Solid specimens of silver were charged at different temperatures in atmospheres of oxygen and of hydrogen. X-ray diffraction line profiles were obtained using powders filed from the treated specimens. Fourier analysis of the diffraction lines was conducted following the method of Warren. The effective particle sizes and root-mean-square strains were obtained from the line shape analysis. Stacking fault and twin fault probabilities were determined from peak-position and center-of gravity displacements, respectively. For the purpose of comparison, two vacuum-melted silver samples and two different silver-tin alloys were studied. The stacking fault and twin fault probabilities were observed to be nearly unaffected by charging in either oxygen or hydrogen. These results are consistent with recent direct determinations of the effect of oxygen on the stacking fault energy of silver. In contrast, the rms strains and particle sizes changed significantly after charging in oxygen. A decrease in the root-mean-square strain and a corresponding increase in the particle size was found. These effects were opposite to those obtained by adding substitutional solute to silver. These findings are interpreted to indicate the effect of oxygen clustered with impurities on the dislocations and stacking faults in silver.

The broadening of the (110)α and (220)α powder pattern peaks of bulk samples from the product of two commercial 70 ton heats of a 9Ni-4Co-0.45C alloy (HP 9-4-45) was studied by Fourier analysis. Both heats were heat treated to a variety of bainitic and tempered martensitic structures and the effective particle size and r.m.s. strain in the <110> direction were measured. Comparison of the mechanical properties with the parameters derived from the Fourier analysis suggests that neither the strain nor the particle size are controlling the strength or toughness. Strength increases with Strain while toughness decreases - both as expected. On the other hand, large particle size seems to favor greater toughness in contrast to the known effect of macroscopic (high angle boundary) grain size.

It is concluded that line shape analysis is a convenient method of investigation giving repeatable parameters which represent the transformation strains. The extension of this technique to these highly strained, metallurgically complex structures appears justified.

A precise method for detenaining unit cell dimensions of semiconductor single crystals has been developed. An accuracy of ± 0.00002 Å and a precision of ± 0.00001 Å is achieved.

The method is used to determine variations in the cell parameter of GaAs as a function of surface defects, homogeneity across and along an ingot, evaluation of epitaxial layers, effects of heat treatments and structure defects induced by oxide diffusion masks on a substrate. The application of the technique to GaAsP alloys permit accurate and rapid study of sample composition and homogeneity.

A procedure for extracting the interference distribution from the X-ray diffraction line using measured instrumental and wavelength distributions has been applied to the study of residual stress in copper following extension.

In the direct unfolding method the principal instrumental distributions due to the source and to specimen absorption are unfolded by the Fourier method. The remaining instrumental aberrations are extracted using the centroid displacements. The resultant scattering distribution is transformed by the relations of Mitchell and de Wolff and the spectral distribution unfolded. Conditions limiting the Fourier coefficients are applied to preserve stability. Measurements have been carried out with a diffractometer equipped with a diffracted beam monochromator.

The 420 and 331 interference distributions in copper following 10% extension have been compared with those of stress - free material. The distributions are symmetric both in the annealed specimen and after extension. In extension the distributions are displaced by an amount equivalent to a surface compressive stress of 6 kg/mm2, The symmetry of the interference distributions indicates that the displacement is due to a macrostress distribution.

The development of deformation textures of both initially disordered and ordered Cu3Au has been studied and discussed in terms of the stacking fault energies and twinning probabilities. No differences in the deforir.ation textures of the ordered and disordered materials were detected at deformations up to 40% reduction. However, at reductions of approximately 42% a marked difference was observed. This difference is attributed both to the degree of order and to the stacking fault energy of the ordered sample being conducive to deformation twinning at this stage of the rolling schedule. For samples which were reordered below the racrystallization temperature at intermediate stages of deformation (every 10% in total reduction) no differences between the ordered and initially disordered samples were observed up to 90% total reduction in thickness.

The texture of sheet metal Is best described, by means of pole figures, which are very expensive and time-consuming to prepare. About 8 to 12 hours of effort by a specially trained, and. highly skilled technician are needed to prepare each pole figure. Accordingly, pole figures are not used as extensively in research studies as they would, be if they could be obtained more easily.

A method has been developed for automatically producing pole figures by printing results directly from a digital computer. This method does not require the use of additional plotting attachments and, is therefore less expensive and time consuming than other methods. With this method, any laboratory with access to a digital computer can produce pole figures automatically.

X-ray diffraction intensities are recorded on punched tape or on punched cards and are fed into the digital computer. A computer program corrects X-ray data obtained, by either transmission or reflection X-ray techniques, maps the stereographic projection, and prints pole figures directly. The time required, to prepare an accurate pole figure is reduced from 8 to 12 hours to 20 minutes or less depending on the type of digital computer used.

The Picker Four-Angle Computer System (FACS-1), a computercontrolled x-ray diffractometer originally designed for single crystal studies, has been adapted for use with polycrystalline samples. The system is controlled by a PDP-8S, a small time-sharing computer with teletype input and output. Programs have been written to take advantage of the high degree of flexibility inherent in online computer control. Four basic operations are possible: (a) simple 2θ step-scanning with variable step width; (b) 2θ stepscanning with randomization of orientation; (c) determination of Legendre expansion coefficients for oriented specimens; and (d) determination of pole figures. In operation (a), data is gathered at a series of 2θ values at a prefixed count and/or time. In (b), the sample is rotated to average out orientation, giving the “randomized” intensity (2θ) at various 2θ values. The on-line computer reads the scaler and timer every two degrees of x rotation and forms the appropriate integrals for calculating (2θ) as the sample rotates. Operation (c) is an extension of (b): not only is (2θ) determined, but also various moments of the orientation distribution of the form , where Pn is the nth order Legendre polynomial. Operation (d) may be used to measure pole figures of sheet specimens in reflection or transmission, or of fibers or small particles. Optional modes of operation allow for (a) use of the Ross “balanced filter” technique; (b) integration across diffraction peaks by continuous scanning in 2θ, with background correction; and (c) application of absorption corrections.

A significant problem of the National Bureau of Standards Standard Reference Materials program is the provision of standards suitable in homogeneity for use with microanalytical techniques such as the spark source mass spectrograph and the electron probe microanalyser. An interim approach to the problem has been the extended homogeneity characterization of selected existing standards. This paper describes the preparation and evaluation of the first NBS standard tested specifically from the beginning for application to electron probe microanalysers. The standard designated SRM 480 is a tungsten-20 weight percent molybdenum alloy prepared by a powder metallurgy process. Based on the results of about 1500 determinations for both tungsten and molybdenum by electron probe microanalysis, the material was found to be of high homogeneity at about the micrometer leve 1 of spatial re so lution. The coefficient of variation for molybdenum was 2.5% and that for tungsten 1.5%. Correction of relative intensity ratios to obtain concentrations is discussed in terms of input parameter uncertainties such as mass absorption coefficients, and electron backscatter factors. The result of studies for atomic number correction and effects on operating voltage on the microprobe absorption correction, will be given. It is concluded that SRM 480 should be a valuable addition to any microprobe laboratory doing quantitative analyses.

The use of beta excited x-ray sources in solution absorptiometry has been reported upon by this laboratory. These sources produce characteristic K x-ray lines superimposed upon a bremsstrahlung continuum. The work previously reported made use of the entire beta excited spectrum. The resulting plots of In (Io/I) versus concentration of solute were not straight lines, (I0 is the intensity of the x-rays after passing through the solvent, and I is the intensity after passing through both the solvent and solute), but curved downward, as is characteristic of polychromatic sources due to preferential absorption of the low energy photons. These sources have a bremsstrahlung end-point equal to the beta energy E-max. This continuum energy usually is relatively high, resulting in a high mean x-ray energy for the polychromatic beam. This is not a desirable condition. Ideal sources for absorptiometry would be monochromatic in nature with an overall energy range covering the values for the elements. The use of a monochromatic source has two advantages for solution absorptiometry. First, the sensitivity is much greater since the K peak is of lower energy than the bremsstrahlung continuum. Secondly, the plot of In (Io/I) versus solute concentration is a straight line since a fixed value of the mass absorption coefficient can be assigned to this fixed energy, thus simplifying calibration. To use only the monochromatic K energy from these beta excited sources, we have applied the method of balanced filters. These monochromatic K x-rays, with the remainder of the continuum filtered out, have been employed in solution absorptiometry. The resulting plots of In (Io/I) versus solute concentration are straight lines with very good sensitivity. Using this technique, a very wide range of monochromatic energies are available for use in solution absorptiometry.

The fluorescence x-ray method of the coating thickness measurement is theoretically studied. The theoretical formulas of the fluorescence x-ray intensity from the coating and from the substrate are obtained which include the enhancement effect between the coating and the substrate. The formulas are applied to the following examples and are confirmed experimentally: zinc on steel; tin on steel; aluminum on steel and on PMMA. The experimental results agree well with the theoretical results.

Characteristic soft x-ray emission lines created as a result of low energy (10-100keV) positive ion excitation of a target material has been applied to the study of thin films of oxygen that form naturally on metal surfaces. The specific metals investigated were aluminum and nickel aa well as a number of their corresponding alloys.

Low energy positive ions are well suited as projectiles for the study of thin films in that they have low penetrating powers, yet produce high x-ray yields for K-, L-, and M-shell elemental x-ray lines. Perhaps the most important feature is the fact that no background bremsstrahlung continuum x-radiation is generated in the slowing processes of the ions. Exploiting the advantages of these facts, a special high sensitivity detection technique has been developed for the soft x-ray analysis of the constituents that make up a particular thin surface film.

The physical basis for thickness and atomic density measurements of thin surface films by positive ion excitation are discussed. Calibration data are presented for thin films of known thickness and composition establishing the sensitivity of the method for the measurement of the elemental atomic densities of thin films down to a few monolayers or approximately 10 Å in thickness. Finally, the results of the analysis of oxide surface films of aluminum, nickel and some of their alloys are presented and discussed.

The objective of this continuing program is to gain an understanding of the mechanisms and principles governing the formation of oxide surface films on metals with the basic goal of producing new alloying techniques for metal passification in a saline environment.

Research is being conducted on low energy x-ray and electron physics and application in the 100-1000 eV region. Work on the optimization of fluorescence analysis for light elements, including the development of efficient dispersive systems as the multilayer analysers of the Langmuir-Blodgett type, has been continued. Along with the ultrasoft x-ray program has been one on photoelectron and Auger electron spectroscopy. An evaluation of the application of low energy electron spectroscopy for surface characterization is being made.

A novel x-ray spectrograph for the analysis of light elements has been developed based on previous computations and confirming experiments by one of as (H. K. Herglotz). The major components of the instrument are an efficient fluorescent source, a totally reflecting mirror, and an open window photomultiplier. Identification of wavelengths in the range 15 < λ < 80 Å is achieved by the wavelength dependence of the critical angle of reflection of an x-ray beam incident on a suitably chosen low absorption reflector. As the incident angle is increased through the critical angle for a particular wavelength, the reflected beam intensity is sharply reduced; hence, a periodic vibration of the incident beam through a small angular range about the critical angle furnishes a strong a.c. reflected signal characteristic of one narrow wavelength band only.

Initial results promise a simple, easy-to-operate instrument for the routine analysis of elements boron to fluorine.

X-ray emission spectra have been obtained with a single crystal spectrometer having a resolution λ/dλ of 6,640 for iron Kβ5 spectra. The spectrometer consisted of a Jarrel-Ash Microfocus unit in conjunction with a Picker diffractometer which was operated as a rocking crystal spectrometer with a stepping angle of 1/400th of a degree. Kβ5 and Kβ1,3 spectra have been obtained from iron, Fe0.9Mn0.9Ge, Fe1.60Ge, FeGe, FeGe2 and germanium. Because the K05 spectra are due to 4p - Is transitions from the 3d bands, the Kβ5 intensities are a direct measure of the overlap or hybridization of the 3d - 4s - 4p bands. The experimental results can be correlated either with orbitals derived on the basis of coordination, or more roughly with possible band distributions.

A simple two-crystal spectrometer was constructed, making use of a commercial one-crystal spectrometer. A detector attached to a 2θ-arm has been replaced by a second crystal. After adjustment, the second crystal is fixed to the 2θ-arm. The detector is placed at an appropriate position apart from the 2θ-arm. The spectrometer is operated as if it were a one-crystal spectrometer. The dispersion of the spectrometer in this mode of operation is the sum of those of the first and second crystals. Ge(220) crystals gave the best resolution which was about 3-3 eV fwhm for Cu Kα1. The strongest Intensity was observed using etched LIF crystals. The intensity was about 1/4 of that of a one-crystal spectrometer and the resolution was about 9.5 eV fwhm for Cu Kα1. The spectrometer was applied to the determination of hafnium in zirconium and sircaloy. The Zr Kα 2nd order line, which coincides with the Hf Lα1 line, was completely eliminated by successive reflections by Ge(111) crystals owing to the extinction rule and the high resolving power. The detection limit was about 7 ppm of hafnium. Another example is the separation of the As Kα2 line from the Pb Lα1 line. Arsenic in low alloy steel was determined without the correction for the well known Interference of lead. The very common interference of Cr Kβ on Mn Kα was also eliminated completely by this spectrometer. Though the spectrometer is not very suitable for a wide wavelength range spectroscopic measurement, owing to its narrow window region (about ±1°), the chemical effects on Kg spectra of sulfur in various compounds were investigated. Fine structures were found In the sulfur Kβ spectra of several thiocyanates and elemental sulfur. To observe long wavelength region spectra, the X-ray path of the spectrometer was modified to a hybrid path (vacuum and helium).

Brieketted rock powder discs provide a convenient type of specimen for use in x-ray fluorescent analysis for both major and trace elements. The method is limited because variation of matrix makes calibration difficult over more than a limited range of specimen compositions.

Standard grinding and preparation procedures together with a ratio technique of measurement eliminate variation except that arising from chemical, mineralogical and particle size differences. Several series of standard rock samples have been prepared and the variation of measured intensity per unit concentration have been investigated for the analytical Lines of silicon, aluminium, iron, selenium and cadmium. Computer methods of multivatiate regression analysis have been used to calculate empirical correction factors based on this data and the mineralogical and chemical composition of the specimens. These correction factors have been used to correct the measured intensities of other specimens measured in separate experiments.

The double dilution method has many important advantages. For any element to be determined, let us say A, It enables us to control or calculate the matrix factor (sum of the absorption end enhancement effects) for the sample being Investigated towards A radiation, and it furnishes corrected Intensities which are strictly proportional to A concentration. Thus the results are exact, whatever the general composition of the sample, their accuracy depending only on the quality of measurement and preparation. Another major practical advantage is that the method does not require systematic calibration but only a few permanent standards consisting of a pure compound or of an accurately known sample.

The procedure has been tested successfully for accurate determination of rare earth elements using, for solid materials such as ores and oxide mixtures, the borax fusion technique. It also can be readily applied to liquids. All the rare earth elements can be titrated by that method, as well as yttrium, thorium and, if necessary, all the elements relevant to X-ray fluorescence analysis. The concentration range considered for solids is of one comprised between 0.5 and 100 % and, with a lesser accuracy, between 0.1 and 0-5 % Examples are given relative to the analysis of various ores. Finally it rcust be pointed out that the method is universal and applies to the analysis of every solid, especially ores, provided that they can be converted to solid or liquid solutions. It appears that most industrial analyses can be worked on In this way.

Consideration of variations in atomic potentials permits an estimation of the mass absorption coefficients of many elements for the characteristic emission lines of carbon, nitrogen and oxygen. The accuracy of (μ/p) values presented is better than 5% in most cases. These values are tested in the microanalysis of defect titanium carbides. Limitations of the theoretical correction procedures lead to the definition, of a set of analysis conditions which permit carbon analysis to within 1.0% absolute.

Solid solutions of cadmium sulf ide and cadmium selenide prepared by several methods at temperature in excess of 1000°C in inert atmosphere are described. The compositions of these solid solutions varied widely, ranging from pure cadmium sulfide to pure cadmium selenide.

Films of solid solutions have been successfully prepared by the vacuum deposition on various substrates using a procedure previously reported. Conditions for the depositions have been investigated to prevent noticeable alteration of the composition of the solid solution dur ing vacuum deposit ion. Films of various thicknesses have been prepared to date.

Determination of film thicknesses have been carried out by the fluorescence method based on the elements existing in these solid solutions. Chemical analysis in which specimens were consumed have also been used primarily for the construction of calibration curves. X-ray fluorescence intensity versus film thickness are established for films of 0, 25, 50, 77, and 100 percent by weight of cadmium selenide up to 10.00 microns thick on aluminum substrates. Thickness determinations using selenium Kα and cadmium Kα spectra give accuracies of 0.200 ± 0.05 to 10.00 ± 0.40 microns depending on film thickness and composition. Correction factors are given for films on substrates other than aluminum.

Procedures have been established for the determinat ion of both composition and film thickness based on fluorescence intensity data. There is a relationship between the fluorescence intensity and appearance when deposited films are not of the same thickness. Correlation of these phenomena will be demonstrated.

Argon is commonly used as the sputtering medium for RF sputtering of insulators and is entrapped in the deposits. X-ray emission determination of argon in RF sputtered SiO2 was required as part of a study of the relationships between argon concentration in the deposits and their electrical and physical properties.

Concentrations ranging from 0.05 to 7.4 weight % argon were measured in deposits 0.5 to 5μ thick. Two techniques were used for standardization: (1) weight loss of deposits heated for several hours in a helium atmosphere at 600°C; (2) potassium Kα and chlorine Kα measurements on a KCl film of known thickness to infer argon mass/argon Kα net counts. Calibrations made using these procedures agreed to within 10% and are reliable to about ±25% on an absolute basis. Absorption of radiation by the deposits was taken into account and used to correct measured argon intensities for absorption.

Sputtering parameters which had major effects on argon concentration were the substrate temperature and the magnetic field applied during sputtering. Argon pressure and RF power were found to have lesser effects.