This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Benznidazole (N-benzyl-2-(2-nitro-1H-imidazol-1-yl)acetamide), is a nitro-heterocyclic drug used in the treatment of Chagas' disease. Despite the fact that this drug was released more than 30 years ago, little information about its solid state properties is available in the literature. The aim of this work is to investigate the thermal stability of benznidazole, providing a detailed characterization of the solid form used in formulations. In addition, it was verified that this drug exhibits three polymorphs, which were characterized by X-ray powder diffraction, thermal analysis and hot stage microscopy. The thermodynamic relationship among these polymorphs has also been established.
Keywords: benznidazole; polymorphism; hot stage microscopy; thermal analysis; X-ray powder diffraction; Raman spectroscopy; infrared spectroscopy
Chagas' disease is a zoonosis caused by Trypanosoma cruzi, which was discovered 100 years ago. It is endemic the American continent and present from the southern part of the United States of America, Chile and Argentina. Despite recent advances in the control of its vectorial and transfusional transmission, it remains the major parasitic disease burden in Latin America. It is estimated that 28 million people are at risk of contracting Chagas' disease.1,2
Many attempts to develop a treatment for Chagas' disease, have been made during the last century.3 However, the pharmacological treatment of T. cruzi infection, has varied little during the last 30 years. The currently available drugs are nifurtimox (3-methyl-N-[(1E)-(5-nitro-2-furyl) methylene] thiomorpholin-4-amine1,1-dioxide), released in 1967 by Bayer (Lampit®), and benznidazole (N-benzyl-2-(2-nitro-1H-imidazol-1-yl)acetamide), released in 1972 by Roche (Rochagan® or Radanil®). Benznidazole is better tolerated and so is favored by most experts as the first-line treatment for Chagas disease. Good results have been achieved using these drugs in the acute phase, congenital infection and laboratory accidents, but variability in efficacy may occur for chronic infections. This was because both drugs induce significant side effects and some strains of T. cruzi are resistant to treatment.
Benznidazole (BZN, Figure 1), is a nitroimidazole derivative, whose action is related to the nitroreduction of components of the parasite and the binding of metabolites to the nuclear DNA and k-DNA of T. cruzi and, the lipids, and proteins of the parasite.4,5 Despite the fact that BZN has been applied to the treatment of Chagas' disease for a long time, there is little information about its physicochemical properties. Furthermore, at the best of our knowledge, no polymorphs of BZN were previously reported. In view of the relevance of BZN for public health programs, the aim of this study and research is to provide a detailed physicochemical data about this drug and its polymorphs. Thus, this paper deals with the crystallization of BZN, as well as, the structural, thermal and spectroscopic solid state characterization.
The raw material of BZN used in the crystallization essays, was supplied by Fundação Oswaldo Cruz (FioCruz, Instituto de Tecnologia em Fármacos-FarManguinhos). The first problem in developing a crystallization protocol is the choice of a solvent with the right solubility profile. For satisfactory results, the material must be moderately soluble at room temperature and completely soluble close to the boiling point of the solvent. In our case the effect of different common solvents on BZN, has been studied (Table 1) at room and low temperatures. Following these results, BZN crystals have grown by the slow evaporation method using acetone, ethyl acetate and ethanol as solvents at 5 °C and 20 °C. The best crystals were obtained at low temperature using ethanol. BZN crystals grown with an acicular habit in a radiating manner, as shown in Figure 2.
Data collection of the selected BZN crystals were made at room temperature and 150 K. All X-ray diffraction intensities were collected using an Enraf-Nonius Kappa-CCD diffractometer (graphite monochromated MoKα X-ray beam with l = 0.71073 Å, j scans and w scans with k offsets, 95 mm CCD detector on a k-goniostat), equipped with a cold N2 gas blower cryogenic apparatus (Oxford Cryosystem). The X-ray diffraction data were processed as follows: data acquisition, indexing, integration and scaling with the COLLECT6 and the HKL Denzo-Scalepack softwares,7 no absorption correction (m = 0.104 mm-1), solving by direct methods of phase retrieval with SHELXS-978 within the WinGX,9 merge of Friedel pairs before refinement (Flack parameter was not refined), refinement by full-matrix least squares on F2 with SHELXL-978 also within the WinGX,9 constrained positions and fixed isotropic thermal parameters for C-H hydrogen atoms (bond distances of 0.950 Å and 0.989 Å for Csp2-H and Csp3-H in methine groups, respectively, Uiso(H) = 1.2Ueq(C), assignment of N8-H hydrogen atom directly from the difference Fourier map and refinement riding to the Nitrogen atom with fixed isotropic thermal parameters (Uiso(H) = 1.2Ueq(N)). Structural analysis and drawings were made using the MERCURY10 and ORTEP-311 softwares. The crystallographic information file (CIF file), loading all crystal data of BZN structure determination, were deposited with the Cambridge Structural Data Base under deposit code CCDC 812245 and 812246. Free of charge, copies of these files may be solicited from The Director, CCDC, 12 Union Road, Cambridge, CB2 1EZ, UK, fax: +44123-336-033; e-mail: email@example.com or http:www.ccdc.cam.ac.uk.
High-resolution synchrotron X-ray powder diffraction data, were collected at the X-ray Powder Diffraction beamline (D10B - XPD) of the Brazilian Synchrotron Light Laboratory (LNLS, Campinas, SP, Brazil).12 X-rays of wavelength 1.23858 Å were selected by a double-bounce Si(111) monochromator with water cooling for the first crystal, while the second one is bent for sagittal focusing.13 Conventional X-ray powder diffraction profiles of the samples were also recorded using a X-ray powder diffraction system Bruker D8 Advance working in the Bragg-Brentano geometry and equipped with a Göbel mirror and a Lynxeye detector. Patterns were collected at laboratory temperature (about 294 K), using Cu Ka radiation, operated at 40 kV and 25 mA.
Simultaneous thermogravimetric (TG) and differential scanning calorimetry (DSC) experiments were performed in a Netzsch STA 409 PC/PG equipment. A Bruker Tensor 27 Fourier Transformed infrared spectrometer coupled to the TG/DSC system was used to analyze the released gases. The measurements were carried out from room temperature up to 500 oC, at 10 oC min-1, under nitrogen flow, by using an open aluminum pan, in which approximately 5 mg of the sample was placed. In addition, DSC thermograms of heating/cooling cycles were recorded in a Netzsch Phoenix 204 system, using heating (cooling) rates, ranging from 1 to 10 K min-1 (2 to 100 K min-1). Hot stage microscopy (HTM) investigations were registered by a Leica DM2500P polarization microscope equipped with a Kofler hot stage. The microphotographs were taken with a digital camera (Leica EC 3).
Scanning electron microscopy
Scanning electron microscopy (SEM) was utilized in order to assess the morphological characteristics of the raw materials and recrystallized samples using a Tescan Model XMU, VEGA II. The samples were fixed on aluminum stubs with a double-sided tape and gold sputter-coated. They were examined in the microscope using an accelerating voltage of 30 kV at a working distance of 8 mm.
Mid- and near-infrared spectra were recorded on a Bruker Vertex 70 Fourier Transform Infrared (FT-IR) spectrometer. KBr pellets of solid samples were prepared from mixtures of KBr and the sample in 400:1 ratio using a hydraulic press. The Raman spectra were recorded with a Jobin Yvon T64000 triple spectrometer equipped with a N2-cooled CCD detector. The excitation source was the 514.5 nm line of an Argon laser operating with a power lower than 5 mW on the sample surface.
RESULTS AND DISCUSSION
The crystal structure of BZN was recently determined by Soares-Sobrinho et al. at 100 K.14 Nevertheless, in order to investigate any possible structural change with the temperature, it was re-determined at room temperature and 150 K. The main crystallographic data are presented on Table 2. These studies show small changes in the crystal structures being the molecules at both temperatures easily superimposed. Thus, we may assume that no solid-solid phase transition occurs between 100 K and 293 K.
The analysis of the asymmetric unit shows the presence of two intramolecular interactions that help the stabilization of the molecular conformation (see Table 3). One of them, C11-H11b···O18, involve the oxygen of the nitro group bonded to the imidazole group leaving it almost coplanar with the ring (Figure 1), but showing a small twist of only 5.3(3)°. The other intramolecular interaction, C7-H7b···O10, helps the planarity of the group involving the atoms C7, N8, C9, O10 and C11. The analysis of the molecular bond length indicates a resonance of the acetamide group, involving the atoms N, O and C with relative distances of O10-C9 = 1.233(3)Å, C9-N8 = 1.334(3) Å, and O18-N17 = 1.227(4)Å.
Our results show that the crystal lattice of BZN consists of infinite chains along the a-axis, stabilized by a bifurcate intermolecular interaction involving the oxygen atom of the acetamide group (Figure 3a). In addition to those interactions, the structure presents four non-classical intermolecular interactions of C-H···O and C-H···N, a types that connect neighboring chains along the crystallographic axis b forming a three-dimensional network (Figure 3b and 3c).
Figure 4 shows high-resolution synchrotron X-ray diffraction measurements, recorded from the BZN raw material. All the diffraction peaks were indexed using the crystalline structure determined by single crystal X-ray diffraction. This result clearly confirms that the samples selected for the single-crystal experiment were in the same crystalline structure than the raw material.
In order to complete the solid state characterization of BZN, which could be applied in the quality control of the raw materials and in the implementation of Process Analytical Technology (PAT), the vibrational spectrum was investigated through Raman scattering, mid- (MIR) and near-infrared (NIR) spectroscopies. Raman and MIR spectra are compared in Figure 5a.
- High-frequency region (2800-3600 cm-1): In this spectral region the stretching vibrations of NH, CH and CH2 are expected. The most prominent feature of the IR spectrum is the band observed at 3272 cm-1, which is weak in Raman spectroscopy, and can be ascribed to the stretching of the NH group of the primary amide. The remaining bands are associated with the CH stretching modes. Above 3000 cm-1, two groups of bands can be identified: the modes from the imidazole group (3100 to 3150 cm-1) and those belonging to the benzyl group (3000 to 3100 cm-1). On the other hand, the symmetric and anti-symmetric stretching modes of the CH2 are clearly observed in the Raman spectrum, between 2800 and 3000 cm-1.
- Fingerprint region (500-1700 cm-1): Some relevant features of this spectral region are the intense modes in the infrared absorbance, corresponding to the C=O stretching (1660 cm-1) and the NH (1537 cm-1) in plane bending of the primary amide. These bands are expected to be shifted, respectively, toward lower and higher frequencies, due to their participation in the intermolecular hydrogen bonds NHâˆ™âˆ™âˆ™O stabilizing the crystal structure. The Raman spectrum is dominated by the angular deformation of the CH and CH2 moieties (1360 cm-1) and the breathing mode of the benzyl ring (1158 cm-1). This ring also contributes with two modes at 1588 cm-1 and 1609 cm-1 referring to the C=C stretching. The anti-symmetric and symmetric stretching modes of the nitro group (NO2) are observed, respectively, around 1555 cm-1 and 1380 cm-1, superimposed with the d(CH) deformations. Below 1000 cm-1, the C-C stretching modes and the remaining deformations and twisting modes of the molecule, also are being expected. The wide IR band at 683 cm-1 exhibit the characteristics of an out-of-plane NH deformation from a hydrogen bonded amide group, in good agreement with the crystal structure.
- Lattice Mode Vibrations (‹200cm-1): The bands observed below 200 cm-1 are associated with deformations of the molecular skeleton together with those related to the lattice vibrations. This spectral region provides direct information about the crystal packing and lattice symmetry. However, a detailed classification in these modes is not possible without the use of quantum mechanical calculations.
- Near-infared spectrum (4000-10000 cm-1): Figure 5b shows the near-infrared spectra of BZN. This technique has been demonstrated to be very useful not only to characterize the different solid forms of an active principle, as well as to provide quantitative results, when combined with chemometric methods. The NIR spectrum is determined by overtones and combinations of the fundamental vibrations of the molecule. Therefore, the bands observed below 5000 cm-1 correspond to the combination of n(CH) vibrations and n(NH) vibrations with deformation (d(CH) and d(NH)) and other stretching (n(CC) and n(CN)) vibrations. The overtones of the n(CH) and n(NH) modes are observed around 6000 cm-1 (first overtones) as well as 8500-9000 cm-1 (second overtones). The large number of combination bands observed below 5000 cm-1 is also reflected in the very complex broad band around 7200 cm-1, which corresponds to the combination of those bands with the first overtone of the fundamental stretching vibrations. The comparison of the first overtone and fundamental n(CH) bands (inset in Figure 5b), shows that the fundamental n(CH) modes have their counterparts in the NIR spectrum with similar relative intensities, however, shifted towards lower energies due to the anharmonicity of the vibrational modes. As in the case of the fundamental vibrations, hydrogen bonds also influence the behavior of the NIR bands. Thus, vibrational modes associated with this kind of interaction become more harmonic and, as a consequence, their NIR activity is strongly reduced. This behavior can be verified in the strong n(NH) band observed as a fundamental vibration in the IR spectrum, which displays almost no contribution in the near infrared region, due to the strong NH OC hydrogen bond associated with this vibration.
The thermal stability of BZN was investigated through thermogravimetric measurements. Figure 6 exhibits a typical TG curve showing that BZN is stable up to approximately 274 °C, where the first decomposition process starts. This process is very sharp, finishing around 286 oC and being accompanied by a mass loss of 25.6%. A simultaneous TG/FT-IR analysis of the released gases phase evidence that the decomposition is characterized by the emission of CO2, H2O and NH3. Considering that just one molecule of each species is released during the decomposition of BZN, they combined correspond to approximately 30% of the BZN mass, in good agreement with the TG results. A second decomposition stage has the onset temperature around 286 °C exhibiting a smooth mass decreasing giving rise to a mass loss of 30.6% at 450 oC. The same gases than in the first process were identified by FT-IR, but, in this case, they follow the same tendency than the mass loss. At the highest temperature registered in our TG measurements, the residual mass was 48%.
Figure 6 also presents the DSC thermogram of BZN, where two well defined anomalies are clearly observed. The first one, without a counterpart in the TG measure, corresponds to a sharp endothermic peak (DT = 5.4 oC) with onset temperature at 190 oC and transformation enthalpy of -138 J/g. This process can be associated to the melting point of BZN, which was confirmed by HTM observations. The second anomaly is exothermic and correlates very well with the first decomposition process. This peak is characterized by an onset temperature of 273 oC and a transformation enthalpy of 1138 J/g. Finally, a very broad peak can be observed in the DSC data above 300 oC corresponding to the second decomposition process.
Unmasking the polymorphic forms
Since thermogravimetric measurements showed that BZN does not decompose on melting, several attempts of recrystallizing this compound from the melt were performed using HTM and DSC. In Figure 7, representative HTM images of the observed crystal habits are presented. The acicular habit of the samples crystallized by slow evaporation is well identified in Figure 7a. Since it is also a characteristic of the raw material and it seems to be the most stable form, therefore, will be labeled as form I. Seeding experiments at about 180 oC showed that the crystal growth rate of form I in the supercooled melt is very fast (Figure 7b). However, the stable form I does not nucleate spontaneously from the melt. By cooling down the melt without seeding, BZN recrystallizes with a spherulite habit (Figure 7c), hereby the form III. Furthermore, no transition from form III to form I was observed on cooling, which lets us to expect that either, in the case of enantiotropism between these forms, the transformation is kinetically hampered, or form III and form I are monotropically related. If the sample is quenched, some scattered spherulites are observed submersed in the supercooled melt, which nucleates spontaneously at temperatures above 50 oC (Figure 7d). Around 100 oC, a new phase (form II) starts to growth destroying the spherulites (Figure 7e). This process is completed above 120 oC (Figure 7f). Upon heating, form II melts around 188 oC.
DSC thermograms were recorded in successive heating/cooling cycles under rates ranging from 1 to 20 K min-1. The higher temperature was limited to 200 oC in order to avoid the decomposition of BZN. Figure 8a shows typical curves of two successive DSC thermograms, where the raw material or crystals obtained from slow evaporation were used as a starting point. The main feature of both curves is, as stated previously, an endothermic peak characteristic of the melting process. However, after the second heating, the melting of BZN is observed at a slightly lower temperature (188 oC). The cooling process, showed in Figure 8b, is characterized by one or several sharp exothermic events, which depend on the thermal history of the sample and are associated to the recrystallization of the scattered spherulites (form III) from the supercooled melted phase. A detailed analysis of the second heating reveals a weak exothermic event at 105 oC (Figure 8c). It is important to point out that after the first melting, successive thermal cycles did not recover the original form (melting point = 190 oC), even combing different heating/cooling rates. On the other hand, after leaving a melted samples several days at atmospheric conditions, the elusive form I was recovered.
The polymorphs of BZN were also identified by XRPD measurements. In Figure 9, diffractograms recorded from BZN samples as raw material, after recrystallizing from the melt and after heating the recrystallized sample at 160 oC were compared. The powder patterns plotted in the Figure 9 were systematically obtained following the previously described procedure in good agreement with the DSC and HTM results.
Thermodynamic relationships between polymorphic forms
The information provided by calorimetric measurements can be used to build energy diagrams relating to the BZN polymorphs.15-18 The relationship among the polymorphs can be established using the melting points combined with the heat of fusion (HFR), and heat of transition (HTR) rules stated by Burger and Ramberger.15 Three polymorphs (n=3) were identified giving rise to pairs, which may be related monotropically or enantiotropically. Thus, the build of semi-schematic energy/temperature diagrams will be very helpful to describe the behavior of this polymorphic compound.
From the HFR, it may be inferred that the polymorph with higher melting point (form I), must be monotropically related to the form III, since the former fusion enthalpy is higher. Despite the fact that it is not possible to measure the melting point of form III, we may conclude that there is a monotropic relationship between this polymorph and form II, since there is an exothermic transformation relating them (HTR). Therefore, one may expect that the melting point of form II is lower than the one of form III. The fusion enthalpy of the form II may be estimated using the Hess's law, as the sum of the II→III transition heat and the heat of fusion of form III. Additionally, the heat of fusion from form II should be 105 J/g. The fact it has the lower fusion heat suggests that form II is the less stable one and is monotropically related to the other two polymorphs. As a consequence, the isobaric free energy curve representing form II cannot intersect the curves corresponding to Forms I and III, as shown in Figure 10.
Grunenberg et al.16 developed a method to determine the relationship among the phase transitions based on the melting points and heats of fusion. These calculations allowed us to confirm the conclusions raised up from the energy/temperature diagrams (Figure 10) and to verify the relationship between polymorphic pairs (monotropic: Ttrs > Tfus and enantiotropic: Ttrs < Tfus). The calculations are based on the following equation:
where Ttrs is the thermodynamic transition point (in K), DfusH the heat of fusion (J/g) and Cp the specific heat (J/gK). Subscripts 1 and 2 stand for the forms of higher and lower melting points, respectively. As the specific heats of the liquid (Cp,liq) and higher melting point forms (Cp,1) are usually not available, an empiric correction Cp,liq-Cp,1 = kDfusH1 can be applied. The typical values of k are 0.001 to 0.007 K, having and average value of k~0.003 K.17 Thus, we can write equation (1) as follows:
Using this equation, the transition temperature I-III is estimated at 200 oC. Since this value is higher than the melting points of these polymorphs, we can confirm that they are monotropically related. On the other hand, neither the interconversion between form III and I nor the melting point of form III were observed. Furthermore, the relationship of the II-I pair can only be estimated from the course of the free enthalpy curves and calculations utilizing the obtained thermochemical data. Unfortunately, the melting point and heat of fusion of form II are not available to apply equation (2) to the II-III e II-I pairs. However, this equation was adapted by Henck19 in order to determine the melting point from the transition temperature:
Considering the heat of fusion of form II calculated from the Hess's law, its melting point was estimated at 194 oC. According the Burger and Ramberger rules, the melting point of form II should be the lowest one if it is monotropically related to form III. Nevertheless, the virtual melting point of form II is also slightly higher than the one of forms I and II. Notice that the calculated and measured melting points differ less than 3% of the average value. Moreover, equations (2) and (3) are very prone to errors only allowing rough estimations, which does not allow us to confirm the conclusion raised from the HTR and HFR.
A detailed physicochemical characterization of BZN was performed in this study. Thermal analysis results showed that BZN melts without decomposition at 190 oC. The room temperature and 150K structures were discussed in detail, based on single-crystal X-ray diffraction data, which allowed us to compare the predicted powder pattern with the one recorded from the raw material. The vibrational fingerprint of BZN, was obtained by Raman scattering and infrared spectroscopy (mid and near), providing valuable information to support the development of new synthesis routes of BZN, and the definition of quality control procedures to be applied in the validation of raw materials and formulated products. Three polymorphs of BZN were identified by independent experimental techniques. These crystalline forms exhibit special kinetic relationships, whose transformation pathways are illustrated in Figure 11.
Table 1. Crystallization assays of BZN.
Table 2. Crystal data and structure determination statistics for BZN.
Table 3. Classical hydrogen bonding interactions. Distances and angles are given in angstroms (Å) and degrees (°), respectively.
Figure 1. Asymmetric unit of the BZN crystal structure (form I) showing the intramolecular hydrogen bonds.
Figure 2. SEM images of the (a) BZN raw material and (b) BZN crystals grown in ethanol at 5 °C.
Figure 3. Unit cell of the BZN crystal structure (form I) showing the intermolecular interactions.
Figure 4. High-resolution synchrotron X-ray diffraction pattern of BZN (form I).
Figure 5. (a) Raman scattering, mid-infrared absorption and (b) near-infrared absorption spectra of BZN. Insets correspond to the near-infrared spectrum in the region of 5400-6900 cm-1 and also to the mid-Infrared spectrum in the region of 2700-3450 cm-1.
Figure 6. Simultaneous TG/DSC thermograms of BZN showing the FT-IR absorbance spectrum of the released gases (inset).
Figure 7. Polarized light microphotographs of BZN melt film preparations. (a) slow evaporation crystallization; (b) nucleation and growth of form I from the melt under seeding; (c) growth of the form III spherulite under quenching; (d) growth of the form III on heating (~50 oC); (e) growth of form II over the form III spherulites (100 oC); and (f) complete conversion to form II (160 oC).
Figure 8. DSC thermograms of heating/cooling sucessive cicles of BZN.
Figure 9. X-ray powder diffraction patterns of BZN polymorphs.
Figure 10. Semi-schematic energy/temperature diagram of BZN polymorphs (Tf, melting point; G, Gibbs free energy; H, enthalpy; liq, liquid phase (melt); DHf, enthalpy of fusion; Tt, transition point; DHt, transition enthalpy). Bold vertical arrows represent experimentally measured enthalpies.
Figure 11. Schematic illustration of the BZN polymorphs relationship.