) > I (1.81) > II (1.68). The habit with aspect ratio value near to one generally shows better solubility and IDR due to symmetrical morphology. We observed that, the rank order of BFDH aspect ratio (III > I > II), correlated well with the sequence of solubility and IDR (II > I > III). Similar kind of rank order correlation was observed between MG aspect ratio and solubility/IDR. The sequence order of polar/non polar ratio (II > I > III), also resembled the sequence of 8

solubility and IDR. Polarity can be directly related to solubility and IDR in aqueous medium [30]. Increased polarity of crystal surface usually leads to improved dissolution of the drug crystal in aqueous media. Fitting in the calculated values of various parameters of form IV morphology, a rank order correlation of II > I/IV > III was obtained for solubility and IDR for Fel polymorphs. The solubility and IDR of form IV were predicted by using simple linear correlations (y = mx + c, where y = solubility or IDR, x = aspect ratio or polar/nonpolar ratio). The solubility and IDR for form IV was calculated as 0.0156 mol.l-1 (from the equation y = 0.00180x + 0.01839, r2=0.737) and 0.263 mg.min-1.cm-2 (y = -0.1005x + 0.4164, r2=0.999) respectively from BFDH aspect ratio equation, whereas from MG aspect ratio equation, the solubility and IDR for form IV were predicted as 0.0154 mol.l-1 (y = 0.00092x + 0.01695, r2=0.937) and 0.246 mg.min-1.cm-2 (y = -0.0436x + 0.3193, r2=0.909) respectively. The solubility and IDR for form IV were calculated as 0.0148 mol.l-1 (y = 0.00304x + 0.1216, r2=0.989) and 0.223 mg.min-1.cm-2 (y = 0.115x + 0.119, r2=0.624) respectively from polar/non polar correlation. All the predicted values for form IV showed good agreement with each other. However, from all the above models obtained, based on the higher coefficient of determination values (r2 > 0.9), morphology growth model was selected to predict solubility and IDR for polymorph IV. Thus, the predicted solubility and IDR of form IV was found between solubility and IDR values of form I and form III. This rank order in solubility and IDR was also consistent with the facet chemistry of the given forms. Form II showed greater solubility due to a large number of polar functional groups at the morphologically important surfaces (section 3.2). The proportion of polar facet area (46%) and nonpolar facet area (54%) for form IV was similar to the proportion of polar facet area (46%) and non-polar facet area (54%) for form I. These values were also closer to the proportion of polar facet area (44%) and non-polar facet area (56%) for form III, resulting in the solubility and IDR values of form IV lesser than form II but closer to form I and form III. Hirshfled surface analysis was employed to identify and characterize various aspects of different crystal environments. The sum of the molecular interactions is depicted as a fingerprint plot in fig. 9. The largest % of contacts were found between H···H atoms which may not play a significant role in dissolution and solubility. Most of the 9

difference in solubility and dissolution is anticipated to arise from the polar or relatively non polar (O···H, H··Cl, N···H, O···Cl, C···Cl) contacts which showed a rank order of II (35.3%) > IV (35.2%) > I (34.9%) > III (34.6%) for Fel polymorphs. This also showed good agreement with predicted values substantiating the suitability of the model in prediction of solubility and IDR of polymorphs. 4. Conclusions The difference in vacuum morphologies of different polymorphs of Fel was discussed. The solubility and IDR of different polymorphs was correlated to the relative presence of polar and non-polar functional groups. Form II showed greater solubility by virtue of its large number of polar functional groups at the morphologically important surfaces. The solubility of form IV was predicted based on facet structure, which was found between solubility values of form I and form III. In conclusion, the concept of using molecular modeling to predict the crystal morphology and to determine the surface structure of particular faces of a crystal has been successfully applied to predict and calculate the solubility of a polymorph of Fel. This kind of research work can be helpful in understanding the impact of polymorphism on solubility and dissolution rate with limited resources available. Similarly, it is presumed that such type of study would broaden our information related to the face-specific properties of drug crystals and have implications when considering formulation technology and drug delivery.

Acknowledgments The authors acknowledge financial support from the National Institute of Pharmaceutical Education & Research (NIPER), Hyderabad, India and Indian Institute of Chemical Technology (IICT), Hyderabad, India. 5. References [1] D. Kaul, N.T. Nguyen, S. Venkataram, Crystal habit modifications and altered tabletting characteristics, Int. J. Pharm., 88 (1992) 345-350. [2] A.O. Surov, K.A. Solanko, A.D. Bond, G.L. Perlovich, A. Bauer-Brandl, Crystallization and polymorphism of felodipine, Cryst. Growth Des., 12 (2012) 4022-4030. [3] G. Clydesdale, K.J. Roberts, R. Docherty, HABIT95-a program for predicting the morphology of molecular crystals as a function of the growth environment, J. Cryst. Growth, 166 (1996) 78-83. [4] A. Llinas, J.M. Goodman, Polymorph control: past, present and future, Drug Disc. Today, 13 (2008) 198-210. 10

[5] S.L. Price, The computational prediction of pharmaceutical crystal structures and polymorphism, Adv. Drug Deliv. Rev., 56 (2004) 301-319. [6] J.J. Lu, J. Ulrich, An improved prediction model of morphological modifications of organic crystals induced by additives, Cryst. Res. Tech., 38 (2003) 63-73. [7] H.M. Burt, A.G. Mitchell, Effect of habit modification on dissolution rate, Int. J. Pharm., 5 (1980) 239-251. [8] R. Witzleb, V.-R. Kanikanti, H.-J. Hamann, P. Kleinebudde, Influence of needle-shaped drug particles on the solid lipid extrusion process, Powder Tech., 207 (2011) 407-413. [9] F. Otalora, J. Garcia-Ruiz, Nucleation and growth of the Naica giant gypsum crystals, Chem. Soc. Rev., (2014). [10] L.-F. Huang, W.-Q.T. Tong, Impact of solid state properties on developability assessment of drug candidates, Adv. Drug Deliv. Rev., 56 (2004) 321-334. [11] D. Winn, M.F. Doherty, Predicting the shape of organic crystals grown from polar solvents, Chem. Eng. Sci., 57 (2002) 1805-1813. [12] D. Winn, M.F. Doherty, Modeling crystal shapes of organic materials grown from solution, AIChE Journal, 46 (2000) 1348-1367. [13] E. Ramachandran, S. Ramukutty, Growth, morphology, spectral and thermal studies of gel grown diclofenac acid crystals, J. Cryst. Growth, 389 (2014) 78-82. [14] Z. Liang, Q. Yi, W. Wang, X. Han, J. Chen, Y. Le, J. Wang, C. Xue, H. Zhao, A systematic study of solvent effect on the crystal habit of dirithromycin solvates by computer simulation, Comp. & Chem. Eng., 62 (2014) 56-61. [15] V. Natarajan, M. Arivanandhan, P. Anandan, K. Sankaranarayanan, G. Ravi, Y. Inatomi, Y. Hayakawa, In-situ observation of faceted growth of benzophenone single crystals, Mat. Chem. Phys., (2014). [16] M.A. Deij, J. van Eupen, H. Meekes, P. Verwer, P. Bennema, E. Vlieg, Experimental and computational morphology of three polymorphs of the free base of Venlafaxine: A comparison of morphology prediction methods, Int. J. Pharm., 353 (2008) 113-123. [17] D. Kumar, N.R. Shastri, Designed Isomorphism of Nifedipine: A Joint Experimental and Molecular Simulation Study with Screened Solvents and Antisolvents, Cryst. Growth Des., 14 (2013) 326-338. [18] C. Aubrey Medendorp, S. Parkin, T. Li, The confusion of indexing aspirin crystals, J. Pharm. Sci., 97 (2008) 1361-1367. [19] G.L. Destri, A. Marrazzo, A. Rescifina, F. Punzo, Crystal morphologies and polymorphs in tolbutamide microcrystalline powder, J. Pharm. Sci., 102 (2013) 73-83. [20] U.S. Kestur, L.S. Taylor, Evaluation of the crystal growth rate of felodipine polymorphs in the presence and absence of additives as a function of temperature, Cryst. Growth Des., 13 (2013) 43494354. [21] R. Fossheim, Crystal structure of the dihydropyridine calcium antagonist felodipine. Dihydropyridine binding prerequisites assessed from crystallographic data, J. Med. Chem., 29 (1986) 305-307. [22] S. Srcic, J. Kerc, U. Urleb, I. Zupancic, G. Lahajnar, B. Kofler, J. Smid-Korbar, Investigation of felodipine polymorphism and its glassy state, Int. J. Pharm., 87 (1992) 1-10. [23] B. Lou, S.P. Velaga, Polymorph control of felodipine form II in an attempted cocrystallization, Cryst. Growth Des., 9 (2009) 1254-1257. [24] C. Schmidt, J. Ulrich, Predicting crystal morphology grown from solution, Chem. Eng. Technol., 35 (2012) 1009-1012. [25] J. Chen, B.L. Trout, Computer-aided solvent selection for improving the morphology of needlelike crystals: A case study of 2, 6-dihydroxybenzoic acid, Cryst. Growth Des., 10 (2010) 4379-4388. [26] J.J. Lu, J. Ulrich, An improved prediction model of morphological modifications of organic crystals induced by additives, Cryst. Res. Technol., 38 (2003) 63-73.

11

[27] Y.H. Kiang, C.-Y. Yang, R.J. Staples, J. Jona, Crystal structure, crystal morphology, and surface properties of an investigational drug, Int. J. Pharm., 368 (2009) 76-82. [28] A.O. Surov, K.A. Solanko, A.D. Bond, A. Bauer-Brandl, G.L. Perlovich, Polymorphism of felodipine co-crystals with 4, 4-bipyridine, CrystEngComm, 16 (2014) 6603-6611. [29] L. Yang, Y. Dong, Crystal morphology study of N, N'-diacetylchitobiose by molecular dynamics simulation, Carbohydr. Res., 346 (2011) 2457-2462. [30] T. Li, B. Li, M.S. Tomassone, Surface characterization of aspirin crystal planes using molecular dynamics simulations, Chem. Eng. Sci., 61 (2006) 5159-5169.

Fig. 1. Vacuum morphology of Polymorph I a. Crystal structure b. BFDH, c. Morphology growth, d. Equilibrium morphology

12

Fig. 2 Crystal slices of different facets of Fel polymorph I expressing the presence of polar and non-polar functional groups, a. {1 0 0} b. { 0 1 1} c. {1 1 0} d. {1 1 -1} e. {1 0 -2} f. {0 2 0}

Fig. 3. Vacuum morphology of Polymorph II a. Crystal structure b. BFDH, c. Morphology growth, d. Equilibrium morphology 13

Fig. 4 Crystal slices of different facets of Fel polymorph II expressing the presence of polar and non-polar functional groups, a. {1 1 0} b. {2 0 0} c. {1 1 -1}d. {1 1 0} e. {1 1 1} f. { 0 0 2}

Fig. 5. Vacuum morphology of Polymorph III a. Crystal structure b. BFDH, c. Morphology growth, d. Equilibrium morphology 14

Fig. 6 Crystal slices of different facets of Fel polymorph III expressing the presence of polar and non-polar functional groups, a. {1 0 -1} b. {1 0 1} c. {0 1 1} d. {1 1 0} e. {1 1 -1}

Fig. 7. Vacuum morphology of Polymorph IV a. Crystal structure b. BFDH, c. Morphology growth, d. Equilibrium morphology

15

Fig. 8. Crystal slices of different facets of Fel polymorph IV expressing the presence of polar and non-polar functional groups, a. {1 0 -1} b. {0 1 1} c. {1 1 0} d. {1 1 -1} e. {1 0 1}

Fig. 9 Comparisons of Fel polymorphs relative contributions to the Hirshfeld surface for intermolecular close contacts.

16

Table 1. Surface Chemistry of the Dominant Crystal Faces (hkl) of different Fel polymorphs Number of functional groups exposed to crystal surface Polymorph I hkl

CH3

CO/COO

Cl

Aromatic ring

{ 1 0 0}

2

2

{ 0 1 1}

1

{ 1 1 0}

5

2

1

{ 1 1 -1}

1

2

1

1

{ 1 0 -2}

2

1

{ 0 2 0}

2

2

1

1

Polymorph II { 1 1 0}

5

1

{ 2 0 0}

3

{ 1 1 -1}

3

2

2

{ 1 1 1}

3

1

3

{ 0 0 2}

1

1

1

2

1

Polymorph III { 1 0 -1}

2

1

{ 1 0 1}

3

{ 0 1 1}

3

{ 1 1 0}

4

2

2

{ 1 1 -1}

2

1

1

2 2

Polymorph IV { 1 0 -1}

2

{ 0 1 1}

3

{ 1 1 0}

4

{ 1 1 -1}

3

{ 1 0 1}

1 2

4

17

Table 2. Morphology predictions for polymorphs I, II, III & IV by means of BFDH, GM, and EM Calculations (TFA=total facet area) BFDH

GM

EM

Multiplicity

dhkl(Ǻ)

% TFA

Eatt (kcal mol−1)

% TFA

Esur (kcal mol−1 Ǻ-2)

% TFA

{ 1 0 0}

2

10.85

26.49

-26.16

32.95

0.080

17.69

{ 0 1 1}

4

8.53

37.48

-36.64

43.56

0.089

28.95

{ 1 1 0}

4

8.07

9.24

-42.84

3.11

{ 1 11}

4

8.03

21.69

-41.92

17.50

0.096

11.32

{ 1 02}

2

6.65

5.05

-52.87

2.72

0.104

6.10

{ 0 2 0}

2

6.03

0.02

-51.50

0.14

0.093

7.68

{ 2 11}

4

5.40

0.094

9.477

{ 1 1 0}

4

16.20

29.51

-262.25

23.76

{ 2 0 0}

2

16.19

22.53

-224.31

25.67

0.125

16.01

{ 1 11}

4

13.43

18.66

-307.51

5.53

0.142

4.64

{ 1 1 1}

4

13.32

17.07

-266.38

17.60

0.124

9.35

{ 0 0 2}

2

11.85

12.15

-218.23

27.42

0.091

23.51

{ 2 02}

2

9.64

0.05

{ 1 1 2}

4

9.52

0.112

12.49

hkl

Form I

Form II

18

2

9.35

0.135

4.77

{ 1 01}

2

13.17

40.97

-30.34

45.38

0.113

19.88

{ 1 0 1}

2

9.21

23.57

-46.77

30.00

0.125

14.19

{ 0 0 2}

2

8.10

6.28

{ 0 1 1}

4

6.60

17.27

-76.79

8.87

{ 1 1 0}

4

6.44

9.85

-80.25

1.11

{ 1 11}

4

6.33

2.03

-70.10

14.62

0.132

24.15

{ 0 1 2}

4

5.39

0.157

5.65

{ 2 1 0}

4

5.06

0.152

5.74

{ 1 01}

2

9.71

21.58

-41.01

24.66

0.110

12.39

{ 0 1 1}

4

9.00

44.13

-47.34

40.64

0.119

12.26

{ 1 1 0}

4

8.11

22.54

-56.04

12.97

{ 1 11}

4

7.68

3.70

-53.72

6.74

0.116

18.41

{ 1 0 1}

2

7.23

7.58

-53.16

12.02

0.108

9.15

{ 0 0 2}

2

6.45

0.45

-57.84

2.81

0.106

11.80

{ 0 2 0}

2

6.28

0.118

9.23

{ 1 1 1}

4

6.26

0.117

11.25

{ 0 2 0} Form III

Form IV

19

Table 3. Prediction of solubility and IDR of polymorph IV Polymorph

BFDH aspect ratio

Solubility (mol.l-1)

IDR (mg.min-1.cm-2)

I

1.81

0.0148

0.235

II

1.68

0.0156

0.247

III

2.2

0.0145

0.195

IV

1.53

0.0156

0.263

Morphology growth aspect ratio I

2.15

0.0148

0.235

II

1.55

0.0156

0.247

III

2.74

0.0145

0.195

IV

1.66

0.0154

0.246

Morphology growth polar/non polar ratio I

0.846

0.0148

0.235

II

1.133

0.0156

0.247

III

0.785

0.0145

0.195

IV

0.9

0.0148

0.223

Biography

Dinesh Kumar Dinesh Kumar is currently a PhD scholar under the supervision of Dr. Nalini Shastri at Department of pharmaceutics, NIPER, Hyderabad. He did his M. S. Pharm under the supervision of Dr Sanyog Jain at Department of Pharmaceutics, NIPER, Mohali, India. He received his B. Pharm degree from Maharshi Dayanand University Rohtak in year 2009. His 20

research interests include crystal habit modification and its simulations, solubility prediction, enhancement of oral bioavailability of anticancer drugs.

Dinesh Kumar

Rajesh Thipparaboina Rajesh Thipparaboina is currently a PhD scholar under the supervision of Dr. Nalini Shastri at Department of pharmaceutics, NIPER, Hyderabad. He did his M. S. Pharm at Department of Pharmaceutics, NIPER, Hyderabad. He received his B. Pharm degree from Kakatiya University in year 2010. His research interests include co-crystal generation and prediction, solubility prediction, enhancement of oral bioavailability of BCS class II & IV drugs.

Rajesh Thipparaboina

Dr. Nalini Shastri Her area of research interests are crystal habit modification, simulations, molecular modeling, polymorph prediction, solubility prediction, drug delivery to geriatric and pediatrics, QbD; Mesoporous and nanoporous materials, solid state characterization; Techniques to improve bioavailability, Micro and nanoemulsion technology.

Dr. Nalini Shastri

21

22