The crystal structure of hexagonal barium titanate has b~en determined by X-ray methods. The unit cell, which contains six units of BaTiOs, ha: %= 5.735 A. and co= 14.05 A. In the space group C6Jmmc all the ions occur at special positions: 2 Ba at (b), 4 Ba at (f), 2 Ti at (a), 4 Ti at (f), 6 O at (h), and 12 O at (k). The structure is built up of six close-packed layers of Ba and O ions, each layer consisting of one Ba to three O ions, having the packing sequence ABCACB. The Ti ions are located in the oxygen octahvdral holes between the layers. The structure is remarkable in that two-thirds of the Tie e octahedra occur in pairs which share a face to form Ti209 co-ordination groups. The compensating distortion which occurs in the Ti209 groups increases the Ti-Ti distance by 0.33 A. while the O-0 distance in the shared face is decreased by 0.38 A. The existence of Ti809 groups in hexagonal barium titanate, an unusually stable substance under these circumstances, constitutes a notable exception to certain of Pauling's rules for complex ionic structures.

A technique has been developed for handling the extremely reactive interhalogen compounds which makes it possible to investigate their structures by x-ray diffraction at low temperatures. As the first step in a program of study of the interhalogen group the crystal structure of chlorine trifluoride has been determined at −120°C. The chlorine trifluoride molecule is planar with the point group symmetry mm. The Cl atom is bonded to one F atom at 1.621A and to two F atoms at 1.716A. The F–Cl–F bond angle is 86°59′. In the succeeding pages of this journal, a parallel investigation of the vapor phase by microwave spectroscopy is presented by Dr. D. F. Smith. The molecular configuration is identical in the two cases.

To maintain intensity measurements on a fixed scale an identical and constant spectral distribution must be integrated in either the (o or 20 scan. Both the vertical and horizontal dimensions of the detector aperture are critical for either scan. With the mechanical arrangements currently in use for upper level o~ scans both aperttwe dimensions are dependent on the mosaic spread of the crystal as well as the spectral dispersion. The optimum type of scan for a given set of conditions is that which yields the true integrated intensity plus the least amount of thermal diffuse scattering. This will be the scan which defines the smallest illuminated volume in reciprocal space. When other factors, such as anomalous dispersion, are equal, a long characteristic wavelength is desirable in that it will give rise to a smaller illuminated volume at a given point, in reciprocal space than will a short wavelength.

No abstract

No abstract

PrAlO3 undergoes a first‐order phase transformation at 205°K, and a second order transformation at 151°K. Single crystal X‐ray measurements at 293°K, 172°K, and 135°K display the Laue symmetries m, mmm, and 2/m respectively. With a specimen ground to a thin crystal plate, and mounted with minimal mechanical restraints, the 205°K transformation was single crystal to single crystal; the 151°K transformation was single crystal to twinned crystal. Both transformations were completely reversible with no sign of fatigue; the twinned crystal always transformed back to a single crystal. The true space group symmetries of the low temperature phases are masked by the effects of two types of domains on the observed intensities. The most probable space groups, expressed in unconventional orientations to preserve continuity with the ideal perovskite structure, are: at 293°K F2/c, α = 90°21′, V/8 = 53.56 Å3, at 172°K I112/m, γ = 90°, V/4 = 53.22 Å3, at 135°K I, α = γ = 90°, β = 90°40′, V/4 = 53.02 Å3. To unravel the true symmetries it was necessary to solve all three crystal structures. The first order transformation is characterized by a shear of (0k0) planes along the rhombohedral [101] direction to form an orthogonal space lattice. The AlO6 groups are rotated into new orientations and Pr atoms are subjected to displacements which are normal to the shear direction. The Pr displacements lower the symmetry from orthorhombic to monoclinic even though the space lattice is orthogonal. The second order transformation is characterized by a shear of (00l) planes along ±[100] in the orthogonal space lattice to form twinned monoclinic space lattices. The AlO6 groups do not assume new orientations. The Pr displacements are increased, again in directions normal to the shear directions. Because of the Pr displacements the symmetry is triclinic even though the space lattice is monoclinic.

The CoAl distances determined by Mrs Douglas vary somewhat about the average value discussed above. The variations might be attributed either to differences in bond numbers or to differences in the amount of d character of the cobalt orbitals. Their distribution in the crystal (especially the 180 ° angle between the strongest bonds) indicates that the second rather than ¢he first effect is mainly involved. The smallest observed distance, 2.375 A., would result from a 2/3 bond with the cobalt bond orbital having 60% d character. The possibility of unequal distribution of d character among the bond orbitals permits a variety of behavior of the transition elements that is not shown by elements such as aluminum.Additional experimental evidence about the electronic structure proposed here for Co~Al9 could be obtained by study of the magnetic properties of the substances. The above theory requires that the magnetic moment of cobalt in Co~h_l 9 be greater than in the elementary substance, whereas the theory of Raynor & Waldron requires that it be less. The crystal structure of a-monoclinie selenium has been determined by the application of the Harker-Kasper phase inequalities and two-dimensional Fourier syntheses. The a-selenium molecule is an eight-membered, puckered ring similar to the rhombie sulfur molecule. It conforms, within the limit of experimental error, to the non-crystallographic point-group symmetry ~2m; the covalent bond distance is 2.34 _+ 0.02 A., and the covalent bond angle is 105.3 _+ 2-3 °. Certain abnormally short packing distances are observed which are discussed in terms of the metallic character believed to be present in the intermolecular bonding of hexagonal selenium.

Platinblau, the anomalously blue platinum amide complex previously thought to have the composition Ptn(CH3CONH)2 • H20, has been shown by crystallographic, analytical, infrared, and nmr studies to be PtIV(CHs-CONH)2(OH)2, the corresponding blue dichloride also having been prepared. A related divalent platinum complex, Ptn[(CH3)3CCONH2]2Cl2 (I), contains trimethylacetamide in its iminol form, RC(OH)=NH, with coordination of the square-planar Pt by the =NH group. In a second compound of the formula PtIV[(CH3)3CCONH]2-Cl2 (II), one of the amide ligands is present as an amide anion, -C(=0)NH~, and the other as an iminol anion, -C(OH)=N~. Yellow II is readily converted into its blue tautomer (III), which is the trimethylacetamide analog of the original Platinblau, and contains both amide groups as amide anions. The crystallographic study was performed on a crystal of I doped with 20% of II and 10% of III.The complexes of platinum having first-row atoms as ligands exhibit a certain uniformity of color, being only infrequently red, and quite often yellow or even colorless. Such would seem to be the case almost without exception, for both the divalent and tetravalent oxidation states. In view of this, it is highly anomalous that the simple complex generally agreed to have the composition Ptn(CH3CONH)2 H20 has a deep blue color. Called Platinblau by the early German workers,1 this anomalous material was first prepared in 1909 by shaking Ag2S04 with the yellow complex PtnCl2(CH8CN)2. The blue reaction product has a monomeric molecular weight in water and was concluded to be a divalent platinum complex, since on treatment with concentrated HC1, 90% of the platinum was recovered as PtnCl42-. Actually, other blueviolet platinum complexes are known,2 but they are in all cases polymeric materials having strong metalmetal interactions.As can be seen from Figure 1, Platinblau has been assigned a variety of structures, all of which have the common feature that the platinum is divalent. The structure most generally accepted involves squareplanar platinum and chelating acetamido groups (A),3•4 whereas other postulated structures involve six-coordinate dimers (B),5 polymeric chains with bridging acetamido groups (C),6 and one novel suggestion involving diazocyclobutadiene as a ligand (D).7 Structures B and C are contradicted, of course, by the molecular weight first reported by Hofmann and Bugge.1 It was our original intention to study the electronic structure and electronic spectrum of this material after first having satisfied ourselves as to its molecular structure. This first step has now been completed, and we report the results of this study here. Inasmuch as unambiguous conclusions came rather slowly, we have (1) K. A. Hofmann and G.