- Email: [email protected]
Influence of pH and method of crystallization on the solid physical form of indomethacin
International Journal of Pharmaceutics 473 (2014) 536–544
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Influence of pH and method of crystallization on the solid physical form of indomethacin Alessandra Dubbini a , Roberta Censi a , Valentina Martena a , Ela Hoti b , Massimo Ricciutelli a , Ledjan Malaj b , Piera Di Martino a, * a b
University of Camerino, School of Pharmacy, Via S. Agostino, Camerino, Italy University of Medicine, Tirana, Faculty of Pharmacy, Street of Dibra, Tirana, Albania
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 April 2014 Received in revised form 21 July 2014 Accepted 22 July 2014 Available online 23 July 2014
The purpose of this study was to investigate the effect of pH and method of crystallization on the solid physical form of indomethacin (IDM). IDM, a non steroidal anti-inflammatory drug poorly soluble in water, underwent two different crystallization methods: crystallization by solvent evaporation under reduced pressure at 50.0 C (method A), and crystallization by cooling of solution from 50.0 to 5.0 C (method B). In both cases, several aqueous ethanolic solutions of IDM of different pHs were prepared. pHs were adjusted by adding acidic solutions (HCl 2 M) or alkali (NaOH or NH4OH 2 M) to an aqueous ethanolic solution of IDM. Thus, several batches were recovered after crystallization. The chemical stability of IDM was verified through 1H NMR and mass spectroscopy (FIA-ESI-MS), that revealed that IDM degraded in strong alkali media (pH 12). Crystals obtained under different crystallization conditions at pHs of 1.0, 4.5, 7.0, 8.0, 10.0 and chemically stable were thus characterized for crystal habit by scanning electron microscopy, for thermal behaviour by differential scanning calorimetry, and thermogravimetry, and for solid state by X-ray powder diffractometry. Under the Method A, IDM always crystallized into pure metastable alpha form when solutions were acidified or alkalized respectively with HCl and NH4OH. On the contrary, in presence of NaOH, IDM crystallized under a mixture of alpha and sodium trihydrate form, because the presence of the sodium counter ion orientates the crystallization towards the formation of the trihydrate salt. Under the method B, at pH of 1.0, IDM crystallized under the alpha form; at pH 4.5, IDM crystallized under the form alpha in presence of some nuclei of gamma form; at pH 7.0, 8.0, and 10.0 for NH4OH, IDM crystallized under the most stable polymorph gamma form, whereas in presence of NaOH, a mix of alpha, and salt forms was formed whatever the pH of the solution. ã 2014 Elsevier B.V. All rights reserved.
Keywords: Indomethacin Polymorphism pH Crystallization
1. Introduction Knowledge of solid state transitions occurring in pure drugs during processing or storage is fundamental for developing stable pharmaceutical formulations (Yu et al., 1998). Polymorphic transitions, hydration–dehydration processes, amorphization, and salt formation occurring intentionally or unintentionally during drug processing must be taken into account, because different solid forms lead to different physical properties (solubility and dissolution rate, compression ability and attitude to grinding, long term physical stability, etc.) (Yoshinari et al., 2002; Nair et al., 2002; Vippagunta et al., 2001; Zhang et al., 2004; Yoshinari et al., 2003; Martínez-Ohárriz et al., 2002; Singhal and Curatolo, 2004; Yu, 2001).
* Corresponding author. Tel.: +39 737 402215; fax: +39 737 637345. E-mail address: [email protected] (P. Di Martino). http://dx.doi.org/10.1016/j.ijpharm.2014.07.030 0378-5173/ ã 2014 Elsevier B.V. All rights reserved.
Molecules containing ionizable groups (carboxyl, amino, etc.) can crystallize in neutral form or as salts with their counter ions. It was demonstrated that it is possible to crystallize different polymorphs according to the solution pH. In aqueous solution, a sodium salified form of sulindac exists in equilibrium between two different polymorphic forms; this fact allows to isolate in the solid state the two different polymorphic forms I and II according to the pH of the crystallization solution (Llinàs et al., 2007). Yu and Ng (2002) investigated the effect of pH on glycine crystallization during spray drying. This study revealed the crystallization of glycine on different polymorphs and salt forms according to the crystallization pH. The authors of the present study intend to consider in dept the possibility to induce a change in the solid physical form of a ionizable drug by changing crystallization conditions and solution pH. In particular, the effect of pH and method of crystallization on the solid state of indomethacin will be investigated. Indomethacin (IDM) (the molecular structure is given in Fig. 1), a non steroidal anti-inflammatory drug (NSAID), is a poorly water
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
O
2' 14
7 6
9
5
8
O
4
4'
Cl 2.2. Preparation of IDM solution at different pHs
N1 2
H3C
Crystals were stored in a desiccator in presence of P2O5 as desiccant during the experiment time interval.
3'
1'
CH3
3 10
OH
11
537
O Fig. 1. Molecular structure of indomethacin.
soluble active substance with poor bioavailability after oral administration. IDM exists in four different polymorphic forms: the g form (also named form I) (Tm: 157.9 C) that is the most stable thermodynamic form (triclinic); the a form (also named form II) (Tm: 151.4 C) (monoclinic); the b form (also named III) (Borka, 1974; Yamamoto, 1968; Allen and Kwan, 1969; Monkhouse and Lach, 1972; Legendre and Feutelais, 2004), and the d form (also named form IV) (Tm: 130.6 C) (Crowley and Zografi, 2002). A crystalline IDM sodium trihydrate has also been described (Tong and Zografi, 1999; Tong and Zografi, 2004; Kulkarni et al., 2011). During experiments devoted to describe the crystallization behaviour of quench cooled amorphous IDM in media of different pHs and temperatures, new polymorphs were detected, named e, z, and h (Surwase et al., 2013). In the present study, two different crystallization methods were used: 1. Method A: crystallization by solvent evaporation under
reduced pressure at 50 C. 2. Method B: crystallization by cooling of solution from 50.0 to
5.0 C.
The aqueous ethanolic solutions of indomethacin were prepared by adding an accurately weighed amount of IDM to a solution of water and ethanol (1:1). The initial pH was measured (Jenway 3510 pHmeter, Staffordshire, UK). Then, the solution was charged with HCl, NaOH, or NH4OH 2 M up to the desired pH (1.0, 7.0, 8.0 10.0, 12.0). Only for solutions at pH 4.5, neither acid or alkali were added exploiting the fact that this pH value, that corresponds to the pKa of indomethacin, may be easily reached by adjusting the amount of IDM added to the aqueous ethanolic solutions. The actual pH was again measured. Table 1 gives details about the preparation conditions of IDM solutions. Results are the mean of three different measurements. 2.3. Indomethacin equilibrium solubility in solutions of different pH The evaluation of the equilibrium solubility was performed with the only scope to obtain saturated solutions at 25 C at different pHs. An excess of g IDM (native crystals) was added to 50 ml of the aqueous ethanolic solution of the required pH and maintained at 25 0.5 C under continuous stirring in an incubator (Velp Scientifica, FTC 90E, Usmate, Italy). The required pH was adjusted either with HCl 1 N, NH4OH, or NH4OH 2 M. Equilibrium between the undissolved drug and the drug in solution was tested and judged to have been reached when three consecutive spectrophotometric measurements differed by no more than 1%, generally after 24 h. The equilibrium solubility was thus determined by withdraws aliquots, centrifuged at 5000 rpm at 25 0.5 C for 10 min (High Speed Micro-Centrifuge Scilogex D3024R, Berlin, CT, USA); after appropriate dilution, the concentration of the solution was determined by UV spectrophotometry at a wavelength of 318 nm (Cary 1E UV–VIS, Varian, Leinì, Italy). Assays were performed in triplicate.
For each method, different aqueous ethanolic solutions of different pHs (1.0, 4.5, 7.0, 8.0, 10.0, and 12.0) were tested to crystallize IDM.
2.4. Crystallization of indomethacin through different methods
2. Experimental
Taking into account the solubility measurements, and after the preparation of solutions at different pHs, IDM crystals were recovered by using two different crystallization methods:
2.1. Materials Indomethacin (Native Crystals: NCs) (g form) was kindly supplied by Fabbrica Italiana Sintetici (F.I.S., Vicenza, Italy) as white crystalline powder. Ultrapure water was produced by Gradient Milli-Q1 (Millipore, Molsheim, France). Chemicals, all of analytical grade, were supplied by Sigma–Aldrich (Steinheim, Germany).
2.4.1. Method A: crystallization by solvent evaporation under reduced pressure at 50 C The saturated solution prepared at 25 0.5 C was heated up to 50 0.5 C in order to completely dissolve IDM. The solvent was thus totally evaporated under reduced pressure at a temperature of 50 C (Rotavapor1 R-210, Buchi, Flawil, Switzerland; vacuum
Table 1 Aqueous ethanolic solutions of indomethacin obtained after addition of acids or alkali to reach solutions of the desired pH. pH required
Ethanol + Water (g)
IDM (mg)
1.0 4.5 (pKa) 7.0 8.0 10.0 12.0 7.0 8.0 10.0 12.0
25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0
50.730 50.400 50.695 50.800 50.400 51.225 50.425 50.255 50.265 50.255
+ + + + + + + + + +
25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0
Starting pH of ethanol/water/IDM solution 0.028 0.607 0.714 0.567 0.607 0.035 0.318 0.0778 0.0636 0.0778
4.323 4.457 4.309 4.410 4.457 4.248 4.426 4.430 4.453 4.431
0.052 0.060 0.255 0.050 0.060 0.169 0.043 0.014 0.0184 0.012
Added volume of acid or alkali (ml)
Final pH
4400.0 (HCl 2 M) – 60.0 (NaOH 2 M) 67.0 (NaOH 2 M) 73.0 (NaOH 2 M) 90.0 (NaOH 2 M) 48.0 (NH4OH 2 M) 59.0 (NH4OH 2 M) 352.0 (NH4OH 2 M) 47680.0 (NH4OH 2 M)
1.062 4.457 7.060 8.095 10.948 12.019 7.071 8.184 10.063 11.947
0.016 0.060 0.094 0.071 0.080 0.013 0.153 0.072 0.081 0.160
538
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
pump, V-710, Buchi, vacuum level 680 mmHg). The residual solvent was then eliminated by placing crystals in a desiccator in presence of P2O5 as desiccant. Crystals were stored in these conditions for the entire duration of the experiments. 2.4.2. Method B: crystallization by cooling of solution The saturated solution prepared at 25 0.5 C was heated at 50 0.5 C until the complete dissolution of the drug. The solution was then quench cooled under stirring at 5.0 1.0 C by simply placing the solution container in a glass bath and this temperature was maintained until the crystallization of the IDM (approximately for 24–48 h). Crystals were filtered out and maintained in a desiccator in presence of P2O5 as desiccant. Crystals were stored in these conditions for the entire duration of the experiments. 2.5. Preparation of indomethacin a and amorphous forms The a form of IDM was prepared by a modification of the method of Borka (Borka 1974; Kaneniwa et al., 1985). 10 g of IDM bulk powder were dissolved in 10 ml of ethanol at 80 0.5 C; the undissolved drug was filtered off, and 20 ml of distilled water at room temperature was added to the IDM-saturated ethanol solution at 80 0.5 C. The precipitated crystals were removed by filtration, using a glass funnel, and then dried overnight in a P2O5 desiccator under vacuum at room temperature. The X-ray powder diffraction (XRPD) patterns of this sample were compared to those described in literature (Kaneniwa et al., 1985) to confirm the obtainment of a pure a form. Amorphous IDM (Andronis and Zografi, 2000) was obtained by cooling in liquid nitrogen after melting the bulk powder at 165 C for 5 min. The crystallinity of this material, assessed by X-ray powder diffractometry, was assumed to be 0%. IDM a and amorphous forms were stored in a desiccator in presence of P2O5 as desiccant during the experiment time interval. 2.6. Preparation of the indomethacin sodium trihydrate The IDM sodium trihydrate was prepared according to Kulkarni et al. (2011). An appropriate amount of IDM was added to 50 ml of methanol and maintained at 50 0.5 C until the complete dissolution of the drug. The solution was left to cool spontaneously until room temperature. An appropriate amount of NaOH 2 M was then added to the solution and kept under stirring for 2 h. The solution was then evaporated under reduced pressure at 35 0.5 C. The solid was desiccated in presence of P2O5 as desiccator for 24 h. Powder is then kept in well closed vials. X-ray powder diffraction patterns of this sample were compared to those described in literature (Tong and Zografi, 2004) to confirm the obtainment of a pure trihydrate sodium indomethacin.
2.8. Evaluation of the chemical stability by 1H NMR The chemical stability of IDM samples was also evaluated through 1H NMR (Varian, Mercury Console 400, Palo Alto, USA). Sample were analysed in deuterated methanol at different pHs. 2.9. Thermogravimetric analysis Thermogravimetric analysis was used to determine the powder water content and it was carried out by the simultaneous thermal analysis (STA) which enables to simultaneously analyze a sample for change in weight (Thermogravimetric analysis, TGA) and change in enthalpy flow (Differential Scanning Calorimetry, DSC). In this article, the thermogravimetric analysis is referred to as STA–TGA. The analysis was performed using a Simultaneous Thermal Analyser (STA 6000, PerkinElmer, Inc., Waltham, MA, USA), under-nitrogen atmosphere (20 ml/min) in 0.07 ml open aluminium oxide pans. Samples were heated from room temperature to 100 C at a heating rate of 5.0 C/min. STA was calibrated for temperature and heat flow with three standard metals (indium and tin), taking into account their expected melting temperatures (156.75 and 232.08 C respectively), and for weight with an external PerkinElmer standard (Calibration Reference Weight P/N N520-0042, Material lot 91101GB, Weight 55.98 mg, 01/23/08 VT). Calibration was repeatedly checked to assure deviation 0.3 K. 2.10. Differential scanning calorimetry analysis Differential scanning calorimetry (DSC) analysis was performed on a Pyris 1 (PerkinElmer, Co., Norwalk, USA) equipped with a cooling device (Intracooler 2P, Cooling Accessory, PerkinElmer, Co., Norwalk, USA). A dry purge of nitrogen gas (20 ml/min) was used for all runs. DSC was calibrated for temperature and heat flow using a pure sample of indium and zinc standards. Sample mass was about 4–5 mg and aluminium perforated pans of 50 ml were used. Samples were heated from room temperature to 200 C or 300 C at different heating rates 1.0, 5.0, 10.0, 20.0 and 40.0 C min 1. 2.11. X-ray powder diffractometry X-ray powder diffractometry (XRPD) was carried out on a Philips PW 1730 (Philips Electronic Instruments Corp., Mahwah, NJ, USA) as X-ray generator for Cu Ka radiation (la1= 1.54056 Å, la2 = 1.54430 Å) was used. The experimental X-ray powder patterns were recorded on a Philips PH 8203. The goniometer supplied was a Philips PW 1373 and the channel control was a Philips PW 1390. Data were collected in the discontinuous scan mode using a step size of 0.01 2u. The scanned range was 2 –40 (u). 2.12. Scanning electron microscopy
2.7. Evaluation of the chemical stability by FIA-ESI-MS The evaluation of the chemical stability of IDM during processing was evaluated by FIA-ESI-MS analysis (Flow Injection Analysis-Electrospray Ionization-Mass Spectrometry) (purity test) performed by an HPLC 1090Hewlett Packard Series II (Agilent, Santa Clara, CA, USA) equipped with a Hewlett Packard Mass spectrometer HP1100 MSD Chemstation Rev. A.08.03 (Agilent, Santa Clara, CA, USA). 5 ml of each sample (1 mg/ml) at different pHs were injected in SCAN mode (range 50–1500 m/z) in positive and negative polarity; the parameters of the ionization source were the following: flow 0.3 ml/min of water–ethanol 50:50; drying gas flow 10 l/min; nebulizer pressure 30 psi; drying gas temperature 350 C; capillary voltage 4000 V for positive ionization and 3500 V for negative ionization.
Nanocrystal morphology was determined using a Scanning Electron Microscope (SEM) (Stereoscan 360, Cambridge Instruments, Cambridge, United Kingdom). Samples were mounted on a metal stub with double-sided adhesive tape and then sputtered under vacuum with a gold layer of about 200 Å thickness using a metallizator (Balzer MED 010, Linchestein). The particle size of coarsest crystals was determined by measuring the Ferret’s diameter of 500 particles. 2.13. Statistical analysis Data were analyzed by one-way analysis of variance (ANOVA), using a Bonferoni test. The statistical analysis was conducted using Origin1 software (version 8.5) (Northampton, MA, USA). Results
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
are shown as mean S.D. (standard deviation), and considered significantly different when P < 0.05.
Table 3 Batches produced under different conditions of pH and crystallization method. Batch
Crystallization pH of the crystallization conditions solution (nominal) and chemical used to adjust the pH
Solid form Obtained under crystallization
B1_A B1_B B4.5_A B4.5_B
pH 1 (HCl) pH1 (HCl) pH 4.5 pH 4.5
Method Method Method Method
A B A B
B7_ANaOH B7_BNaOH B7ANH4 OH B7BNH4 OH B8_ANaOH B8_BNaOH B8ANH4 OH B8BNH4 OH B10_ANaOH B10_BNaOH B10ANH4 OH B10BNH4 OH B12_ANaOH Trihydrate a form
pH 7 (NaOH) pH 7 (NaOH) pH 7 (NH4OH) pH7 (NH4OH) pH 8 (NaOH) pH 8 (NaOH) pH 8 (NH4OH) pH 8 (NH4OH) pH 10 (NaOH) pH 10 (NaOH) pH 10 (NH4OH) pH 10 (NH4OH) pH 12 (NaOH) – –
Method Method Method Method Method Method Method Method Method Method Method Method Method
A B A B A B A B A B A B A
a form a form a form a form + some nuclei of g form a form + salt a form + salt a form g form a form + salt a + salt a form g form a form + salt a + salt a form g form
3. Results 3.1. Physicochemical characterization of indomethacin crystals obtained by different methods In order to prepare IDM solutions of appropriate concentration, an evaluation of the IDM solubility at 25 C in aqueous ethanolic solutions at different pHs was carried out (Table 2). IDM solubility remained very low at acidic pHs, while progressively increased in neutral pH and with the pH increase. This result is in agreement with Jain (2008), carried out in aqueous solutions at different pHs. In addition, in the present study it was possible to highlight that IDM solubility in solutions alkalized with NaOH 2 M is always higher than those alkalized with NH4OH, probably because in presence of NaOH sodium trihydrate is formed, that is more soluble than non salified forms. Thus, the IDM solubility values allowed the preparation of solutions for further crystallizations. Actually, to obtain crystals, IDM saturated solutions at 25 C were prepared taking into account the solubility results; the complete dissolution of IDM, necessary to destroy any nuclei, was favored by heating IDM solutions up to 50 C for both crystallization methods. In Table 3, a list of all the IDM batches produced with the two methods at different pHs is given. SEM photomicrographs of the crystals are shown in Fig. 2. DSC thermograms are given in Fig. 3, while in Table 4 both results from DSC and STA–TGA are reported. XRPD patterns for both crystallization methods A and B are respectively given in Figs. 4 and 5. To obtain solutions at a pH 1.0 (batches B1_A and B1_B), the aqueous ethanolic solution was acidified with HCl 0.1 N. By both methods A and B, it was possible to obtain crystals characterized by a needle-shaped form, as showed by SEM microphotographs. It was previously showed (Kaneniwa et al., 1985), that particles of polymorphic a form were needle-shaped and this same conclusion was reported by Martena et al. (2012). Our reference a form, obtained in ethanol solutions, showed very thin needle-shaped crystals. DSC and XRPD analyses confirmed that pure a form crystals were recovered at pH 1.0 by both methods. The DSC thermograms showed the endothermic peak corresponding to the melting of the a form and no other peaks were evident whatever the heating rate (5.0, 10.0, 40.0 C/min). The XRPD patterns are completely superimposable to the reference a form by both methods. The percentage of water content of both batches was lower than 1.0% and thus in agreement with the reference materials. During preliminary tests, assays at pHs 2.0 and 3.0 were also carried out. Results are identical to those obtained at pH 1.0 and thus they are not given for simplicity. To obtain a solution of pH 4.5 that corresponds to the pKa of IDM, no acids or alkali were added to the IDM aqueous ethanolic solution.
Table 2 Indomethacin solubility in aqueous ethanolic solutions at different pHs at 25 C. Theoretical pH of solution
Experimental pH of solution Solubility mg ml
1.0 (HCl 2 M) 4.5 7.0 (NaOH 2 M) 8.0 (NaOH 2 M) 10.0 (NaOH 2 M) 4.5 (NH4OH 2 M) 7.0 (NH4OH 2 M) 8.0 (NH4OH 2 M) 10.0 (NH4OH 2 M)
1.062 4.457 7.060 8.095 10.948 4.648 7.071 8.184 10.063
0.016 0.060 0.094 0.071 0.080 0.033 0.153 0.072 0.081
0.013 0.056 4.204 4.811 5.625 0.051 4.012 4.502 5.089
0.002 0.020 0.110 0.203 0.340 0.023 0.237 0.324 0.372
1
539
Degradation
Two batches were recovered: the B4.5_A and B4.5_B. Even in this case, the crystals recovered were characterized by a needle-shaped form typical of the a form, which leads to suppose that IDM crystallized under the polymorphic a form at a pH of 4.5 whatever the crystallization method. However, the DSC thermograms of the B4.5_B batch carried out at 5 or 10 C/min revealed a double melting peak, the first one of higher enthalpy change (DH 80.25 1.26 J g 1) corresponding to the melting of the a form and the second one of lower enthalpy change (DH 15.13 0.28 J g 1) corresponding to the melting of the g form. Conversely, when the heating rate was carried out at 20.0, or 40.0 C/min it was possible to observe only the endotherm corresponding to the melting of the a form. The XRPD patterns are typical of the a form for both batches. Thus, the presence of the double endotherm evident at the heating rates of 5 or 10 C/min may depend on the presence of some nuclei of g form that, when the a form is completely melted, favoured the crystallization of the melt under g form that then melts. Faster heating rates are unable to highlight this transition. These nuclei are not detectable by XRPD, because this technique was revealed, in this case, less sensitive than thermal techniques. The water content % of both batches was lower than 1.0% and thus in agreement with the reference materials. To obtain IDM solutions at pH 7.0, it was necessary to add alkali. Two different alkali were used, NaOH 2 M that undoubtedly adds sodium molecules into the solution with the potential to form a salified form, and NH4OH 2 M that, on the contrary, avoids the presence of stable cations in the solution preventing the formation of a stable salt. At pH 7.0, for NaOH (Batches B7_ANaOH, B7_BNaOH), IDM crystallized in a mixture of a and sodium salt forms whatever the method A or B used. The SEM microphotographs showed needleshaped crystals very similar to those previously described prevalently composed by a form crystals. For that matter, the reference trihydrate sodium salt showed an agglomeration of irregular coarse and needle-shaped crystals. However, no irregular crystals in addition to the needle-shaped one were observed in the batches B7_ANaOH and B7_BNaOH, indicating that the a needle-shaped morphology prevails. The DSC thermograms showed three different peaks, the first one corresponding to the dehydration of the trihydrated form, the second one corresponding to the melting of the a form, and the third one corresponding to the melting of the anhydrous sodium salt. DSC thermogram of those batches were compared to the thermogram of the pure IDM sodium trihydrate. In
540
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
regard to the pure forms, the melting enthalpies are lower than those of pure compounds. The XRPD patterns simply correspond to the sum of peaks typical of the two forms. It is noteworthy that the peak intensity of the two batches are lower than that of pure compounds. The lower enthalpy change and lower peak intensity may lead to conclude in a decrease in apparent crystallinity occurring during crystallization for these two batches. The water content % confirms that a certain amount of hydration water is lost during heating, corresponding to the presence of the trihydrate sodium salt. When the alkalinization was carried out by NH4OH (B7ANH4 OH and B7BNH4 OH ), IDM crystallized under the pure a form by the method A and pure g form by the method B. The SEM microphotograph of crystals of batch B7ANH4 OH are needle-shaped typical of the a form, while those of batch B7BNH4 OH are irregular particles (coarse crystals exist together with small irregular crystals). In the case of these two batches, the thermograms and the XRPD confirm their crystalline form. The water content % is in agreement with anhydrous crystalline forms. At pH 8.0, in presence of NaOH, two different batches were obtained: the batches B8_ANaOH and B8_BNaOH that crystallize under a mixture of a and trihydrate sodium form. The particles of the batch B8_ANaOH lost the typical needle-shaped habit to assume a shape more similar to a parallelepiped. The particles of the batch B8_BNaOH lost the parallepipedic shape to assume an irregular isodimensional one. The DSC thermogram of both batches is characterized by the presence of endotherms corresponding to the dehydration of the trihydrate form, and to the melting of the a and sodium salt forms. The enthalpies for these three curves are lower than those of corresponding pure forms. XRPD patterns show the presence of the peaks typical of the two forms, but the peak intensity is very low. Thus, in the case of both batches, a decrease in
apparent crystallinity must be emphasized. The water content % is in accordance with the presence of the trihydrate sodium form. At pH 8.0 in presence of NH4OH, two different batches were obtained: the batch B8ANH4 OH , that crystallizes under the a form, and the batch B8BNH4 OH , that crystallizes under the g form. The crystal habit of particles of the batch B8ANH4 OH is that typical of the a form, with the typical needle-shaped crystal, while the shape of batch B8BNH4 OH is that typical of the g form. The thermograms of the two batches are very close to those of the reference particles for both extrapolated onset temperature and enthalpy content. XRPD patterns are completely superimposed to those of the reference materials for both peak distances and intensity. The water content % complies with anhydrous crystalline forms. At pH 10.0, when alkalinization was carried out by NaOH, two different batches were obtained: the batches B10_ANaOH and B10_BNaOH, that crystallize under a mixture of a and trihydrate sodium form. The particles of the batch B10_ANaOH are characterized by an irregular shape and size. The particles of the batch B10_BNaOH are characterized by an irregular isodimensional shape. The DSC thermogram of both batches is characterized by the presence of endotherms corresponding to the dehydration of the trihydrate form, and to the melting of the a and sodium salt forms. The enthalpies for these three curves are lower than those of corresponding pure forms. XRPD patterns show the presence of the peaks typical of the two forms, but the peak intensity is very low. Hence, in both batches, a decrease in apparent crystallinity must be highlighted. The water content % is in agreement with the presence of the trihydrate sodium form for both batches. At pH 10.0, when alkalinization was carried out by NH4OH, two different batches were obtained: the batch B10ANH4 OH , that crystallizes under the a form, and the batch B10BNH4 OH , that
Fig. 2. SEM microphotographs of indomethacin samples (Magnification 5000).
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
Fig. 2. (Continued)
541
542
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
Fig. 3. DSC Thermograms of indomethacin samples: (a) reference g form; (b) reference a form; (c) B4.5_B (a form + some nuclei of g form); (d) reference sodium trihydrate salt; (e) B7_ANaOH (a form + salt) (heating rate: 10 C/min).
crystallizes under the g form. The crystal habit of particles of the batch B10ANH4 OH is that typical of the a form, with the typical needle-shaped crystals, while the particle shape of the batch B10ANH3 is irregular. The thermograms of the two batches are very close to those of the reference particles for both extrapolated onset temperature and enthalpy content. XRPD patterns are completely superimposed to those of the reference materials for both peak distances and intensity. The water content % is in agreement with the anhydrous crystalline forms. IDM was analyzed by mass spectrometry (MS) and 1H NMR in deuterated methanol at different pHs. These analyses were not suitable for establish the kind of polymorphic form or discriminate one polymorph from another, but allowed to evaluate the instability of IDM at basic pHs. IDM crystallized by both methods A and B at pHs 1 to pH 10.0 was stable, and under 1H NMR analysis all the protons were perfectly integrated (refer to Supplementary Material). At basic
pHs (12) the amidic bound was broken into the indol ring and acid 4-chlorobenzoic. 1H NMR spectrum showed the presence of acid 4-chlorobenzoic protons between 7 and 8 ppm. Moreover, it was possible to note a chemical shift of the methylene protons. The MS analysis confirmed results of 1H NMR. The spectrum of the positive ionization of products at pHs 1.0 to 10.0 highlighted the presence of IDM with a molecular weight of 357.08 (M) and the following ionization products: M + H+ = 358.1; M + Na+ = 380.0; 2M + Na+ = 737.2 (refer to Supplementary Material). The negative ionization of products at pHs 1.0–10.0 showed the formation of the following species: M H+ = 356.1; M + Cl = 392.0; 2M + H = 713.3. The positive ionization of the product at pH 12.0 showed the presence of the indol ring, with molecular weight of 218.08 (M), as M + Na+ = 242.1. The negative ionization of the product at pH 12.0 showed the peak of the acid 4chlorobenzoic, with molecular weight of 156.0 (M), present as M H+ = 154.9.
k
h g f e
k
Intensity (a.u.)
Intensity (a.u.)
j i
j i h g f e
d
5
10
15
20
25
30
c
d c
b a
b a
35
2θ Fig. 4. Comparison of indomethacin X-ray powder diffraction patterns of samples produced with the method A: (a) reference g form; (b) reference a form; (c) reference sodium trihydrate salt; (d) B1_A; (e) B4.5_A; (f) B7_ANaOH; (g) B7ANH4 OH ; (h) B8_ANaOH; (i) B8ANH4 OH ; (j) B10_ANaOH; (k) B10ANH4 OH .
5
10
15
20
25
30
35
2θ Fig. 5. Comparison of indomethacin X-ray powder diffraction patterns of samples produced with the method B: (a) reference g form; (b) reference a form; (c) reference sodium trihydrate salt; (d) B1_B; (e) B4.5_B; (f) B7_BNaOH; (g) B7BNH4 OH ; (h) B8_BNaOH; (i) B8BNH4 OH ; (j) B10_BNaOH; (k) B10BNH4 OH .
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
Summarizing, both mass spectrometry (MS) and 1H NMR studies proved that IDM degraded at pHs 12.0. These results confirm previous evidences that proved the chemical degradation of IDM in strong alkali (Hajratwala and Dawson, 1977; Archontaki, 1995). Hence, in this study, the products obtained at pHs 12.0 were not characterized from the physicochemical point of view. 4. Discussion In a previous study (Martena et al., 2012), it was shown that the crystallization of IDM under spray drying from an aqueous ethanolic solution induced the formation of a mixture prevalently composed by the a form and a certain amount of the g form as proven by the XRPD. The prevalence of the metastable a form (lower melting temperature as compared to the g form) by a spray drying process was ascribed to the fact that the two polymorphs a and g have a monotropic relationship and therefore, in a bottom up technique such as the spray drying, the generation of the metastable a polymorph is expected: the nucleation that occurs in the metastable zone and is followed by subsequent crystal growth favours the nucleation and growth of the metastable a form, which is the form of higher energy content (free Gibbs energy) and higher disorder degree (entropy) (Grant 1999). Results in the present study showed that IDM always crystallized in the a form when the method A was applied whatever the pH. Traces of the g form appeared with the method B at the pKa, and the pure g form was obtained in presence of NH4OH at pHs 7.0, 8.0, and 10.0. The trihydrate sodium salt was always formed in presence of NaOH, but it was never pure and always coexisted with the metastable a form whatever the crystallization method. Actually, the moles of NaOH added to reach the requested pHs are not enough to salify all the IDM molecules (Table 1), and thus the IDM excess crystallized under an acidic form. It is
543
interesting to note that the presence of NaOH always favoured the crystallization into the metastable a form even under the Method B that is characterized by a lower temperature of nucleation. The crystal packing of the g form has been described in depth (Kistenmacher and Marsh, 1972; Masuda et al., 2006): crystals are triclinic and each unit cell accommodates two molecules of same conformation (Z = 2). No hydrogen bonds are established between the two molecules. The crystal packing of the a form has been also described in depth (Chen et al., 2002; Masuda et al., 2006): crystals are monoclinic and each unit cells accommodates a couple of three different conformation molecules (A, B, and C) (Z = 6). These three molecules mutually form a hydrogen bond: a hydrogen-bonded carboxylic acid dimer between carbon number 11 of molecule A and B, and hydrogen bonding occurs between the carboxylic acid group of molecule C (carbon number 11) and the amide carbonyl group of molecule B (carbon number 14) (Fig. 1). At the acidic pHs (lower than pKa), IDM only exists as neutral undissociated form (R COOH) and, whatever the crystallization conditions (method A or B), it may easily establish hydrogen bonds orienting the crystallization under the formation of the a form. At the pKa (pH 4.5), indomethacin exists in equilibrium between the dissociated (R COO ) and the undissociated form (R COOH). The method A favours the crystallization under the pure a form: the metastable form is thermodynamically favoured and, as far as the crystallization proceeds, the equilibrium is displaced towards the undissociated form that favours the formation of the a form. With the method B, the a form seems still predominant, but this different crystallization condition (lower crystallization rate and temperature) may favour the appearance of some nuclei of the g form which is the thermodynamically favoured form. The IDM existing as undissociated form may easily establish hydrogen bonds and thus preferentially orientates the
Table 4 Thermal analysis of indomethacin samples. Batch
Crystalline form
Water content %a
0.23 0.11 0.47 0.13 12.45* 12.30 0.24** 0.55 0.16 0.49 0.20 0.45 0.10 0.48 0.14
Native crystals (g form) a form (reference) Sodium salt (reference)
g form a form
B1_A B1_B B4.5_A B4.5_B B7_ANaOH
a form a form a form a form + some nuclei of g form a form + salt
B7_BNaOH
a form + salt
7.88 1.05
B7ANH4 OH B7BNH4 OH B8_ANaOH
a form g form a form + salt
0.53 0.12 0.28 0.10 6.49 1.16
B8_BNaOH
a form + salt
5.24 1.08
B8ANH4 OH B8BNH4 OH B10_ANaOH
a form g form a form + salt
0.50 0.16 0.25 0.14 8.12 1.08
B10_BNaOH
a form + salt
5.27 1.35
B10ANH4 OH B10BNH4 OH B12_ANaOH
a form g form
Trihydrate sodium form
Degradation
7.20 0.89
0.52 0.09 0.22 0.08 –
Dehydration endothermb
Melting endothermb
Tm C (DH J g
Tm C
DH J g
157.67 0.13 153.95 0.32 266.25 1.24
110.63 1.72 85.66 2.24 112.96 3.44
– – 74.33 1.76 (75.06 2.89) – – – – 70.24 (25.28 69.57 (37.13 – – 71.38 (27.89 70.58 (25.21 – – 70.24 (30.12 69.25 (24.15 – – –
*Theoretical value. **Experimental value. a Determined by TGA–STA. b Determined by conventional DSC at 10 C/minDeDetermined by conventional DSC at 10 C/min.
2.28 1.14) 1.15 1.18)
1.89 2.77) 2.66 2.13)
1.08 2.68) 3.12 1.68)
1
)
153.38 153.02 153.05 153.12 157.66 150.22 265.06 151.65 264.56 153.24 157.55 150.27 265.27 149.88 265.98 153.12 157.08 149.55 265.90 148.22 266.14 153.04 157.66 –
0.56 0.75 0.24 0.85 1.86 0.27 1.25 0.78 1.84 0.55 1.56 1.18 2.35 2.23 2.76 0.65 1.78 1.22 1.67 3.55 3.27 0.77 1.24
87.28 84.22 52.88 80.25 15.13 56.92 27.12 50.29 29.30 87.37 108.27 45.37 25.31 46.28 32.71 86.24 110.63 40.25 27.40 44.22 30.11 86.18 110.63 –
1
1.67 3.55 3.55 1.26 0.28 1.37 1.44 1.47 2.45 1.18 2.45 2.36 3.28 2.17 2.44 3.18 1.72 3.19 2.79 2.57 1.12 1.27 1.72
544
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
crystallization under the a form. Nevertheless, the presence of a certain amount of dissociated IDM may limit the establishment of hydrogen bond thus orienting IDM in different molecule arrangement that favours the crystallization under the g form. Nevertheless, at pHs higher than the pKa and Method A, the a form is always present (sometimes together with the salt when the sodium counterion is present in the crystallization medium). At these pHs, the dissociated form (R COO ) is predominant, but the a form is still favoured. Once more, this is due to the crystallization conditions (rate and temperature) that favour the formation of the metastable form. On the contrary, method B conditions in presence of NH4OH favoured the crystallization under the stable g form, which is for that matter independent to the dissociation of the IDM. Differently, under method B in presence of NaOH the form a is still favoured: probably the presence of the counterion favoured the establishment of hydrogen bonds and probably the typical crystal packing of the a form. 5. Conclusions This study proved the effect of pH and crystallization conditions on the crystallization of indomethacin. Crystallization conditions, pH of crystallization solution, and the presence of the sodium counterion may orientate the crystallization of IDM towards the a, g polymorphic forms and/or the formation of the trihydrate sodium salt. Since polymorphs and salts significantly influences the physicochemical properties and particle morphology of indomethacin, consequences on formulation and technological behavior of the drug may be expected. Acknowledgements The authors would like to thank Mrs Laura Petetta for her contribution to SEM analysis and Mr Giuliano Camaro Blanco for his contribution to the experimental section. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijpharm.2014. 07.030. References Allen, D.J., Kwan, K.C., 1969. Determination of degree of crystallinity in solid-solid equilibria. J. Pharm. Sci. 58, 1190–1193. Andronis, V., Zografi, G., 2000. Crystal nucleation and growth of indomethacin polymorphs from the amorphous state. J. Non-Cryst. Sol. 271, 236–248. Archontaki, H.A., 1995. Kinetic study on degradation of indomethacin in alkaline aqueous solutions by derivative ultraviolet spectrophotometry. Analyst 120, 2627–2634. Borka, L., 1974. The polymorphism of indomethacine. New modifications, their melting behavior and solubility. Acta Pharm. Suec. 11, 295–303. Chen, X., Morris, K.R., Griesser, U.J., Byrn, S.R., Stowell, J.G., 2002. Reactivity differences of indomethacin solid forms with ammonia gas. J. Am. Chem. Soc. 124, 012–15019.
Crowley, K.J., Zografi, G., 2002. Cryogenic grinding of indomethacin polymorphs and solvates: assessment of amorphous phase formation and amorphous phase physical stability. J. Pharm. Sci. 91, 492–507. Grant, D.J.W., 1999. Theory and origin of polymorphism. In: Brittain, H.G. (Ed.), Polymorphism in Pharmaceutical Solids. Marcel Dekker, New York, pp. 227–278. Hajratwala, B.R., Dawson, J.E., 1977. Kinetics of indomethacin degradation I: presence of alkali. J. Pharm. Sci. 66, 27–29. Jain, A.K., 2008. Solubilization of indomethacin using hydrotropes for aqueous injection. Eur. J. Pharm. Biopharm. 68, 701–714. Kaneniwa, N., Otsuka, M., Hayashi, T., 1985. Physicochemical characterization of indomethacin polymorphs and the transformation kinetics in ethanol. Chem. Pharm. Bull. 33, 3447–3455. Kistenmacher, T.J., Marsh, R.E., 1972. Crystal and moleculr structure of an antiinflammatory agent, indomethacin, 1-(p-chlorobenzoyl)-5-methoxy-2-methylindole-3-acetic acid. J. Am.Chem. Soc. 23, 1340–1345. Kulkarni, S., Gupta, S.P., Upmanyu, N., Tonpay, S.D., 2011. Solubility enhancement of water insoluble drug for ophthalmic formulation. Int. J. Drug Deliv. 3, 141–148. Legendre, B., Feutelais, Y., 2004. Polymorphic and thermodynamic study of indomethacin. J. Therm. Anal. Calorim. 76, 255–264. Llinàs, A., Box, K.J., Burley, J.C., Glen, R.C., Goodman, J.M., 2007. A new method for the reproducible generation of polymorphs: two forms of sulindac with very different solubilities. J. Appl. Cryst 40, 379–381. Martena, V., Censi, R., Hoti, E., Malaj, L., Di Martino, P., 2012. Indomethacin nanocrystals prepared by different laboratory scale methods: effect on crystalline form and dissolution behavior. J.Nanopart. Res. 14, 1275–1289. Martínez-Ohárriz, M.C., Rodríguez-Espinosa, C., Martín, C., Goñi, M.M., TrosIlarduya, M.C., Sánchez, M., 2002. Solid dispersions of diflunisal-PVP: polymorphic and amorphous states of the drug. Drug Dev. Ind. Pharm. 28, 717–725. Masuda, K., Tabata, S., Kono, H., Sakata, Y., Hayase, T., Yonemochi, E., Terada, K., 2006. Solid-state 13C NMR study of indomethacin polymorphism. Int. J. Pharm. 318, 146–153. Monkhouse, D.C., Lach, J.L., 1972. Use of adsorbents in enhancement of drug dissolution. J. Pharm. Sci. 61, 1435–1441. Nair, R., Gonen, S., Hoag, S.W., 2002. Influence of polyethylene glycol and povidone on the polymorphic transformation and solubility of carbamazepine. Int. J. Pharm. 240, 11–22. Singhal, D., Curatolo, W., 2004. Drug polymorphism and dosage form design: a practical perspective. Adv. Drug Deliv. Rev. 56, 335–347. Surwase, S.A., Boetker, J.P., Saville, D., Boyd, B.J., Gordon, K.C., Peltonen, L., Strachan, C.J., 2013. Indomethacin: new polymorphs of an old drug. Mol. Pharm. 10, 4472–4480. Tong, P., Zografi, G., 1999. Solid-state characteristics of amorphous sodium indomethacin relative to its free acid. Pharm. Res. 16, 1186–1192. Tong, P., Zografi, G., 2004. Effects of water vapor absorption on the physical and chemical stability of amorphous sodium indomethacin. AAPS Pharm. Sci. Tech. 5, 1–8. Vippagunta, S.R., Brittain, H.G., Grant, D.J.W., 2001. Crystalline solids. Adv. Drug Deliv. Rev. 48, 3–26. Yamamoto, H., 1968. 1-acyl-indoles. II. A new syntheses of 1-(ion-chlorobenzoyl)-5methoxy-3-indolacetic acid and its polymorphism. Chem. Pharm. Bull. 16, 17–19. Yoshinari, T., Forbes, R.T., York, P., Kawashima, Y., 2002. Moisture induced polymorphic transition of mannitol and its morphological transformation. Int. J. Pharm. 247, 69–77. Yoshinari, T., Forbes, R.T., York, P., Kawashima, Y., 2003. The improved compaction properties of mannitol after a moisture-induced polymorphic transition. Int. J. Pharm. 258, 121–131. Yu, L., Reutzel, S.M., Stephenson, G.A., 1998. Physical characterization of polymorphic drugs: an integrated characterization strategy. Pharm. Sci. Technol. Today 1, 118–127. Yu, L., 2001. Amorphous pharmaceutical solids: preparation, characterization and stabilization. Adv. Drug Deliv. Rev. 48, 27–42. Yu, L., Ng, K., 2002. Glycine crystallization during spray drying the pH effect on salt and polymorphic forms. J. Pharm. Sci. 91, 2367–2375. Zhang, G.G.Z., Law, D., Schmitt, E.A., Qiu, Y., 2004. Phase transformation considerations during process development and manufacture of solid oral dosage forms. Adv. Drug Deliv. Rev. 56, 371–390.
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Influence of pH and method of crystallization on the solid physical form of indomethacin Alessandra Dubbini a , Roberta Censi a , Valentina Martena a , Ela Hoti b , Massimo Ricciutelli a , Ledjan Malaj b , Piera Di Martino a, * a b
University of Camerino, School of Pharmacy, Via S. Agostino, Camerino, Italy University of Medicine, Tirana, Faculty of Pharmacy, Street of Dibra, Tirana, Albania
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 April 2014 Received in revised form 21 July 2014 Accepted 22 July 2014 Available online 23 July 2014
The purpose of this study was to investigate the effect of pH and method of crystallization on the solid physical form of indomethacin (IDM). IDM, a non steroidal anti-inflammatory drug poorly soluble in water, underwent two different crystallization methods: crystallization by solvent evaporation under reduced pressure at 50.0 C (method A), and crystallization by cooling of solution from 50.0 to 5.0 C (method B). In both cases, several aqueous ethanolic solutions of IDM of different pHs were prepared. pHs were adjusted by adding acidic solutions (HCl 2 M) or alkali (NaOH or NH4OH 2 M) to an aqueous ethanolic solution of IDM. Thus, several batches were recovered after crystallization. The chemical stability of IDM was verified through 1H NMR and mass spectroscopy (FIA-ESI-MS), that revealed that IDM degraded in strong alkali media (pH 12). Crystals obtained under different crystallization conditions at pHs of 1.0, 4.5, 7.0, 8.0, 10.0 and chemically stable were thus characterized for crystal habit by scanning electron microscopy, for thermal behaviour by differential scanning calorimetry, and thermogravimetry, and for solid state by X-ray powder diffractometry. Under the Method A, IDM always crystallized into pure metastable alpha form when solutions were acidified or alkalized respectively with HCl and NH4OH. On the contrary, in presence of NaOH, IDM crystallized under a mixture of alpha and sodium trihydrate form, because the presence of the sodium counter ion orientates the crystallization towards the formation of the trihydrate salt. Under the method B, at pH of 1.0, IDM crystallized under the alpha form; at pH 4.5, IDM crystallized under the form alpha in presence of some nuclei of gamma form; at pH 7.0, 8.0, and 10.0 for NH4OH, IDM crystallized under the most stable polymorph gamma form, whereas in presence of NaOH, a mix of alpha, and salt forms was formed whatever the pH of the solution. ã 2014 Elsevier B.V. All rights reserved.
Keywords: Indomethacin Polymorphism pH Crystallization
1. Introduction Knowledge of solid state transitions occurring in pure drugs during processing or storage is fundamental for developing stable pharmaceutical formulations (Yu et al., 1998). Polymorphic transitions, hydration–dehydration processes, amorphization, and salt formation occurring intentionally or unintentionally during drug processing must be taken into account, because different solid forms lead to different physical properties (solubility and dissolution rate, compression ability and attitude to grinding, long term physical stability, etc.) (Yoshinari et al., 2002; Nair et al., 2002; Vippagunta et al., 2001; Zhang et al., 2004; Yoshinari et al., 2003; Martínez-Ohárriz et al., 2002; Singhal and Curatolo, 2004; Yu, 2001).
* Corresponding author. Tel.: +39 737 402215; fax: +39 737 637345. E-mail address: [email protected] (P. Di Martino). http://dx.doi.org/10.1016/j.ijpharm.2014.07.030 0378-5173/ ã 2014 Elsevier B.V. All rights reserved.
Molecules containing ionizable groups (carboxyl, amino, etc.) can crystallize in neutral form or as salts with their counter ions. It was demonstrated that it is possible to crystallize different polymorphs according to the solution pH. In aqueous solution, a sodium salified form of sulindac exists in equilibrium between two different polymorphic forms; this fact allows to isolate in the solid state the two different polymorphic forms I and II according to the pH of the crystallization solution (Llinàs et al., 2007). Yu and Ng (2002) investigated the effect of pH on glycine crystallization during spray drying. This study revealed the crystallization of glycine on different polymorphs and salt forms according to the crystallization pH. The authors of the present study intend to consider in dept the possibility to induce a change in the solid physical form of a ionizable drug by changing crystallization conditions and solution pH. In particular, the effect of pH and method of crystallization on the solid state of indomethacin will be investigated. Indomethacin (IDM) (the molecular structure is given in Fig. 1), a non steroidal anti-inflammatory drug (NSAID), is a poorly water
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
O
2' 14
7 6
9
5
8
O
4
4'
Cl 2.2. Preparation of IDM solution at different pHs
N1 2
H3C
Crystals were stored in a desiccator in presence of P2O5 as desiccant during the experiment time interval.
3'
1'
CH3
3 10
OH
11
537
O Fig. 1. Molecular structure of indomethacin.
soluble active substance with poor bioavailability after oral administration. IDM exists in four different polymorphic forms: the g form (also named form I) (Tm: 157.9 C) that is the most stable thermodynamic form (triclinic); the a form (also named form II) (Tm: 151.4 C) (monoclinic); the b form (also named III) (Borka, 1974; Yamamoto, 1968; Allen and Kwan, 1969; Monkhouse and Lach, 1972; Legendre and Feutelais, 2004), and the d form (also named form IV) (Tm: 130.6 C) (Crowley and Zografi, 2002). A crystalline IDM sodium trihydrate has also been described (Tong and Zografi, 1999; Tong and Zografi, 2004; Kulkarni et al., 2011). During experiments devoted to describe the crystallization behaviour of quench cooled amorphous IDM in media of different pHs and temperatures, new polymorphs were detected, named e, z, and h (Surwase et al., 2013). In the present study, two different crystallization methods were used: 1. Method A: crystallization by solvent evaporation under
reduced pressure at 50 C. 2. Method B: crystallization by cooling of solution from 50.0 to
5.0 C.
The aqueous ethanolic solutions of indomethacin were prepared by adding an accurately weighed amount of IDM to a solution of water and ethanol (1:1). The initial pH was measured (Jenway 3510 pHmeter, Staffordshire, UK). Then, the solution was charged with HCl, NaOH, or NH4OH 2 M up to the desired pH (1.0, 7.0, 8.0 10.0, 12.0). Only for solutions at pH 4.5, neither acid or alkali were added exploiting the fact that this pH value, that corresponds to the pKa of indomethacin, may be easily reached by adjusting the amount of IDM added to the aqueous ethanolic solutions. The actual pH was again measured. Table 1 gives details about the preparation conditions of IDM solutions. Results are the mean of three different measurements. 2.3. Indomethacin equilibrium solubility in solutions of different pH The evaluation of the equilibrium solubility was performed with the only scope to obtain saturated solutions at 25 C at different pHs. An excess of g IDM (native crystals) was added to 50 ml of the aqueous ethanolic solution of the required pH and maintained at 25 0.5 C under continuous stirring in an incubator (Velp Scientifica, FTC 90E, Usmate, Italy). The required pH was adjusted either with HCl 1 N, NH4OH, or NH4OH 2 M. Equilibrium between the undissolved drug and the drug in solution was tested and judged to have been reached when three consecutive spectrophotometric measurements differed by no more than 1%, generally after 24 h. The equilibrium solubility was thus determined by withdraws aliquots, centrifuged at 5000 rpm at 25 0.5 C for 10 min (High Speed Micro-Centrifuge Scilogex D3024R, Berlin, CT, USA); after appropriate dilution, the concentration of the solution was determined by UV spectrophotometry at a wavelength of 318 nm (Cary 1E UV–VIS, Varian, Leinì, Italy). Assays were performed in triplicate.
For each method, different aqueous ethanolic solutions of different pHs (1.0, 4.5, 7.0, 8.0, 10.0, and 12.0) were tested to crystallize IDM.
2.4. Crystallization of indomethacin through different methods
2. Experimental
Taking into account the solubility measurements, and after the preparation of solutions at different pHs, IDM crystals were recovered by using two different crystallization methods:
2.1. Materials Indomethacin (Native Crystals: NCs) (g form) was kindly supplied by Fabbrica Italiana Sintetici (F.I.S., Vicenza, Italy) as white crystalline powder. Ultrapure water was produced by Gradient Milli-Q1 (Millipore, Molsheim, France). Chemicals, all of analytical grade, were supplied by Sigma–Aldrich (Steinheim, Germany).
2.4.1. Method A: crystallization by solvent evaporation under reduced pressure at 50 C The saturated solution prepared at 25 0.5 C was heated up to 50 0.5 C in order to completely dissolve IDM. The solvent was thus totally evaporated under reduced pressure at a temperature of 50 C (Rotavapor1 R-210, Buchi, Flawil, Switzerland; vacuum
Table 1 Aqueous ethanolic solutions of indomethacin obtained after addition of acids or alkali to reach solutions of the desired pH. pH required
Ethanol + Water (g)
IDM (mg)
1.0 4.5 (pKa) 7.0 8.0 10.0 12.0 7.0 8.0 10.0 12.0
25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0
50.730 50.400 50.695 50.800 50.400 51.225 50.425 50.255 50.265 50.255
+ + + + + + + + + +
25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0 25.0
Starting pH of ethanol/water/IDM solution 0.028 0.607 0.714 0.567 0.607 0.035 0.318 0.0778 0.0636 0.0778
4.323 4.457 4.309 4.410 4.457 4.248 4.426 4.430 4.453 4.431
0.052 0.060 0.255 0.050 0.060 0.169 0.043 0.014 0.0184 0.012
Added volume of acid or alkali (ml)
Final pH
4400.0 (HCl 2 M) – 60.0 (NaOH 2 M) 67.0 (NaOH 2 M) 73.0 (NaOH 2 M) 90.0 (NaOH 2 M) 48.0 (NH4OH 2 M) 59.0 (NH4OH 2 M) 352.0 (NH4OH 2 M) 47680.0 (NH4OH 2 M)
1.062 4.457 7.060 8.095 10.948 12.019 7.071 8.184 10.063 11.947
0.016 0.060 0.094 0.071 0.080 0.013 0.153 0.072 0.081 0.160
538
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
pump, V-710, Buchi, vacuum level 680 mmHg). The residual solvent was then eliminated by placing crystals in a desiccator in presence of P2O5 as desiccant. Crystals were stored in these conditions for the entire duration of the experiments. 2.4.2. Method B: crystallization by cooling of solution The saturated solution prepared at 25 0.5 C was heated at 50 0.5 C until the complete dissolution of the drug. The solution was then quench cooled under stirring at 5.0 1.0 C by simply placing the solution container in a glass bath and this temperature was maintained until the crystallization of the IDM (approximately for 24–48 h). Crystals were filtered out and maintained in a desiccator in presence of P2O5 as desiccant. Crystals were stored in these conditions for the entire duration of the experiments. 2.5. Preparation of indomethacin a and amorphous forms The a form of IDM was prepared by a modification of the method of Borka (Borka 1974; Kaneniwa et al., 1985). 10 g of IDM bulk powder were dissolved in 10 ml of ethanol at 80 0.5 C; the undissolved drug was filtered off, and 20 ml of distilled water at room temperature was added to the IDM-saturated ethanol solution at 80 0.5 C. The precipitated crystals were removed by filtration, using a glass funnel, and then dried overnight in a P2O5 desiccator under vacuum at room temperature. The X-ray powder diffraction (XRPD) patterns of this sample were compared to those described in literature (Kaneniwa et al., 1985) to confirm the obtainment of a pure a form. Amorphous IDM (Andronis and Zografi, 2000) was obtained by cooling in liquid nitrogen after melting the bulk powder at 165 C for 5 min. The crystallinity of this material, assessed by X-ray powder diffractometry, was assumed to be 0%. IDM a and amorphous forms were stored in a desiccator in presence of P2O5 as desiccant during the experiment time interval. 2.6. Preparation of the indomethacin sodium trihydrate The IDM sodium trihydrate was prepared according to Kulkarni et al. (2011). An appropriate amount of IDM was added to 50 ml of methanol and maintained at 50 0.5 C until the complete dissolution of the drug. The solution was left to cool spontaneously until room temperature. An appropriate amount of NaOH 2 M was then added to the solution and kept under stirring for 2 h. The solution was then evaporated under reduced pressure at 35 0.5 C. The solid was desiccated in presence of P2O5 as desiccator for 24 h. Powder is then kept in well closed vials. X-ray powder diffraction patterns of this sample were compared to those described in literature (Tong and Zografi, 2004) to confirm the obtainment of a pure trihydrate sodium indomethacin.
2.8. Evaluation of the chemical stability by 1H NMR The chemical stability of IDM samples was also evaluated through 1H NMR (Varian, Mercury Console 400, Palo Alto, USA). Sample were analysed in deuterated methanol at different pHs. 2.9. Thermogravimetric analysis Thermogravimetric analysis was used to determine the powder water content and it was carried out by the simultaneous thermal analysis (STA) which enables to simultaneously analyze a sample for change in weight (Thermogravimetric analysis, TGA) and change in enthalpy flow (Differential Scanning Calorimetry, DSC). In this article, the thermogravimetric analysis is referred to as STA–TGA. The analysis was performed using a Simultaneous Thermal Analyser (STA 6000, PerkinElmer, Inc., Waltham, MA, USA), under-nitrogen atmosphere (20 ml/min) in 0.07 ml open aluminium oxide pans. Samples were heated from room temperature to 100 C at a heating rate of 5.0 C/min. STA was calibrated for temperature and heat flow with three standard metals (indium and tin), taking into account their expected melting temperatures (156.75 and 232.08 C respectively), and for weight with an external PerkinElmer standard (Calibration Reference Weight P/N N520-0042, Material lot 91101GB, Weight 55.98 mg, 01/23/08 VT). Calibration was repeatedly checked to assure deviation 0.3 K. 2.10. Differential scanning calorimetry analysis Differential scanning calorimetry (DSC) analysis was performed on a Pyris 1 (PerkinElmer, Co., Norwalk, USA) equipped with a cooling device (Intracooler 2P, Cooling Accessory, PerkinElmer, Co., Norwalk, USA). A dry purge of nitrogen gas (20 ml/min) was used for all runs. DSC was calibrated for temperature and heat flow using a pure sample of indium and zinc standards. Sample mass was about 4–5 mg and aluminium perforated pans of 50 ml were used. Samples were heated from room temperature to 200 C or 300 C at different heating rates 1.0, 5.0, 10.0, 20.0 and 40.0 C min 1. 2.11. X-ray powder diffractometry X-ray powder diffractometry (XRPD) was carried out on a Philips PW 1730 (Philips Electronic Instruments Corp., Mahwah, NJ, USA) as X-ray generator for Cu Ka radiation (la1= 1.54056 Å, la2 = 1.54430 Å) was used. The experimental X-ray powder patterns were recorded on a Philips PH 8203. The goniometer supplied was a Philips PW 1373 and the channel control was a Philips PW 1390. Data were collected in the discontinuous scan mode using a step size of 0.01 2u. The scanned range was 2 –40 (u). 2.12. Scanning electron microscopy
2.7. Evaluation of the chemical stability by FIA-ESI-MS The evaluation of the chemical stability of IDM during processing was evaluated by FIA-ESI-MS analysis (Flow Injection Analysis-Electrospray Ionization-Mass Spectrometry) (purity test) performed by an HPLC 1090Hewlett Packard Series II (Agilent, Santa Clara, CA, USA) equipped with a Hewlett Packard Mass spectrometer HP1100 MSD Chemstation Rev. A.08.03 (Agilent, Santa Clara, CA, USA). 5 ml of each sample (1 mg/ml) at different pHs were injected in SCAN mode (range 50–1500 m/z) in positive and negative polarity; the parameters of the ionization source were the following: flow 0.3 ml/min of water–ethanol 50:50; drying gas flow 10 l/min; nebulizer pressure 30 psi; drying gas temperature 350 C; capillary voltage 4000 V for positive ionization and 3500 V for negative ionization.
Nanocrystal morphology was determined using a Scanning Electron Microscope (SEM) (Stereoscan 360, Cambridge Instruments, Cambridge, United Kingdom). Samples were mounted on a metal stub with double-sided adhesive tape and then sputtered under vacuum with a gold layer of about 200 Å thickness using a metallizator (Balzer MED 010, Linchestein). The particle size of coarsest crystals was determined by measuring the Ferret’s diameter of 500 particles. 2.13. Statistical analysis Data were analyzed by one-way analysis of variance (ANOVA), using a Bonferoni test. The statistical analysis was conducted using Origin1 software (version 8.5) (Northampton, MA, USA). Results
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
are shown as mean S.D. (standard deviation), and considered significantly different when P < 0.05.
Table 3 Batches produced under different conditions of pH and crystallization method. Batch
Crystallization pH of the crystallization conditions solution (nominal) and chemical used to adjust the pH
Solid form Obtained under crystallization
B1_A B1_B B4.5_A B4.5_B
pH 1 (HCl) pH1 (HCl) pH 4.5 pH 4.5
Method Method Method Method
A B A B
B7_ANaOH B7_BNaOH B7ANH4 OH B7BNH4 OH B8_ANaOH B8_BNaOH B8ANH4 OH B8BNH4 OH B10_ANaOH B10_BNaOH B10ANH4 OH B10BNH4 OH B12_ANaOH Trihydrate a form
pH 7 (NaOH) pH 7 (NaOH) pH 7 (NH4OH) pH7 (NH4OH) pH 8 (NaOH) pH 8 (NaOH) pH 8 (NH4OH) pH 8 (NH4OH) pH 10 (NaOH) pH 10 (NaOH) pH 10 (NH4OH) pH 10 (NH4OH) pH 12 (NaOH) – –
Method Method Method Method Method Method Method Method Method Method Method Method Method
A B A B A B A B A B A B A
a form a form a form a form + some nuclei of g form a form + salt a form + salt a form g form a form + salt a + salt a form g form a form + salt a + salt a form g form
3. Results 3.1. Physicochemical characterization of indomethacin crystals obtained by different methods In order to prepare IDM solutions of appropriate concentration, an evaluation of the IDM solubility at 25 C in aqueous ethanolic solutions at different pHs was carried out (Table 2). IDM solubility remained very low at acidic pHs, while progressively increased in neutral pH and with the pH increase. This result is in agreement with Jain (2008), carried out in aqueous solutions at different pHs. In addition, in the present study it was possible to highlight that IDM solubility in solutions alkalized with NaOH 2 M is always higher than those alkalized with NH4OH, probably because in presence of NaOH sodium trihydrate is formed, that is more soluble than non salified forms. Thus, the IDM solubility values allowed the preparation of solutions for further crystallizations. Actually, to obtain crystals, IDM saturated solutions at 25 C were prepared taking into account the solubility results; the complete dissolution of IDM, necessary to destroy any nuclei, was favored by heating IDM solutions up to 50 C for both crystallization methods. In Table 3, a list of all the IDM batches produced with the two methods at different pHs is given. SEM photomicrographs of the crystals are shown in Fig. 2. DSC thermograms are given in Fig. 3, while in Table 4 both results from DSC and STA–TGA are reported. XRPD patterns for both crystallization methods A and B are respectively given in Figs. 4 and 5. To obtain solutions at a pH 1.0 (batches B1_A and B1_B), the aqueous ethanolic solution was acidified with HCl 0.1 N. By both methods A and B, it was possible to obtain crystals characterized by a needle-shaped form, as showed by SEM microphotographs. It was previously showed (Kaneniwa et al., 1985), that particles of polymorphic a form were needle-shaped and this same conclusion was reported by Martena et al. (2012). Our reference a form, obtained in ethanol solutions, showed very thin needle-shaped crystals. DSC and XRPD analyses confirmed that pure a form crystals were recovered at pH 1.0 by both methods. The DSC thermograms showed the endothermic peak corresponding to the melting of the a form and no other peaks were evident whatever the heating rate (5.0, 10.0, 40.0 C/min). The XRPD patterns are completely superimposable to the reference a form by both methods. The percentage of water content of both batches was lower than 1.0% and thus in agreement with the reference materials. During preliminary tests, assays at pHs 2.0 and 3.0 were also carried out. Results are identical to those obtained at pH 1.0 and thus they are not given for simplicity. To obtain a solution of pH 4.5 that corresponds to the pKa of IDM, no acids or alkali were added to the IDM aqueous ethanolic solution.
Table 2 Indomethacin solubility in aqueous ethanolic solutions at different pHs at 25 C. Theoretical pH of solution
Experimental pH of solution Solubility mg ml
1.0 (HCl 2 M) 4.5 7.0 (NaOH 2 M) 8.0 (NaOH 2 M) 10.0 (NaOH 2 M) 4.5 (NH4OH 2 M) 7.0 (NH4OH 2 M) 8.0 (NH4OH 2 M) 10.0 (NH4OH 2 M)
1.062 4.457 7.060 8.095 10.948 4.648 7.071 8.184 10.063
0.016 0.060 0.094 0.071 0.080 0.033 0.153 0.072 0.081
0.013 0.056 4.204 4.811 5.625 0.051 4.012 4.502 5.089
0.002 0.020 0.110 0.203 0.340 0.023 0.237 0.324 0.372
1
539
Degradation
Two batches were recovered: the B4.5_A and B4.5_B. Even in this case, the crystals recovered were characterized by a needle-shaped form typical of the a form, which leads to suppose that IDM crystallized under the polymorphic a form at a pH of 4.5 whatever the crystallization method. However, the DSC thermograms of the B4.5_B batch carried out at 5 or 10 C/min revealed a double melting peak, the first one of higher enthalpy change (DH 80.25 1.26 J g 1) corresponding to the melting of the a form and the second one of lower enthalpy change (DH 15.13 0.28 J g 1) corresponding to the melting of the g form. Conversely, when the heating rate was carried out at 20.0, or 40.0 C/min it was possible to observe only the endotherm corresponding to the melting of the a form. The XRPD patterns are typical of the a form for both batches. Thus, the presence of the double endotherm evident at the heating rates of 5 or 10 C/min may depend on the presence of some nuclei of g form that, when the a form is completely melted, favoured the crystallization of the melt under g form that then melts. Faster heating rates are unable to highlight this transition. These nuclei are not detectable by XRPD, because this technique was revealed, in this case, less sensitive than thermal techniques. The water content % of both batches was lower than 1.0% and thus in agreement with the reference materials. To obtain IDM solutions at pH 7.0, it was necessary to add alkali. Two different alkali were used, NaOH 2 M that undoubtedly adds sodium molecules into the solution with the potential to form a salified form, and NH4OH 2 M that, on the contrary, avoids the presence of stable cations in the solution preventing the formation of a stable salt. At pH 7.0, for NaOH (Batches B7_ANaOH, B7_BNaOH), IDM crystallized in a mixture of a and sodium salt forms whatever the method A or B used. The SEM microphotographs showed needleshaped crystals very similar to those previously described prevalently composed by a form crystals. For that matter, the reference trihydrate sodium salt showed an agglomeration of irregular coarse and needle-shaped crystals. However, no irregular crystals in addition to the needle-shaped one were observed in the batches B7_ANaOH and B7_BNaOH, indicating that the a needle-shaped morphology prevails. The DSC thermograms showed three different peaks, the first one corresponding to the dehydration of the trihydrated form, the second one corresponding to the melting of the a form, and the third one corresponding to the melting of the anhydrous sodium salt. DSC thermogram of those batches were compared to the thermogram of the pure IDM sodium trihydrate. In
540
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
regard to the pure forms, the melting enthalpies are lower than those of pure compounds. The XRPD patterns simply correspond to the sum of peaks typical of the two forms. It is noteworthy that the peak intensity of the two batches are lower than that of pure compounds. The lower enthalpy change and lower peak intensity may lead to conclude in a decrease in apparent crystallinity occurring during crystallization for these two batches. The water content % confirms that a certain amount of hydration water is lost during heating, corresponding to the presence of the trihydrate sodium salt. When the alkalinization was carried out by NH4OH (B7ANH4 OH and B7BNH4 OH ), IDM crystallized under the pure a form by the method A and pure g form by the method B. The SEM microphotograph of crystals of batch B7ANH4 OH are needle-shaped typical of the a form, while those of batch B7BNH4 OH are irregular particles (coarse crystals exist together with small irregular crystals). In the case of these two batches, the thermograms and the XRPD confirm their crystalline form. The water content % is in agreement with anhydrous crystalline forms. At pH 8.0, in presence of NaOH, two different batches were obtained: the batches B8_ANaOH and B8_BNaOH that crystallize under a mixture of a and trihydrate sodium form. The particles of the batch B8_ANaOH lost the typical needle-shaped habit to assume a shape more similar to a parallelepiped. The particles of the batch B8_BNaOH lost the parallepipedic shape to assume an irregular isodimensional one. The DSC thermogram of both batches is characterized by the presence of endotherms corresponding to the dehydration of the trihydrate form, and to the melting of the a and sodium salt forms. The enthalpies for these three curves are lower than those of corresponding pure forms. XRPD patterns show the presence of the peaks typical of the two forms, but the peak intensity is very low. Thus, in the case of both batches, a decrease in
apparent crystallinity must be emphasized. The water content % is in accordance with the presence of the trihydrate sodium form. At pH 8.0 in presence of NH4OH, two different batches were obtained: the batch B8ANH4 OH , that crystallizes under the a form, and the batch B8BNH4 OH , that crystallizes under the g form. The crystal habit of particles of the batch B8ANH4 OH is that typical of the a form, with the typical needle-shaped crystal, while the shape of batch B8BNH4 OH is that typical of the g form. The thermograms of the two batches are very close to those of the reference particles for both extrapolated onset temperature and enthalpy content. XRPD patterns are completely superimposed to those of the reference materials for both peak distances and intensity. The water content % complies with anhydrous crystalline forms. At pH 10.0, when alkalinization was carried out by NaOH, two different batches were obtained: the batches B10_ANaOH and B10_BNaOH, that crystallize under a mixture of a and trihydrate sodium form. The particles of the batch B10_ANaOH are characterized by an irregular shape and size. The particles of the batch B10_BNaOH are characterized by an irregular isodimensional shape. The DSC thermogram of both batches is characterized by the presence of endotherms corresponding to the dehydration of the trihydrate form, and to the melting of the a and sodium salt forms. The enthalpies for these three curves are lower than those of corresponding pure forms. XRPD patterns show the presence of the peaks typical of the two forms, but the peak intensity is very low. Hence, in both batches, a decrease in apparent crystallinity must be highlighted. The water content % is in agreement with the presence of the trihydrate sodium form for both batches. At pH 10.0, when alkalinization was carried out by NH4OH, two different batches were obtained: the batch B10ANH4 OH , that crystallizes under the a form, and the batch B10BNH4 OH , that
Fig. 2. SEM microphotographs of indomethacin samples (Magnification 5000).
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
Fig. 2. (Continued)
541
542
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
Fig. 3. DSC Thermograms of indomethacin samples: (a) reference g form; (b) reference a form; (c) B4.5_B (a form + some nuclei of g form); (d) reference sodium trihydrate salt; (e) B7_ANaOH (a form + salt) (heating rate: 10 C/min).
crystallizes under the g form. The crystal habit of particles of the batch B10ANH4 OH is that typical of the a form, with the typical needle-shaped crystals, while the particle shape of the batch B10ANH3 is irregular. The thermograms of the two batches are very close to those of the reference particles for both extrapolated onset temperature and enthalpy content. XRPD patterns are completely superimposed to those of the reference materials for both peak distances and intensity. The water content % is in agreement with the anhydrous crystalline forms. IDM was analyzed by mass spectrometry (MS) and 1H NMR in deuterated methanol at different pHs. These analyses were not suitable for establish the kind of polymorphic form or discriminate one polymorph from another, but allowed to evaluate the instability of IDM at basic pHs. IDM crystallized by both methods A and B at pHs 1 to pH 10.0 was stable, and under 1H NMR analysis all the protons were perfectly integrated (refer to Supplementary Material). At basic
pHs (12) the amidic bound was broken into the indol ring and acid 4-chlorobenzoic. 1H NMR spectrum showed the presence of acid 4-chlorobenzoic protons between 7 and 8 ppm. Moreover, it was possible to note a chemical shift of the methylene protons. The MS analysis confirmed results of 1H NMR. The spectrum of the positive ionization of products at pHs 1.0 to 10.0 highlighted the presence of IDM with a molecular weight of 357.08 (M) and the following ionization products: M + H+ = 358.1; M + Na+ = 380.0; 2M + Na+ = 737.2 (refer to Supplementary Material). The negative ionization of products at pHs 1.0–10.0 showed the formation of the following species: M H+ = 356.1; M + Cl = 392.0; 2M + H = 713.3. The positive ionization of the product at pH 12.0 showed the presence of the indol ring, with molecular weight of 218.08 (M), as M + Na+ = 242.1. The negative ionization of the product at pH 12.0 showed the peak of the acid 4chlorobenzoic, with molecular weight of 156.0 (M), present as M H+ = 154.9.
k
h g f e
k
Intensity (a.u.)
Intensity (a.u.)
j i
j i h g f e
d
5
10
15
20
25
30
c
d c
b a
b a
35
2θ Fig. 4. Comparison of indomethacin X-ray powder diffraction patterns of samples produced with the method A: (a) reference g form; (b) reference a form; (c) reference sodium trihydrate salt; (d) B1_A; (e) B4.5_A; (f) B7_ANaOH; (g) B7ANH4 OH ; (h) B8_ANaOH; (i) B8ANH4 OH ; (j) B10_ANaOH; (k) B10ANH4 OH .
5
10
15
20
25
30
35
2θ Fig. 5. Comparison of indomethacin X-ray powder diffraction patterns of samples produced with the method B: (a) reference g form; (b) reference a form; (c) reference sodium trihydrate salt; (d) B1_B; (e) B4.5_B; (f) B7_BNaOH; (g) B7BNH4 OH ; (h) B8_BNaOH; (i) B8BNH4 OH ; (j) B10_BNaOH; (k) B10BNH4 OH .
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
Summarizing, both mass spectrometry (MS) and 1H NMR studies proved that IDM degraded at pHs 12.0. These results confirm previous evidences that proved the chemical degradation of IDM in strong alkali (Hajratwala and Dawson, 1977; Archontaki, 1995). Hence, in this study, the products obtained at pHs 12.0 were not characterized from the physicochemical point of view. 4. Discussion In a previous study (Martena et al., 2012), it was shown that the crystallization of IDM under spray drying from an aqueous ethanolic solution induced the formation of a mixture prevalently composed by the a form and a certain amount of the g form as proven by the XRPD. The prevalence of the metastable a form (lower melting temperature as compared to the g form) by a spray drying process was ascribed to the fact that the two polymorphs a and g have a monotropic relationship and therefore, in a bottom up technique such as the spray drying, the generation of the metastable a polymorph is expected: the nucleation that occurs in the metastable zone and is followed by subsequent crystal growth favours the nucleation and growth of the metastable a form, which is the form of higher energy content (free Gibbs energy) and higher disorder degree (entropy) (Grant 1999). Results in the present study showed that IDM always crystallized in the a form when the method A was applied whatever the pH. Traces of the g form appeared with the method B at the pKa, and the pure g form was obtained in presence of NH4OH at pHs 7.0, 8.0, and 10.0. The trihydrate sodium salt was always formed in presence of NaOH, but it was never pure and always coexisted with the metastable a form whatever the crystallization method. Actually, the moles of NaOH added to reach the requested pHs are not enough to salify all the IDM molecules (Table 1), and thus the IDM excess crystallized under an acidic form. It is
543
interesting to note that the presence of NaOH always favoured the crystallization into the metastable a form even under the Method B that is characterized by a lower temperature of nucleation. The crystal packing of the g form has been described in depth (Kistenmacher and Marsh, 1972; Masuda et al., 2006): crystals are triclinic and each unit cell accommodates two molecules of same conformation (Z = 2). No hydrogen bonds are established between the two molecules. The crystal packing of the a form has been also described in depth (Chen et al., 2002; Masuda et al., 2006): crystals are monoclinic and each unit cells accommodates a couple of three different conformation molecules (A, B, and C) (Z = 6). These three molecules mutually form a hydrogen bond: a hydrogen-bonded carboxylic acid dimer between carbon number 11 of molecule A and B, and hydrogen bonding occurs between the carboxylic acid group of molecule C (carbon number 11) and the amide carbonyl group of molecule B (carbon number 14) (Fig. 1). At the acidic pHs (lower than pKa), IDM only exists as neutral undissociated form (R COOH) and, whatever the crystallization conditions (method A or B), it may easily establish hydrogen bonds orienting the crystallization under the formation of the a form. At the pKa (pH 4.5), indomethacin exists in equilibrium between the dissociated (R COO ) and the undissociated form (R COOH). The method A favours the crystallization under the pure a form: the metastable form is thermodynamically favoured and, as far as the crystallization proceeds, the equilibrium is displaced towards the undissociated form that favours the formation of the a form. With the method B, the a form seems still predominant, but this different crystallization condition (lower crystallization rate and temperature) may favour the appearance of some nuclei of the g form which is the thermodynamically favoured form. The IDM existing as undissociated form may easily establish hydrogen bonds and thus preferentially orientates the
Table 4 Thermal analysis of indomethacin samples. Batch
Crystalline form
Water content %a
0.23 0.11 0.47 0.13 12.45* 12.30 0.24** 0.55 0.16 0.49 0.20 0.45 0.10 0.48 0.14
Native crystals (g form) a form (reference) Sodium salt (reference)
g form a form
B1_A B1_B B4.5_A B4.5_B B7_ANaOH
a form a form a form a form + some nuclei of g form a form + salt
B7_BNaOH
a form + salt
7.88 1.05
B7ANH4 OH B7BNH4 OH B8_ANaOH
a form g form a form + salt
0.53 0.12 0.28 0.10 6.49 1.16
B8_BNaOH
a form + salt
5.24 1.08
B8ANH4 OH B8BNH4 OH B10_ANaOH
a form g form a form + salt
0.50 0.16 0.25 0.14 8.12 1.08
B10_BNaOH
a form + salt
5.27 1.35
B10ANH4 OH B10BNH4 OH B12_ANaOH
a form g form
Trihydrate sodium form
Degradation
7.20 0.89
0.52 0.09 0.22 0.08 –
Dehydration endothermb
Melting endothermb
Tm C (DH J g
Tm C
DH J g
157.67 0.13 153.95 0.32 266.25 1.24
110.63 1.72 85.66 2.24 112.96 3.44
– – 74.33 1.76 (75.06 2.89) – – – – 70.24 (25.28 69.57 (37.13 – – 71.38 (27.89 70.58 (25.21 – – 70.24 (30.12 69.25 (24.15 – – –
*Theoretical value. **Experimental value. a Determined by TGA–STA. b Determined by conventional DSC at 10 C/minDeDetermined by conventional DSC at 10 C/min.
2.28 1.14) 1.15 1.18)
1.89 2.77) 2.66 2.13)
1.08 2.68) 3.12 1.68)
1
)
153.38 153.02 153.05 153.12 157.66 150.22 265.06 151.65 264.56 153.24 157.55 150.27 265.27 149.88 265.98 153.12 157.08 149.55 265.90 148.22 266.14 153.04 157.66 –
0.56 0.75 0.24 0.85 1.86 0.27 1.25 0.78 1.84 0.55 1.56 1.18 2.35 2.23 2.76 0.65 1.78 1.22 1.67 3.55 3.27 0.77 1.24
87.28 84.22 52.88 80.25 15.13 56.92 27.12 50.29 29.30 87.37 108.27 45.37 25.31 46.28 32.71 86.24 110.63 40.25 27.40 44.22 30.11 86.18 110.63 –
1
1.67 3.55 3.55 1.26 0.28 1.37 1.44 1.47 2.45 1.18 2.45 2.36 3.28 2.17 2.44 3.18 1.72 3.19 2.79 2.57 1.12 1.27 1.72
544
A. Dubbini et al. / International Journal of Pharmaceutics 473 (2014) 536–544
crystallization under the a form. Nevertheless, the presence of a certain amount of dissociated IDM may limit the establishment of hydrogen bond thus orienting IDM in different molecule arrangement that favours the crystallization under the g form. Nevertheless, at pHs higher than the pKa and Method A, the a form is always present (sometimes together with the salt when the sodium counterion is present in the crystallization medium). At these pHs, the dissociated form (R COO ) is predominant, but the a form is still favoured. Once more, this is due to the crystallization conditions (rate and temperature) that favour the formation of the metastable form. On the contrary, method B conditions in presence of NH4OH favoured the crystallization under the stable g form, which is for that matter independent to the dissociation of the IDM. Differently, under method B in presence of NaOH the form a is still favoured: probably the presence of the counterion favoured the establishment of hydrogen bonds and probably the typical crystal packing of the a form. 5. Conclusions This study proved the effect of pH and crystallization conditions on the crystallization of indomethacin. Crystallization conditions, pH of crystallization solution, and the presence of the sodium counterion may orientate the crystallization of IDM towards the a, g polymorphic forms and/or the formation of the trihydrate sodium salt. Since polymorphs and salts significantly influences the physicochemical properties and particle morphology of indomethacin, consequences on formulation and technological behavior of the drug may be expected. Acknowledgements The authors would like to thank Mrs Laura Petetta for her contribution to SEM analysis and Mr Giuliano Camaro Blanco for his contribution to the experimental section. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ijpharm.2014. 07.030. References Allen, D.J., Kwan, K.C., 1969. Determination of degree of crystallinity in solid-solid equilibria. J. Pharm. Sci. 58, 1190–1193. Andronis, V., Zografi, G., 2000. Crystal nucleation and growth of indomethacin polymorphs from the amorphous state. J. Non-Cryst. Sol. 271, 236–248. Archontaki, H.A., 1995. Kinetic study on degradation of indomethacin in alkaline aqueous solutions by derivative ultraviolet spectrophotometry. Analyst 120, 2627–2634. Borka, L., 1974. The polymorphism of indomethacine. New modifications, their melting behavior and solubility. Acta Pharm. Suec. 11, 295–303. Chen, X., Morris, K.R., Griesser, U.J., Byrn, S.R., Stowell, J.G., 2002. Reactivity differences of indomethacin solid forms with ammonia gas. J. Am. Chem. Soc. 124, 012–15019.
Crowley, K.J., Zografi, G., 2002. Cryogenic grinding of indomethacin polymorphs and solvates: assessment of amorphous phase formation and amorphous phase physical stability. J. Pharm. Sci. 91, 492–507. Grant, D.J.W., 1999. Theory and origin of polymorphism. In: Brittain, H.G. (Ed.), Polymorphism in Pharmaceutical Solids. Marcel Dekker, New York, pp. 227–278. Hajratwala, B.R., Dawson, J.E., 1977. Kinetics of indomethacin degradation I: presence of alkali. J. Pharm. Sci. 66, 27–29. Jain, A.K., 2008. Solubilization of indomethacin using hydrotropes for aqueous injection. Eur. J. Pharm. Biopharm. 68, 701–714. Kaneniwa, N., Otsuka, M., Hayashi, T., 1985. Physicochemical characterization of indomethacin polymorphs and the transformation kinetics in ethanol. Chem. Pharm. Bull. 33, 3447–3455. Kistenmacher, T.J., Marsh, R.E., 1972. Crystal and moleculr structure of an antiinflammatory agent, indomethacin, 1-(p-chlorobenzoyl)-5-methoxy-2-methylindole-3-acetic acid. J. Am.Chem. Soc. 23, 1340–1345. Kulkarni, S., Gupta, S.P., Upmanyu, N., Tonpay, S.D., 2011. Solubility enhancement of water insoluble drug for ophthalmic formulation. Int. J. Drug Deliv. 3, 141–148. Legendre, B., Feutelais, Y., 2004. Polymorphic and thermodynamic study of indomethacin. J. Therm. Anal. Calorim. 76, 255–264. Llinàs, A., Box, K.J., Burley, J.C., Glen, R.C., Goodman, J.M., 2007. A new method for the reproducible generation of polymorphs: two forms of sulindac with very different solubilities. J. Appl. Cryst 40, 379–381. Martena, V., Censi, R., Hoti, E., Malaj, L., Di Martino, P., 2012. Indomethacin nanocrystals prepared by different laboratory scale methods: effect on crystalline form and dissolution behavior. J.Nanopart. Res. 14, 1275–1289. Martínez-Ohárriz, M.C., Rodríguez-Espinosa, C., Martín, C., Goñi, M.M., TrosIlarduya, M.C., Sánchez, M., 2002. Solid dispersions of diflunisal-PVP: polymorphic and amorphous states of the drug. Drug Dev. Ind. Pharm. 28, 717–725. Masuda, K., Tabata, S., Kono, H., Sakata, Y., Hayase, T., Yonemochi, E., Terada, K., 2006. Solid-state 13C NMR study of indomethacin polymorphism. Int. J. Pharm. 318, 146–153. Monkhouse, D.C., Lach, J.L., 1972. Use of adsorbents in enhancement of drug dissolution. J. Pharm. Sci. 61, 1435–1441. Nair, R., Gonen, S., Hoag, S.W., 2002. Influence of polyethylene glycol and povidone on the polymorphic transformation and solubility of carbamazepine. Int. J. Pharm. 240, 11–22. Singhal, D., Curatolo, W., 2004. Drug polymorphism and dosage form design: a practical perspective. Adv. Drug Deliv. Rev. 56, 335–347. Surwase, S.A., Boetker, J.P., Saville, D., Boyd, B.J., Gordon, K.C., Peltonen, L., Strachan, C.J., 2013. Indomethacin: new polymorphs of an old drug. Mol. Pharm. 10, 4472–4480. Tong, P., Zografi, G., 1999. Solid-state characteristics of amorphous sodium indomethacin relative to its free acid. Pharm. Res. 16, 1186–1192. Tong, P., Zografi, G., 2004. Effects of water vapor absorption on the physical and chemical stability of amorphous sodium indomethacin. AAPS Pharm. Sci. Tech. 5, 1–8. Vippagunta, S.R., Brittain, H.G., Grant, D.J.W., 2001. Crystalline solids. Adv. Drug Deliv. Rev. 48, 3–26. Yamamoto, H., 1968. 1-acyl-indoles. II. A new syntheses of 1-(ion-chlorobenzoyl)-5methoxy-3-indolacetic acid and its polymorphism. Chem. Pharm. Bull. 16, 17–19. Yoshinari, T., Forbes, R.T., York, P., Kawashima, Y., 2002. Moisture induced polymorphic transition of mannitol and its morphological transformation. Int. J. Pharm. 247, 69–77. Yoshinari, T., Forbes, R.T., York, P., Kawashima, Y., 2003. The improved compaction properties of mannitol after a moisture-induced polymorphic transition. Int. J. Pharm. 258, 121–131. Yu, L., Reutzel, S.M., Stephenson, G.A., 1998. Physical characterization of polymorphic drugs: an integrated characterization strategy. Pharm. Sci. Technol. Today 1, 118–127. Yu, L., 2001. Amorphous pharmaceutical solids: preparation, characterization and stabilization. Adv. Drug Deliv. Rev. 48, 27–42. Yu, L., Ng, K., 2002. Glycine crystallization during spray drying the pH effect on salt and polymorphic forms. J. Pharm. Sci. 91, 2367–2375. Zhang, G.G.Z., Law, D., Schmitt, E.A., Qiu, Y., 2004. Phase transformation considerations during process development and manufacture of solid oral dosage forms. Adv. Drug Deliv. Rev. 56, 371–390.