Langmuir 2006, 22, 6905-6909
6905
Wettability of Paracetamol Polymorphic Forms I and II Jerry Y. Y. Heng and Daryl R. Williams* Department of Chemical Engineering, Imperial College London, South Kensington Campus, London SW7 2AZ, United Kingdom ReceiVed March 4, 2006 It is well known that different forms of solid-state polymorphic materials exhibit diverse physicochemical properties. The variations in the wetting and surface energetics of a pair of organic polymorphic solids are reported in detail here for the first time. The growth of macroscopic single crystals (facet area >1 cm2) of paracetamol has enabled for the first time the direct measurement of advancing contact angles, θA for water and diiodomethane on a range of specific facets for two polymorphs; forms I and II. Not only was the wetting behavior found to be anisotropic, as has been recently reported, but the differing polymorphic forms exhibited significant variations in their wetting behavior for the same Miller indexed faces. The (001), (010), and (110) faces were studied, and the observed wettability data differed confirming the independence of facet wettability and Miller indices for both polymorphs. θA was found to be very sensitive to the local surface chemistry for each facet examined, which in turn is a direct consequence of the molecular packing and structure within the crystal lattice. On the basis of the θA value of water, the hydrophilicity rankings for the facet surfaces of form II examined is: (010) ≈ (110) > (001). This experimental study highlights complex surface chemistry of polymorphic solids in which anisotropic surface energies were observed for both forms of paracetamol, strongly suggesting that such anisotropic wetting behavior is the norm for organic crystalline solids. Furthermore, the same Miller indexed facets for forms I and II exhibited very different surface chemical behavior, such that it was not possible to infer understanding about one form based upon knowledge of another form.
Introduction Polymorphism is common in virtually all classes of solids. Molecules which are chemically identical can be packed in different crystal lattices, with differing spacings, angles, or even crystal systems, and such systems are known as polymorphs. Many crystalline materials exhibit polymorphism in which the same molecule may form crystal structures with different internal cell dimensions and crystal lattices such as TiO2; rutile, brookite, and anastase. In the case of crystalline pharmaceutical solids, polymorphism is very common1 with many compounds exhibiting more than 10 polymorphic forms, especially for barbiturates, sulfonamides, and steroids.2 Crystallization conditions such as degree of supersaturation, solvent, and temperature may influence the morphology of the resulting crystal structures and are common processing variables used to facilitate the preparation and identification of new polymorphic forms. Often the real challenge for pharmaceutical researchers is identifying all of the possible polymorphic forms of a compound, with a critical interest in determination of the most thermodynamically stable form. The difficulty in such an objective is illustrated by the recent discover of the second form for aspirin, over 150 years following the first synthesis.3 Recent years has shown significant progress in the use of computation based approaches for predicting polymorph stability.4,5 Polymorphs may exhibit enantiotropic behavior, where the order of thermodynamic stability depends on temperature. Such enantiotropic behavior is exhibited by caffeine, which is known * To whom correspondence should be addressed. Phone, +44 (0) 207 594 5611; e-mail,
[email protected]. (1) Bernstein, J. Polymorphism in Molecular Crystals; Oxford University Press: Oxford, 2002. (2) Byrn, S. R.; Pfeiffer, R. R.; Stowell, J. G. Solid-State Chemistry of Drugs, 2nd ed.; SSCI Inc.: Indiana, 1999. (3) Vishweshwar, P.; McMahon, J. A.; Oliveira, M.; Peterson, M. L.; Zaworotko, M. J. Am. Chem. Soc. 2005, 127, 16802-16803. (4) Price, S. L. AdV. Drug Del. ReV. 2004, 56, 301-319. (5) Karamertzanis, P. G.; Pantelides, C. C. J. Comput. Chem. 2005, 26, 304324.
to form two anhydrous polymorphs, a β form stable at 25 °C with a transition in stability to the R form at 141 °C.6 If the stability of polymorphs is not interchangeable, the polymorphs are termed monotropic. In this work paracetamol is studied, which is reported to have a metastable polymorphic form. The thermodynamically stable form of paracetamol is the monoclinic form, while the orthorhombic form is the metastable form at ambient conditions. Previous experimental results suggesting wide ranging possible transition temperatures for paracetamol from -40 to -120 °C.7 Despite the prevalence and complexity of polymorphism, very little detailed and systematic understanding of their physical chemistry exists. This work reports the first systematic investigation of the face specific hydrophilic nature of paracetamol form I vs form II. The experimental measurement of the wetting behavior of particulate systems faces numerous experimental and theoretical limitations as has been discussed elsewhere.8,9 It has been proposed that these conventional experimental techniques should be abandoned in favor of characterization of large single crystals.10 Wetting studies with probe liquids are often employed to evaluate the surface properties of solids by measuring the liquid contact angle,11 described by Young’s equation12 as shown in eq 1
γSV° ) γSL + γLV cos θY
(1)
where γSV° is the solid surface energy, γLV the liquid vapor surface tension, γSL the solid-liquid surface energy, and θY the (6) Epple, M.; Cammenga, H. K.; Sarge, S. M.; Diedrich, R.; Balek, V. Thermochim. Acta 1995, 250, 29-39. (7) Sacchetti, M. J. Therm. Anal. Calorim. 2001, 63, 345-350. (8) Heng, J. Y. Y. Anisotropic Surface Properties of Crystalline Pharmaceutical Solids. Ph.D. Thesis, University of London, London, 2006. (9) Heng, J. Y. Y.; Bismarck, A.; Williams, D. R. J. Pharm. Sci. In preparation. (10) Duncan-Hewitt, W.; Nisman, R. In Contact Angle, Wettability and Adhesion; Mittal, K. L., Ed.; VSP: Utrecht, The Netherlands, 1993; pp 791-811. (11) Good, R. J. In Surface and Colloid Science; Good, R. J., Stromberg, R. R., Eds.; Plenum Press: New York, 1979; pp 1-29. (12) Young, T. Philos. Trans. R. Soc. London 1805, 95, 65-87.
10.1021/la060596p CCC: $33.50 © 2006 American Chemical Society Published on Web 06/30/2006
6906 Langmuir, Vol. 22, No. 16, 2006
Heng and Williams
Table 1. Properties of Liquid Probes Used liquids
purity (%)
density (kg/m3)
γLV (mJ/m2)
γLd (mJ/m2)
γLp (mJ/m2)
diiodomethane water ethylene glycol formamide
>99 deionized >99 >99.5
3325 998 1109 1139
50.8 72.8 48.0 58.0
50.8 21.8 29.0 39.0
0.0 51.0 19.0 19.0
Young contact angle. Young’s equation was originally derived based on a mechanical equilibrium for forces acting at the three phase contact line. It is common practice for liquids which exhibit finite θA > 10° to assume that the equilibrium spreading pressure is negligible (πe = 0).11 Since only θY and γLV can be measured, the evaluation of surface energy and their components, the use of semiempirical approaches such as Fowkes,13 Owens-Wendt,14 Wu15 and Fowkes and Mostafa,16 van Oss et al.,17 and Chen-Chang18 acid-base analyses is common. These models have been used to evaluate the surface free energy components of many solid materials but most commonly polymer surfaces. All of these models are based on the assumption that the surface free energy of a solid or a liquid can be treated as consisting of a number independent, or partially independent components, each of which represents a distinctly different type of intermolecular interaction. Initially, Fowkes proposed that a solid’s surface free energy γS could be considered to be made up of various components such as the dispersive, dipole, induction, metallic, and hydrogen bonds. Owens-Wendt proposed a model, also known as the “extended Fowkes’ approach”, which consists of a dispersive and a polar component such that
γs ) γSd + γSp
(2)
where γSd is the solid surface energy dispersive component while γSp is the solid surface energy polar component. The surface energy of paracetamol forms I and II were evaluated by this approach. The theoretical basis of these analyses and other semiempirical relations for determination of surface energies has been recently review in detail.19 Experimental Section Materials. Paracetamol (4-acetamidophenol) (98%, SigmaAldrich) and methyl alcohol (>99.9%, Sigma-Aldrich) were used for the preparation of paracetamol crystals without further purification. Deionized water, diiodomethane (>99%, Acros Organics), formamide (>99.5%, Acros Organics), and ethylene glycol (>99%, SigmaAldrich) were used as probe liquids for the contact angle measurements. Liquid properties are shown in Table 1. Diiodomethane was purified by passing through chromatographic columns packed with silica (Merck) and basic alumina (Sigma-Aldrich). Macroscopic Crystal Preparation. Macroscopic crystals were grown which allowed the direct determination of the facet specific wettability of paracetamol forms I and II. Thermodynamically stable monoclinic (form I) paracetamol crystals were grown by the slow evaporation of a saturated methanol solution, and this approach is described in detail elsewhere.20 For growing macroscopic form II (13) Fowkes, F. M. Ind. Eng. Chem 1964, 56, 40-52. (14) Owens, D. K.; Wendt, R. C. J. Appl. Polym. Sci. 1969, 13, 1741-1747. (15) Wu, S. Polymer Interface and Adhesion; Marcel Dekker: New York, 1982. (16) Fowkes, F. M.; Mostafa, M. A. Ind. Eng. Chem. Prod. Res. DeV. 1978, 17, 3-7. (17) van Oss, C. J.; Good, R. J.; Chaudhury, M. K. Langmuir 1988, 4, 884891. (18) Chen, F.; Chang, W. V. Langmuir 1991, 7, 2401-2404. (19) Etzler, F. M. In Contact Angle, Wettability and Adhesion; Mittal, K. L., Ed.; VSP: Ultrecht, The Netherlands, 2003; pp 1-46. (20) Heng, J. Y. Y.; Bismarck, A.; Lee, A. F.; Wilson, K.; Williams, D. R. Langmuir 2006, 22, 2760-2769.
Figure 1. Crystallographic structure of the monoclinic (form I) as viewed along the a and c axes, respectively. (orthorhombic, thermodynamically metastable) paracetamol crystals, a controlled temperature drop approach was employed. The description of a method for the growth of macroscopic (>1 cm) sized metastable form II single crystals was first reported by Mikhailenko.21 Form II crystals could not be grown by seeding a saturated solution with form II seeds. Conversion of the form II seeds back to form I is favorable in solution, resulting in a complete conversion within 6 h, even at 0 °C.22 A mixture of 25 g of paracetamol in 1 L of deionized water was prepared at 40 °C. The mixture was then heated to its boiling point to ensure that all powder was fully dissolved. This solution was continually boiled for 15 min and then filtered. The solution was then placed in a water bath at 40 °C for 24 h before cooling at a rate of 3 °C/day until typically 0-4 °C. Form II crystals were formed at the bottom of the flask, with a reported 95% probability (21) Mikhailenko, M. A. J. Cryst. Growth 2004, 265, 616-618. (22) Nichols, G.; Frampton, C. S. J. Pharm. Sci. 1998, 87, 684-693.
Anisotropic Wettability of Polymorphic Paracetamol
Langmuir, Vol. 22, No. 16, 2006 6907
Figure 3. Paracetamol (a) form II single crystals (scale in cm) and (b) its habit.
Figure 2. Crystallographic structure of the orthorhombic (form II) as viewed along the c and a axes, respectively. within 24 days.21 In this research, it has been observed that crystallization of form II can be expedited by directly cooling the solution from 40 to 4 °C, in which case crystals are typically formed within 48 h. The formation of macroscopic form II crystals was very sensitive to the presence of any contamination within the crystallization system. If the supersaturated solutions became contaminated, then form I recrystallization was favorable. It was observed that these contaminants provided sites for the thermodynamically stable form I to crystallize. Therefore, all of the glassware used was initially washed with alcohol solvents and rinsed thoroughly with deionized water. Furthermore, the flasks were sealed with Parafilm to ensure that no contaminant particles could enter the vessel. Crystals obtained after crystallization period of 3-4 weeks had facet areas >1 cm2. The crystallographic structure of both forms were resolved by Haisa et al.23,24 and were obtained from the Cambridge Structural Database (HXACAN01 and HXACAN).25 The crystallographic structures for forms I and II are shown in Figures
1 and 2, respectively. The crystalline habit of form I was prismatic with external facets of (201), (001), (011), and (110). An internal facet (010) was revealed upon cleaving of the crystal. The habit of form II was rod like and elongated along the c axis, formed by facets (110), (010), and (001) as is shown in Figure 3. Contact Angle Measurements. Sessile drop contact angles were obtained with a Kru¨ss Drop Shape Analyzer (DSA 10, Kru¨ss GmbH, Hamburg, Germany). Initial drops of about 5 µL were dispensed onto the solid surface. By use of a motor-driven syringe, the test liquid was added onto the droplet allowing θA to be obtained. The needle remained immersed within the top half of the droplet and θA were determined with a slowly advancing wetting line. The droplet was monitored with a CCD camera and analyzed by the Drop Shape Analysis software (DSA version 1.0, Kru¨ss). The droplet contour is fitted by the tangent method (function f(x) ) a + bx + cx0.5 + d/ln x + e/x2) at both left and right sides of the drop. The function is differentiated and slope at the three phase contact point gives the contact angle. At least 8 (typically between 8 and 20) contact angle measurements were taken for each liquid on each facet, obtained from numerous single crystals. Measurements were obtained in the open air at ambient conditions (T ) 20 ( 2 °C), and probe liquid properties are shown in Table 1.
Results and Discussion Contact Angle Analysis. The advancing contact angles, θA, for water were measured on all available facets of forms I and II paracetamol crystals. θA could be potentially affected either by surface roughness of the crystal facets or by the potential dissolution of the crystal surface by the wetting liquid. The authors have previously established that neither of these factors are significant for the observed anisotropic wettability of macroscopic form I paracetamol crystals.20 A minimum of 20 droplets on more than 5 single crystals were determined as reported in Tables 2 and 3. θA varied significantly, (23) Haisa, M.; Kashino, S.; Kawai, R.; Maeda, H. Acta Crystallogr. 1976, B32, 1283-1285. (24) Haisa, M.; Kashino, S.; Maeda, H. Acta Crystallogr. 1974, B30, 25102512. (25) Fletcher, D. A.; McMeeking, R. F.; Parkin, D. J. Chem. Info. Comput. Sci. 1996, 36, 746-749.
6908 Langmuir, Vol. 22, No. 16, 2006
Heng and Williams
Table 2. θA for Water on Crystalline Facets of Paracetamol advancing contact angle θA (deg) facets
form I
(201) (001) (011) (110) (010)
38.1 ( 4.6 15.9 ( 3.1 29.8 ( 5.7 50.8 ( 4.9 67.7 ( 2.5
form II 64.5 ( 3.5 16.6 ( 1.4 17.9 ( 2.5
Table 3. θA for Diiodomethane on Crystalline Facets of Paracetamol advancing contact angle θA (deg) facets
form I
(201) (001) (011) (110) (010)
48.8 ( 2.2 49.8 ( 3.2 50.7 ( 2.9 50.2 ( 2.4 27.8 ( 2.5
form II 34.5 ( 1.2 51.6 ( 3.2 53.6 ( 2.3
especially for the probe liquid water, for all facets of form I and form II paracetamol crystals. θA for water varied from 15.9 to 67.7° on facets (001) and (010) of form I, respectively, and for form II θA varied from 16.6 to 64.5° for facets (110) and (001), respectively. These data is direct evidence for the anisotropic wetting behavior for crystalline paracetamol polymorph forms I and II. Polymorphism results in a material having different physicochemical22 and thermodynamic properties26 such as the melting temperature, densities, refractive indices, spectra properties,27 crystal habits, processibility,28 and solubilities. The work by Muster and Prestidge,29 though interesting, was significantly limited by the needlelike shape of sulfathiozole form I crystals formed, which were too small for sessile drop experiments on individual facets to be performed. Hence, no comparison between the wetting behavior of polymorphs examined were available. The classical explanation for variations in θA are differences in the local surface chemistry; both types and densities of surface groups. Individual crystallographic facets will intersect through the crystal lattice at different planes and angles. Consequently, the chemical groups exposed on each crystal facet could reasonably be expected to vary. Therefore, the anisotropic wetting observed here by water is fully consistent with this analysis of facet dependent local surface chemistry. Other types of anisotropic surface phenomena for forms I and II paracetamol crystals as well as solvent etch patterns30 have been reported. These crystallographic structures are shown in Figures 1 and 2. For form I, sOH groups accepts hydrogen bonds from sNH and donates to CdO of an adjacent molecule, forming corrugated sheets along the ac plane, held by weak van der Waals forces. For facet (001), sOH groups are clearly exposed, with near normal orientation to the surface (Figure 1), resulting in the low θA reported. For facet (010), sOH groups of the paracetamol molecule appears to be obscured by an adjacent sNH/CdO groups which form multiple hydrogen bonds. There is consequently negligible hydrogen bonding lability on facet (010), which is consistent with our experimental observation of the hydrophobic nature of this facet. Detailed surface analysis of each (26) Espeau, P.; Ceolin, R.; Tamarit, J.-L.; Perrin, M.-A.; Gauchi, J.-P.; Leveiller, F. J. Pharm. Sci. 2005, 94, 524-539. (27) Ivanova, B. B. J. Mol. Struct. 2005, 738, 233-238. (28) DiMartino, P.; Guyot-Hermann, A.-M.; Conflant, P.; Drache, M.; Guyot, J.-C. Int. J. Pharm. 1996, 128, 1-8. (29) Muster, T. H.; Prestidge, C. A. J. Pharm. Sci. 2002, 91, 1432-1444. (30) Mikhailenko, M. A.; Drebushchak, T. N.; Shakhtsneider, T. P.; Boldyrev, V. V. Arkivoc 2004, xii, 156-169.
Table 4. Results of a Surface Energy Analysis for Form I Paracetamol Using the Owens-Wendt Analysis facet
γSd (mJ/m2)
γSp (mJ/m2)
γ (mJ/m2)
γp/γ
(201) (001) (011) (110) (010)
34.9 34.4 33.9 34.2 45.1
27.5 38.0 32.7 20.2 7.0
62.4 72.4 66.5 54.4 52.1
0.44 0.53 0.49 0.37 0.13
individual facets and resultant quantitative determination of functional groups by XPS reported elsewhere support this interpretation.20 The same hydrogen bond pairings exist for form II but are formed between two adjacent molecules, resulting in 2-D sheets parallel to the ab plane. At (001), the paracetamol molecule lie parallel to the ab plane, with the -OH, -NH, and CdO functional groups present. However, all of these groups are fully engaged in intraplanar hydrogen bonding; no groups, especially -OH, are free to engage with the external environment via the surface. This interpretation agrees very well with contact angle observations, which reveals (001) for form II as being the most hydrophobic facet. Meanwhile for facet (010), these groups do not completely form hydrogen bonds and thus some -OH groups are free to interact with external molecules. This explanation accounts for the observed hydrophilicity of this facet. Only slight differences in wetting behavior were seen for facet (110) as compared to facet (010) due to the similar orientation of molecules and functional groups present, resulting in similar hydrogen bonding potential. From the θA measurements with water, the orders of hydrophilicity for the specific facets of forms I and II are
(001) > (011) > (201) > (110) > (010) (Form I)20 (010) ≈ (110) > (001) (Form II) The results reported here illustrate the complexity of the surface chemical behavior of differing polymorphic materials. Specifically, the wetting behavior of polymorphic crystalline solids is unrelated to their Miller indices. For forms I and II of paracetamol, facets (001) and (010) are the most hydrophilic respectively, whereas facets (010) and (001) are the most hydrophobic, respectively. That is, the hydrophilicity trends as given by these lists are opposite for form I vs form II. The wetting literature for organic solids for the past 50 years has been dominated by studies on polymer-based materials.31 No reported studies contradict the general observation that organic solids exhibit isotropic wetting behavior, despite the intuitive expectation that this should be false for crystalline solids. Our work shows that expecting isotropic behavior for crystalline solids is erroneous, and this behavior could well explain the wide variation in surface properties reported for pharmaceutical crystalline solids.32 Calculation of Solid Surface Energy. The contact angle data has been analyzed using the classic Owens-Wendt14 polar approach for both forms I and II paracetamol crystals (Tables 4 and 5). Surface energies calculated from the θA with the component approaches for form I paracetamol crystals are reported in detail elsewhere.20 For form II paracetamol crystals, γSd for facet (001) of 42.3 mJ/m2, the weakest attachment energy plane, was also found to be about 25% higher than the other facets. The other two (31) Wu, S. Polymer Interface and Adhesion; Marcel Dekker: New York, 1982. (32) Buckton, G. Powder Technol. 1990, 61, 237-249.
Anisotropic Wettability of Polymorphic Paracetamol
Langmuir, Vol. 22, No. 16, 2006 6909
Conclusions
Table 5. Results of a Surface Energy Analysis for Form II Paracetamol Using the Owens-Wendt Analysis facet
γSd (mJ/m2)
γSp (mJ/m2)
γ (mJ/m2)
γp/γ
(110) (010) (001)
33.4 32.2 42.3
38.5 38.9 9.5
71.9 71.1 51.8
0.54 0.55 0.18
faces, (110) and (010), for form II gave γSd of 32.8 ( 0.6 mJ/m2, which is similar to the values found for the external facets of form I, 34.4 ( 0.4 mJ/m2. For the case of form I paracetamol crystals, the facet (010) was an internal facet not available in the native crystal. It may be argued that the lower values of γSd for form I external facets were due to some type of surface contamination. However, the same trend is reported here for form II. Thus, it is concluded that this observation is a real intrinsic property of the material. It is therefore suggested that a higher density of methyl groups may be responsible for this behavior, but this analysis requires further work to confirm this viewpoint. The work reported here not only illustrates that the complex physical chemistry of polymorphic solids extends to their surface chemical properties but also allows us to propose that all organic crystalline solids may well exhibit anisotropic face dependent surface chemical properties.
In summary, this study reports for the first time, the facet specific θA for water and diiodomethane on two polymorphs of paracetamol, forms I and II. A hydrophilicity ranking is derived for all facets based on the anisotropic wetting behavior with water observed. The variations in θA are argued to be a direct consequence of the molecular packing and structure within the crystal lattice. For the crystal lattice, the orientation of the molecules, the hydrogen bonding lability and proportions of functional groups present at each facet, in total, define the local surface chemistry for a facet. This experimental study highlights complex surface chemistry of polymorphic solids in which anisotropic surface energies were observed for both forms of paracetamol. Furthermore, the same Miller indexed facets for forms I and II exhibited very different surface chemical behavior, such that it was not possible to infer understanding about one based upon knowledge of the other. Acknowledgment. The authors would like to acknowledge the use of the EPRSC’s Chemical Database Service at Daresbury. J.Y.Y.H. acknowledges the financial support of an ORS scholarship. LA060596P
