Polymorphism of Sulfathiazole

0 The polymorphism of sulfathiazole has been reexamined. The issue of the separate existence of two of the forms, forms 111 and IV, has now tieen definitively resolved: both forms exist. Single crystals were used to generate X-ray diffraction patterns, equivalent to those normally olbtained from polycrystalline samples, using the ingenious Gandolfi camera. The patterns obtained were distinct and in agreement with tho theoretical powder patterns calculated from the reported structures of the two forms, thus confirming the separate existence. Disagreement between the results of previous workers has been traced to two rnain causes: characterization techniques are not very discriminating for the two forms (Ill and IV), and the samples examined have invariably been mixtures of 111 and IV, since the two forms usually crystallize together. Furthermore, crystals of form 111 exhibit marked preferred orientation, which complicates powder X-ray diffraction (PXRD) analysis. The IR, Raman, and I3C NMR spectra, and PXRD patterns are reported for all four forms. These distinct fingerprints should now enable one to isolate pure samples and to study the properties and the relationships between these forms with full confidence. The polymorphism o f t h e antibacterial sulfathiazole ( N 2 thiazolylsulfanilamide) has been investigated by numerous workers. The literature,l-17 however, contains contradictory reports and there i s s t i l l confusion about the characterization of the various polymorphs which have been isolated. In particiilar, the separate existence of two of the polymorphs, forms I11 and IV, is disputed. Some o f the researchers have either implied or concluded that the two forms are in fact the same, even though crystal structures for the two forms had been ieported.IoJl Assuming that at least four forms have been crystallized, the exist ing evidence for the i r assignment and the chief characteristics are summarized in Table I. The present study reexamines the polymorphism of sulfathiazole, w i th the part icular objective o f confirming, categorically, the existence or otherwise o f forms I11 and IV. Experimental Section Crystalllization of the PolymorphsSulfathiazole of BP grade was supplied by May & Baker, Dagenham, Essex, U.K. The solvents used for crystallization of the polymorphs were all of A.R. grade. Form I was recrystallized from n-propanol at 90 "C. It was harvested by hot filtration and dried in a hot air oven at 105 "C for 15 min. Form I1 was obtained by boiling a supersaturated solution of sulfathiazole in water to dryness using a hot plate. As soon as all the water had evaporated, to avoid excessive uncontrolled heating, the crystals were removed from the hot plate and dried immediately in a hot air oven a t 105 "C for 15 min. On occasions when the transfer of the crystals to the hot air oven was not quite immediate, the crystals obtained were not of form I1 hut of form I11 instead. Form I11 was recrystallized from water. An almost saturated solution 'was obtained using a rotary evaporator and then it was cooled at a rate of -5-10 "Cih. Form IV was recrystallized from a 5050 mixture of acetone and chloroform. An almost saturated solution in acetone was cooled to room temperature and diluted with the chloroform. The solution was mixed and allowed to stand. Crystals of both forms I11 and IV were harvested by filtration and dried in a desiccator under reduced pressure (P205) at room temperature. Verification-Single Crystal X-ray Diffraction-As will become evident in the discussion, it became necessary to resort to the use of single crystals. Diffraction patterns, equivalent to those normally obtained from polycrystalline or powder samples, were generated from individual crystallites using the Gandolfi diffraction camera.'*.'$ This is an ingenious camera adapted from the basic Debye Scherrer (powder) version, but designed specifically to overcome problems of preferred orientation or insufficient numbers of microcrystals in the sample. The camera employs a double rotary motion in an attempt to ensure that all relevant crystal (hkl) planes become involved in the diffraction process with equal probability. The specimen is located on a spindle which is offset a t 45" on a rotating base. The spindle rotates on its own axis, while the rotation of the base causes it to describe a conical surface, the apex of which is the sample locus. Nickel-filtered CuKa radiation was used. Exposure times varied from 2 to 6 h. Computer Simulation of Powder X-ray Diffraction (PXRDj Patterns-The ability to be able to relate a known or proposed crystal structure to a powder X-ray diffraction pattern proved to be an important link in resolving the ambiguity between the polymorphic forms. Thus, theoretical X-ray powder diffraction patterns for forms I, 111, and IV were calculated from the reported single crystal structure data,lO.ll using the computer program Lazy Pulverix.20 This program is a general routine for calculating X-ray and neutron powder diffraction patterns from input comprising the space group, unit cell parameters, and coordinates (2, y, 2) of each atom ( j ) within one asymmetric unit. The program in effect combines Bragg's Law, the structure factor (Fhkl) equation, and various corrections to yield the difiaction peak positions and intensities (Ihkl): kyj + Izj)}exp{ -Bjsin20/h2} where rn is the multiplicity factor, and L and P represent, respectively, the Lorentz and polarization geometrical correction factors. In the structure factor equation, f, represents the atomic scattering factor and B, is the isotropic temperature factor for each atom j . Although anisotropic thermal parameters are available for these polymorphs, this refinement was not necessary for the comparisons required: the main effects of the omission are small intensity errors in the high angle reflections, which are of trivial importance to the present study. Characterization-Differential Scanning Calorimetry (DSCjThe thermal behavior of the polymorphs was examined using a Perkin-Elmer DSC2 differential scanning calorimeter. Dry nitrogen at 20 psi was used as carrier gas, the heating rate was varied from' 0.2 to 80 "Cimin, and sample mass varied between 1 and 5 mg. Pure samples of indium and alumina were used to calibrate the temperature scale and the differential energy axis, respectively, using the method of Richardson.21 Hot-Stage Microscopy-To complement the DSC studies, the thermal behavior of the polymorphs was examined directly using a Mettler FP52 hot-stage microscope. 0022-3549/89/O400-0337$01.OO/D