A relatively inexpensive camera which was designed for taking Borrmann topographs with standard X-ray diffraction equipment is described. This camera has been used to take stereo pairs of Borrmann topographs by rotating the crystal around an axis normal to the diffraction planes. Topographs have been taken of nearly perfect copper crystals up to 0.2 cm thick using silver, molybdenum, copper, and chromium radiation, and comparisons have been made with topographs obtained with crystal monochromated radiation. Geometrical factors affecting the resolving power of the technique are briefly reviewed. In addition, the resolution inherent in the diffraction phenomenon is analyzed on the basis of the theory of anomalous transmission. Comparisons are made between calculated and observed image widths of a few dislocations.

A modification of the scanning X-ray topographic camera, is reported. The specimen and photoplate are traversed back and forth independently in two directions rather than in the same direction as in the case of Lang’s camera. The distortions of the photographs caused by geometrical arrangement can be eliminated through this construction so as to have a one-to-one correspondence. Examples of reflection photographs as well as transmission photographs are shown. Some of them are compared with those taken using Lang’s camera. The dislocation images in the reflection photographs show a good one-to-one correspondence to those in the transmission photographs. The broadening of the dislocation images in the traversing direction is discussed. The present camera is especially useful for the studies of lattice defects in the thick specimens because the reflection photographs can be easily taken.

An electron probe microanalyzer (EPMA) was applied for the deformation analysis of aluminum single crystals. The lattice distortions caused by tensile stresses were observed by Kossel patterns, which are sensitive in their change of shape to lattice distortion. The effects of lattice distortion would appear as splitting, tearing, bending, broadening, disappearance, and shift of Kossel lines. This distortion behavior can be analyzed successfully. The jnhomogeneities and anisotropy appearing on every line were explained by the crystallographic cons (deration of slip mechanisms. The lattice distortions and corresponding changes in Kossel patterns depend on the direction of elongation ; the deformation modes of those crystals which show typical fee behavior in stress-strain curves can be reasonably explained by a fragmentation model. It was also found that there are some portions in Kossel patterns where some specific Kossel lines, i.e., {200} and {111}, are very sensitive to lattice deformation.

The preparation of semiconductor materials by epitaxial deposition has rapidly gained prominence in recent years. GaAs(1-x)Px alloys of variable composition, deposited on GaAs substrates, have been prepared in an open-tube reactor using H2, AsCl3, and PCl3 with GaAs source material, The deposits ranged in composition from GaAs to GaAs0.5P0.5 and were 10-30 microns thick.

X-ray diffraction topography was employed to evaluate the structural perfection of the monocrystalline GaAs substrates and the perfection of the deposited Ga(As, P) films. A scanning-reflection technique was developed to study the concentration, type, and spatial distribution of imperfections over large areas (∼8 cm2) to depths of 25 microns. Some areas of investigation included (1) the poly cry stallinity of some deposits, (2) the effect of substrate perfection on the film quality, (3) the depth of mechanically induced damage, and (4) the structural perfection of magnesium-diffused GaAs and Ga(As, P).

A procedure for conveniently determining the orientation of noncubic crystals is presented. The methods usually employed for cubic crystals, flat-film backreflection Laue patterns interpreted with the aid of a table of interplanar angles, is not readily adaptable to noncubics. In general, tables of interplanar angles for noncubics do not exist and if the effort is expended to generate them, such a vast array of angles result that they are virtually impossible to use. It is thus necessary to use the symmetry of the pattern to identify the low-index planes. However, on flat-film geometry, due to the small angular range of the data, insufficient low-index points are present to permit orientation.

In order to alleviate this problem we have developed the necessary techniques for the interpretation of back-reflection Laue patterns employing cylindrical-film geometry. The necessary overlays and their use are presented along with some of the results obtained.

A study of the reversion process in Al-Zn-Mg alloys has been made using small-angle X-ray scattering and transmission electron microscopy techniques. The rate and mode of Guinier-Preston zone dissolutions was investigated as a function of magnesium content, prior zone radius, and reversion temperature. Results indicate that in this system the reversion process is characterized by the preferential dissolution of the smallest G-P zones present after cold aging with a corresponding decrease in the volume fraction of zones. The amount of reversion at a specific temperature is dependent on magnesium content, however, the rate of reversion is independent of magnesium content.

The substructures of a bet a-quenched and a recrystallized form of high-purity uranium were measured by a method based on statistical fluctuations in X-ray diffraction intensity. For these measurements, Warren's statistical equation for determining grain size was modified to make the equation applicable to materials with high absorption coefficients or moderate-to-large grain size ( > 20 microns) or both, since many metals fall into this category, and to allow for defocusing of the X-ray beam which occurs as a natural consequence of the experiment.

The beta-quenched uranium was found to have numerous subgrains with a range of misorientation angles that was smaller and larger than the limits of the X-ray measurements (Ω = 10−4 to 10−2 steradians). The presence of the large subgrains was corroborated by optical microscopy. The presence of very small subgrains was corroborated by transmission electron microscopy which showed 0.1- to 1-micron subgrains relatively free of dislocations bounded by dense dislocation networks, and by micro Laue diffraction patterns (30-micron beam diameter) which showed partial rings similar to a powder pattern.

The recrystallized uranium had no misorientation within the grains greater than 5.5 × 10−3 steradians. In contrast to the beta-quenched case, no subgrains were found either by transmission electron microscopy (TEM) or micro Laue diffraction patterns. The TEM micrographs showed a uniform distribution of dislocation networks. Since no other substructural elements were observed, the dislocations are believed to be the cause of the misorientation within the grains for solid angles of less than 5 × 10−3 steradiflns.

These preliminary experiments show that the statistical method may be used in conjunction with transmission electron microscopy and micro Laue diffraction for the study of substructure. The statistical method gives quantitative data on “bulk” specimens that can be given a meaningful interpretation with the aid of the other techniques.

The broadening of powder pattern peaks has been studied by three methods—Fourier analysis, integral breadth measurements, and variance of the line profiles. The results obtained from the variances are compared with those obtained from the integral breadths and Fourier coefficients.

Tungsten filings were prepared at room temperature and their powder pattern peaks were recorded with a Norelco diffractometer using filtered Cu Kα radiation. The variances, integral breadths, and Fourier coefficients were calculated with the IBM 7094 computer. The results indicate that the variance is very sensitive to the range of integration s2 − s1 = (2θ2 − 2θ1) cos θ0/λ. An error of ± 10% in this range due to the difficulty in choosing the correct background changes the values of the variance significantly and the integral breadth to a lesser extent. However, the same error does not affect the values of the Fourier coefficients.

Comparing the particle sizes and strains obtained by the three methods, it was found that the strains agreed remarkably well. The particle size calculated from the variance was smaller (DeW = 150Å) than that evaluated from the initial slope of the Fourier coefficients (De – 210Å) and from the integral breadths 2De ≃ D1 = 430Å.

In this work experiments have been undertaken to determine the type of residual stresses of plastically deformed polycrystals of aluminum, copper, and nickel which cause a shift of the X-ray interference lines. In order to get accurate stress values, the lattice strain distribution has been measured in special planes of the deformed specimens, using the sin2 ψ technique. The surface stresses were determined as a function of the macroscopic plastic strain of cylindrical specimens. With increasing strain, the residual surface stress component parallel to the direction of deformation increases. The stresses observed are compressive ones after tension and tensile ones after compression.

The changes in the residual lattice strain distributions which arise on progressive thinning of the plastically deformed specimens have been measured. From the stress values determined in each new surface layer, the stress distribution originally present in the uodestroyed specimen was calculated. Tensile stresses in the interior of the specimens are in equilibrium with compressive stresses in the surface areas. The formation of residual stresses on specimens with a work-hardened surface layer is found to be quite different from that of well-annealed samples.

The results of this work seem to show that in uniaxially deformed cylindrical specimens, residual macrostresses arise which may be a consequence of a macroscopic inhomogeneity of work-hardening between the surface layers and the interior of the polycrystals.

Deformation, introduced into niobium and tantalum specimens by wire drawing at room temperature, produced changes in the shape and position of X-ray diffraction peaks. The resultant peak profiles and locations of all available peaks were recorded using the Debye-Scherrer geometry on a modified dtffractometer with crystal monochromated Cu Kα radiation. The amount of deformation in the surface layers of both metals was found to saturate essentially after only 20% reduction in area. The measured decrease in the lattice parameters of either material was attributed to a residual surface stress; the average value for the deformed saturated state for both tantalum and niobium wires corresponded to an equivalent longitudinal tensile stress of 35 ± 5 kg/mm2. Integral breadth measurements revealed approximately equal X-ray particle sizes in the <100> and <110> directions; the minimum particle size for the micro structures of both metals was around 200 Å and occurred after the first few draws.

Present day nickel-base superalloys are hardened in part by the precipitation of a phase which has variously been identified as Ni3Al, Ni3(Al, Ti) and γ′. X-ray diffraction techniques which include precision lattice parameter measurements, intensity measurements, and phase identification are used to define the structural and chemical relationships upon which this phase is based.

These relationships are developed from the following considerations: crystal chemistry and atomic size factors which relate binary Cu3Au-type T3B phases (e.g., Ni3Al) and ternary Perovskite-type T3BC2 carbide phases (e.g., Y3AlC), the determination of the number and kind of atoms in the unit cell of Ni3Al and certain ternary phases, the crystallographic relationship between the structure of Y3C and Y3AlC, and phase relations in certain quarternary alloys.

From these considerations it is shown that the γ′ phase may best be characterized as a Peiovskite-type carbide phase having the chemical formula TsBCx. A model of the γ′ structure is presented which indicates the position of the various atomic constituents based upon whether they are T or B elements. (An atomic component is considered of the T type if it is capable of substituting for nickel in 3, of the B type if it can replace the aluminum. The essential features of this model are: T and B elements form an ordered T3B lattice of the Cu3Au type ; carbon atoms are located only in octahedral holes in the centers of the Cu3Au-type cells thereby establishing Perovskite-type T3BC3 unit cells; the effective size of T and B atoms in the T3BC3 unit cell is the same: hyperstoichiometric alloys, (ratio of B atoms to T atoms greater than one) will contain B atoms at face-centered positions in addition to a small amount of equilibrium vacant sites; in all alloys aluminum will preferentially occupy the cube corners of the unit cell ; the amount of carbon which is soluble in T3BC3; at any particular temperature is determined both by the distribution of the elements which are carbide-formers and the elements manganese, iron and cobalt. This model accounts for microstructural changes which occur in some nickel-base superalloys as a function of temperature and composition.

Silicon carbide vvas generated by pyrolysis of gas mixtures consisting of silicon tetrachloride, hydrogen, and organic vapors, such as acetone, on fine tungsten wires resistance-heated at 1500°C. Prominent two-dimensional structure was demonstrated for the 220 reflection. All other lines were of the normal threedimensional lattice type.

Elevation of less than 100° in the pyrolysis temperature eliminated the twodimensional reflection, and simultaneously changed the visible crystallite size.

Specialized techniques were used td generate the silicon carbide deposits and also to examine the structure of these deposits by X-ray diffraction to obtain lines from only the silicon carbide while ignoring the tungsten wire core. Diffraction techniques include offset collimation and vertical integration.

X-Ray diffraction measurements of MnO confirm the hypothesis that the structural temperature and the magnetic transition temperature, or Néel temperature) are the same. The usefulness of X-ray diffraction intensity data of MnO, with respect to an atomic structural refinement problem involving α-Mn powders, is discussed. Lattice constant values of MnO are listed between 100 and 310°K

Both ammonium chloride and ammonium bromide undergo a transition, with rise in temperature, from an interpenetrating simple cubic (II) to a face-centered cubic (I) lattice at 183 and 137°C, respectively, and both the low- and high-temperature forms give a complete series of solid solutions. We have determined the lattice constants of the high-temperature solids at about 250° as a function of composition, and redetermined the lattice constants of the low-temperature solids at room temperature. The solutions were made by crystallization from water, followed by stirring in contact with mother liquor for at least three weeks at room temperature. Measurements were made with a Norelco- Philips diffractometer and recorder, with Cu Ko. radiation. For the high-temperature work, a simple, inexpensive heating apparatus was developed. The only previous data reported for the high-temperature forms are the lattice constants of the pure components given by Bartlett and Langmuir.

The low-temperature solutions showed negative deviations from Vegard's rule at both ends of the concentration range and a slight positive deviation elsewhere when high-angle data were used. The high-temperature solutions showed marked positive deviations from Vegard's rule over the whole compositions range. Values for the pure components agreed reasonably well with those of Bartlett and Langmuir.

The progress of the change II → I with time was followed for some of the solutions in the neighborhood of the transition temperature in an attempt to reveal the mechanism of the process.

The thermal transformations which take place in solid methyl-substituted ammonium perchlorates have been studied using high-temperature X-ray diffraction and differential thermal analysis techniques. In the temperature range from 20°C to their decomposition temperature (above 300°C), ammonium perchlorate and tetramethyl ammonium perchlorate undergo only one enantiomorphic phase transition, namely at 240 and 340°C (with decomposition), respectively. This I—II transition is ascribed to the beginning of the free rotation of the ClO4− ions. The rotation of the cations, however, begins below room temperature. If the symmetry of the cation is lowered by having both methyl groups and hydrogens arranged around the nitrogen (as in monomethyl, dimethyl, and trimethyl ammonium perchlorates), there is an additional enantiomorphic phase transition. This I—II transformation is ascribed to the rotation of the cations which have, in the partially substituted ions, two sets of non-equivalent symmetry axes (different moments of inertia). The temperatures of transformation are discussed in terms of the space requirements for rotation. Symmetries and cell dimensions of some modifications were determined.

Inner egg shell membrane was soaked in 2 M urea solution and dried. Cu Kα X-rays directed perpendular to the membrane yielded an urtsymmetrical diffraction pattern with urea 110, 111, 102, 210 and 211 planes diffracting to the right and urea 101, 200 and 201 planes diffracting to the left. The urea appears to be hydrogenbonded diagonally between the lamellae of β-keratin at an angle of 33.5° to the 6.68 Å b-axis (fiber axis) of the parallel pleated sheet structure, shortening the half-axis lengths in the a direction from 4.73 Å to 4.41 Å. Similar results were obtained with formalinized dog mesentery. Keratinous outer egg shell membrane, formal in bed collagenous dog renal capsule, formslinized collagenous pheasant pericardium, and formalinized callogenous human arachnoid yielded symmetrically oriented urea layer lines. Formalinized dog renal fascia and pleura yielded random urea lines.

Demountable rotating anode X-ray tubes with a 7½ kW power dissipation have been built for conventional diffraction work with powder cameras and equi-inclination Weissenberg goniometers, and for use with a tetrahedral press for studying crystalline matter at ultra-high pressures. The tubes employ a highly compact cooling and sealing arrangement on the rotating anode which enables four windows to be used with the focal spot close to the specimen. A rotational speed of 1750 rpm with a focal spot size of 10 × 1 mm enables the tubes to be operated at 250–275 mA at 30 kV DC or at 150 mA, 50 kV DC.

The normal procedure for a quantitative measurement of two phases with an X-ray diffractoraeter is to scan or count the intensities of two diffraction peaks and then to calculate the ratio of the two phases using an appropriate equation. This paper will describe an improved method of quantitatively measuring two phases in a sample. The diffractometer, which was constructed by Philips Electronics Instruments, consists of two fixed scintillation counters mounted at the Seeman-Bohlin focusing positions for the two desired diffraction lines. Each counter is preceded by a lithium fluoride curved crystal monochromator. The output from each counter feeds to a separate sealer which can be arranged to gate both sealers at a preselected number of counts. At this point a tape printer prints out the counts for both sealers. These counts are then related to the proportions of each phase present by an appropriate equation. Most of the electronics are solid-state designs. This paper will describe the instrumentation and show how it can be used to measure rapidly and precisely the ratio of anatase to rutile in a titanium dioxide system. The savings in time over the conventional method can be as much as 100 to 1.

While the theoretical basis for the correction of non-linearity of detector systems is well known, methods for the determination of dead-time effects must be adapted to electron probe microanalyzer systems. Two such methods, one employing both X-ray and current measurements and the other employing simultaneous X-ray measurements on two spectrometers, are described. The effect of pulse-height shrinkage at high counting rates on the linearity of the detector system is discussed. When the proposed corrections for the dead-time of X-ray detector systems employing proportional counters are applied to the X-ray intensity measurements obtained with the electron probe microanalyzer, count rates as high as 50,000 counts/sec can be used.