Band 213, Heft 1

An efficient grid search procedure successfully applied to the solution of three unknown molecular structures from X-ray and neutron powder diffraction data is presented.

3 T mica polytype is a trigonal polytype that can be indexed both in a hexagonal primitive and in an orthohexagonal double cell. The reflection conditions clearly show that when k = 0(mod3) (orthohexagonal indexing) the geometry of the corresponding rows is the same as that of the 1 M or 2 M 1 polytypes, differing just for the violation of the additional reflection conditions. However, while the last two polytypes require an oblique setting, the pattern of the 3 T one should be always indexed in a right-angle setting. Spiral twinning of 1 M polytype according to the commonest mica twin law [310] leads to a situation ideally indistinguishable from that of the 3 T polytype. However, the distortions from the ideal geometry of the layers building up the polytype result in the appearance of weak reflections that allow the distinction. The ambiguity of the axial choice and its solution is shown comparing the theoretical calculations with the experimental pattern of a phengite sample.

The phase transitions in CsZnPO 4 at approximately 583 K and 533 K: Pnma (I) → Pn 2 1 a (II) → P 2 1 /a (III) are analyzed by group theoretical methods and structural data published for the three polymorphic forms. An order parameter treatment of the phase transformations is presented in terms of the Landau formalism and projection operator techniques. A collective tilting of the ZnO 4 - and PO 4 -tetrahedra is taken as the primary order parameter for both transitions. In the case of the I → II transformation, the order parameter transforms according to the irreducible representation Γ 4 − of Pnma . The rotational distortion pattern in phase III can be explained by onset of an order parameter transforming according to the zone boundary representation X 2 of phase I. Other distortions (Zn-, P- and Cs-ion displacements, polarization or strain components) can couple to the primary order parameters. The microscopic distortion patterns for the possible displacements as well as the expected domains are listed.

Neutron powder diffraction data of braunite Mn 2+ Mn 3+ 6 O 8 /SiO 4 (space group I 4 1 / acd ) are presented, which have been collected with the diffractometer D1A at ILL Grenoble and E6 and E2 at BENSE (HMI Berlin) at temperatures between 300 K and 1.4 K. The chemical cell has a layer structure consisting of A sheets with two Mn 3+ positions (M2, M3) and B sheets containing Si, Mn 2+ (M1) and a third Mn 3+ position (M4). Below T N at 90 K additional reflections being caused by antiferromagnetic (af) ordering occur. The magnetic cell coincides with the chemical cell. The magnetic structure is dominated by the magnetic ordering of the A layers thus reflecting the relations of the chemical cell: Within the (001) plane of the A sheets magnetic moments of the two nonequivalent Mn 3+ -ions are ferromagnetically aligned, and neighbouring A sheets are af-coupled to each other. Magnetization experiments indicate that besides the afordered Mn ions weakly coupled spins ordering at temperatures below 2 K exist. In accordance with that, additional neutron diffraction reflections arise at T < 3.0 K, allowing the refinement of the whole structure: Within the B sheets the Mn 2+ ions are responsible for the weakly coupled spins, furthermore, the spin arrangement is not collinear. The spin directions of the Mn 3+ ions (M4) are canted, forming angles of about 120° with those of the Mn 2+ ions (M1). The overall magnetic ordering within each B sheet is antiferromagnetic thus resulting in a zero magnetic moment for these sheets. Refined magnetic saturation moments of all the manganese ions agree well to expected spin-only values.

Low temperature neutron diffraction has been performed on polycrystalline DyPtGe 2 . The compound crystallizes in space group Immm ; structural parameters have been refined for T = 30 K. A magnetic phase transition into an incommensurate antiferromagnetic structure with propagation vector [0.495, 0, 0] is observed to occur at 14(±1) K. At 5(±1) K the propagation vector becomes [0.268, 0, 0] and the amplitude sine modulated structure changes from a nearly transverse to a longitudinal type. The ordered Dy moments are 8.5 ± 0.4 μ B and 9.9 ± 0.5 μ B at 8 K and 4 K, respectively.

The crystal and molecular structure of Sn 2 P 2 Se 6 in the low temperature ferroelectric phase is determined by single crystal X-ray diffraction techniques. These results are compared with the crystal structure of the same compound in the paraelectric phase, which was redetermined at room temperature. In both phases, the structure crystallizes in the monoclinic system. The ferroelectric phase with a = 6.8145(3) Å, b = 7.7170(3) Å, c = 11.694(1) Å, β = 124.549(4)°, space group Pc , V = 506.52(6) Å 3 , Z = 2 was determined at T = 173 K and the paraelectric phase with a = 6.808(2) Å, b = 7.682(3) Å, c = 11.667(7) Å, β = 124.75(6)°, space group P 2 1 / c , V = 501.4(5) Å 3 , Z = 2 was determined at T = 293 K. Both structures were solved by automated Patterson methods and refined by least squares methods. For the ferroelectric phase final refinement resulted in R = 0.038 for 3763 reflections (with I < 2 σ ( I )). Refinement of the paraelectric modification revealed disorder of the Sn 2+ cations (the two Sn sites being separated by approximately 0.34 Å) and resulted in R = 0.062 for 2211 reflections (with I &lt; 2 σ ( I )). Comparison of the two structures showed that the tin ions shift to positions of about 0.13 Å from an individual disorder-site in the high temperature phase (paraelectric) to the corresponding tin position in the low temperature phase (ferroelectric). The shift from the average Sn-position in the paraelectric phase to the Sn-positions in the ferroelectric phase is about 0.30 Å (on average 10° off the vector a + c ), and is clearly related to the spontaneous polarization. Moreover, the average direction of these displacements is perpendicular to the modulation wave vector direction in the incommensurate phase, showing the prime importance of such movements to the incommensurate phase formation.

The three Tl(I)-arsenates(V) Tl 3 AsO 4 ) [ a = 8.517(3) Å, c = 5.233(2) Å, space group P 6 3 , Z = 2], Tl 2 Cu(I)(AsO 4 ) [ a = 6.000(3) Å, = 5.343(2) Å, c = 8.616(3) Å, β = 98.05(3)°, space group P 2 1 , Z = 2], and TlCu(II) 2 (AsO 4 )(AsO 3 OH) [ a = 6.407(2) Å, b = 11.245(4) Å, c = 10.173(3) Å, β = 98.64(2)°], space group P 2 1 / c , Z = 4] have been synthesized under hydrothermal conditions. Their structures were determined from single crystal X-ray diffraction data.

Powder samples of the system Sr 3− x Ca x Al 2 O 6 ( x = 0, 0.3, …, 2.7, 3.0) were prepared by means of solid state reactions at 1723 K. The lattice constants determined exhibit a linear dependence from the molar fraction x in the range 0.9 ≤ x ≤ 3. For smaller values of x distinct deviations from this linearity are observed. As shown by the results of three single crystal investigations which could be carried out on crystals with x = 1.395(8), 1.62(1), and 2.12(1) this is due to a remarkable distribution of the earth alkaline metal atoms, which is far from being only statistical over the six crystallographically different sites.

The cations of the diazapropellanium salt 1 ([C 24 H 32 N 2 O 2+ · 2I − · H 2 O]) and the triazapropellanium salt 2 ([C 31 H 39 N 3 2+ · 2I − · 0.5(CH 3 CN) · 0.72 H 2 O]) have analogous configurations and conformations. Compound 2 decomposes in the Mo K α X-ray beam. Both compounds contain the known (but rather rare) I − …H 2 O…I − anion; previously reported manifestations of such interactions are classified in detail.

The crystal and molecular structures of three R 2 Sn(O 2 CR 1 )Cl compounds [R = Me, t Bu and Ph; − O 2 CR 1 = o -(2-hydroxy-5-methylphenylazo)benzoate] have been determined at room temperature. Brown crystals of Me 2 Sn(O 2 CR 1 )Cl ( 1 ), C 16 H 17 ClN 2 O 3 Sn, are triclinic, space group P [unk] with unit cell dimensions a = 10.649(4) Å, b = 12.292(3) Å, c = 7.236(2) Å, α = 96.68(3)°, β = 104.03(3)°, γ = 109.12(3)°, Z = 2 and D x = 1.662 Mg m −3 . Red crystals of t Bu 2 Sn(O 2 CR 1 )Cl ( 2 ), C 22 H 29 ClN 2 O 3 Sn, are monoclinic space group P 2 1 / n with unit cell dimensions a = 8.323(5) Å, b = 24.990(5) Å, c = 12.403(5) Å, β = 107.94(3)°, Z = 4 and D x = 1.417 Mg m −3 . Red crystals of Ph 2 Sn(O 2 CR 1 )Cl ( 3 ), C 26 H 21 ClN 2 O 3 Sn, are also monoclinic with space group P 2 1 / c and unit cell dimensions a = 14.416(3) Å, b = 16.373(6) Å, c = 10.489(4) Å, β = 97.10(3)°, Z = 4 and D x = 1.524 Mg m −3 The structures were solved by Patterson ( 1 and 2 ) or direct methods ( 3 ) and each refined by a full-matrix least-squares procedure to final R = 0.039 using 2933 reflections for 1 , R = 0.039 for 2935 reflections for 2 , and R = 0.037 using 2863 reflections for 3 . The structure of 1 is dimeric with the two tin atoms being bridged by two − O 2 CR 1 anions each of which coordinates a tin atom via one of the carboxylate oxygen atom and the second tin atom via the phenoxide oxygen atom; this result indicates that the − O 2 CR 1 anion exists as a Zwitterion. The tin atom geometry is trigonal bipyramidal. trans -O 2 SnC 2 Cl. By contrast to the dimeric structure for 1 , the structures of 2 and 3 are monomeric with the tin atoms in cis -O 2 SnC 2 Cl trigonal bipyramidal geometries. The − O 2 CR 1 anion coordinates the tin atom in each of 2 and 3 via the carboxylate oxygen atoms only.

The crystal structure of N -(4-methoxycarbonylphenyl)oxycarbonylaziridino-[2′,3′: 1,2]-[60] fullerene was determined. The solution of the structure failed by direct methods, but succeeded by means of the program system DIRDIF-96. The asymmetric unit contains two fullerene molecules and one benzene molecule. The [1 + 2] cycloaddition leads to the formation of elongated bonds between the bridged C atoms. The average values of the [5, 6] and [6, 6] bond lengths in the C 60 spheres [1.448(6) Å and 1.385(6) Å] are close to those determined by electron diffraction [1.458(6) Å and 1.40(1) Å]. Due to the binding of the aziridino addend, the fullerene cage becomes slightly distorted relative to a free C 60 . The bridged atoms ‘stand out’ of the C 60 surface by approximately 0.2 Å. The shortest stacking distance for the fullerene spheres is the a cell translation. We can consider the parallel stacks, formed by the two independent molecules, as a layer that contains also most of the substituents atoms.