PYRIDINIUM PICRATE - THE STRUCTURES OF PHASE-I AND PHASE-II - CORRECTION OF PREVIOUS REPORT FOR PHASE-I - STUDY OF THE PHASE-TRANSFORMATION

Pyridinium picrate, C5H6N+.C6H2N3O7, was reported [Kofler (1944). Z. Elektrochem. 50, 200-207] to exist in two crystalline phases, one (I) being stable below 343 K and the other (II) between 343 K and the melting point (approximately 438 K). The room-temperature structure of phase I, studied by two-dimensional methods, has been reported [Talukdar & Chaudhuri (1976). Acta Cryst. B32, 803-808]. We were led to reinvestigate the system by a number of unusual features in Kofler's description of the phase behaviour. Single crystals of phase I were grown from solution and those of phase II from the melt. We have determined the structure of both phases, including analysis of the thermal motion of the picrate ions, which was found to be appreciably larger in phase II than in phase I. The reported structure of phase I was found to be incorrect, although there were no warning signs; the error was caused by confusion between a centre and twofold screw axis in projection down [010]. The packing units in the two phases are nearly identical and consist of hydrogen-bonded cation-anion pairs. These are packed in stacks, with the ion-pairs superimposed in parallel array in phase I whereas those in phase II are antiparallel; the transition between the two phases, therefore, cannot be expected to be single crystal to single crystal, as indeed it is not. Differential scanning calorimetry (DSC) and variable-temperature powder X-ray diffraction photography show that the transition occurs at 383 K. Kofler appears to have been misled by a colour change in the phase I crystals at 343 K, which we have also observed but cannot explain. The DSC measurements give DELTAH(trans) = 6.8 kJ mol-1 and DELTAH(fus) = 31.2 kJ mol-1. The transition has proved not to be reversible under our experimental conditions; for example, phase II crystals remain unchanged after 24 h at 353 K. This suggests that the temperature at which the crystalline phases are in thermodynamic equilibrium is appreciably below 383 K; we have not been able to determine the transition temperature. The details of the structure determinations (both at 298 K) are as follows: phase I, M(r) = 308.22, lambda(Mo Kalpha) = 0.71069 angstrom, F(000) = 632, yellow laths, monoclinic, mu(Mo Kalpha) = 0.95 cm-1, P2(1/c), a = 12.122 (2), b = 3.783 (1), c = 26.621 (3) angstrom, beta = 92.56 (5)-degrees, V = 1219.6 angstrom3, Z = 4, D(m) = 1.62 (flotation at 298 K), D(x) = 1.67 g cm-3, R(int) = 0.0167 (based on 25 pairs of equivalent reflections), R(F) = 0.0436, wR = 0.0492 [based on 1645 independent reflections with F > 3sigma(F)], refinement on F; phase II, M(r) = 308.22, lambda(Mo Kalpha) = 0.71069 angstrom, F(000) = 316, yellow prisms, triclinic, mu(Mo Kalpha) = 0.90 cm-1, P1BAR, a = 10.156 (2), b = 8.984 (2), c = 7.230 (1) angstrom, alpha = 86.38 (5), beta = 80.10 (5), gamma = 89.97 (5)-degrees, V = 648.6 angstrom3, Z = 2, D(m) = 1.60 (flotation at 298 K), D(x) = 1.58 g cm-3, R(F) = 0.0716, wR = 0.0694 [based on 1478 independent reflections with F > 3sigma(F)], refinement on F. Cell dimensions have been measured as a function of temperature for both phases.