A time-sharing computer was used in conjunction with a semiconductor detector and radioisotopic sources to evaluate energy dispersion x-ray analysis for the rapid analysis of copper base alloys. The time-sharing computer served two purposes: to aid in the development of the analytical method and to simulate the action of a dedicated computer that would eventually be used as part of an on-line analytical device. Some of the general capabilities of time-sharing computers are discussed in this paper with some emphasis on the differences between the computer requirements of energy dispersive x-ray methods and conventional x-ray spectrography.

An annular source of americium 24l was used with a secondary fluorescer to excite 23 brass and bronze alloys. The 400 channel spectrum from a multichannel analyzer was recorded on paper tape. A telephone coupled teletype was used to transmit the X-ray spectrum to the central processing facility. Curve fitting and statistical analysis programs similar to those available from most commercial time-sharing computer operations were utilised as well as other programs written at this laboratory for processing digitized spectra.

The electron beam of the electron microscope is an excellent source of energy for producing localized observable reactions in thin films.

The normal operating beam of a double condenser lens System focussed for studying a metal film in good thermal contact with its mount gives a calculated temperature rise of about 20° to 50°C. However, by changing these specified conditions, the sample may be heated to temperatures well over 1000°C. Copper and gadolinium--melting points 1083°C and 1312°C respectively--have been melted in the electron microscope beam, as have BeO whiskers--melting point 2530° + 30° C--by the authors and their colleagues.

In order to develop an optimum system for flash x-ray diffraction, consideration must be given to both optimum x-ray generation and optimum x-ray detection in the correct wavelength regime suitable for diffraction. Historically, most workers have concentrated their efforts in either the generation area or detection area, but not both. As early as 1942, experimental recording of Laue diffraction patterns was reported using a pulsed x-ray generator and exposure times of milliseconds. Recently, successful x-ray diffraction experiments have been reported with exposure times less than 100 nanoseconds.

The purpose of the present paper is to trace the development of generation and detection systems for flash x-ray diffraction and to summarize the present state-of-the-art for such systems. A comparative evaluation is presented for flash x-ray diffraction systems using generators which rely on increased electron beam current and those which rely on higher potential difference. Comparison is also made between detection systems incorporating film recording, scintillators fiber-optically coupled to photomultiplier tubes, and image-intensifier systems both lens and fiber-optically coupled to fluorescent screens.

A detailed description of the most rapid flash x-ray diffraction system developed to date is given. This system uses a Field Emission Fexitron single channel 300 kilovolt pulsed x-ray generator incorporating an x-ray tube with a beryllium output window. A fluorescent screen converts the x-ray diffraction image into a visible one and this visible image is focused on the first stage photocathode of an image intensifier tube either by direct fiber-optic coupling or by using a coupling lens. The image intensifier tube used is a cascaded three-stage electrostatic focus type with fiber-optic input and output faceplates and inter-stage couplers. Using this system Laue transmission diffraction patterns of single crystals and powder patterns of polycrystalline aggregates have been obtained with exposure times of 30 nanoseconds.

As a result of interest stemming from shock wave studies carried out at Lawrence Livermore Laboratory, we have developed a capability to conduct x-ray diffraction studies in submicrosecond time intervals. This involves the use of a low impedance flash x-ray device. While there are many applications to which these techniques can be put, our first experiments deal with samples undergoing shock wave compression. These particular experiments are conducted by synchronizing a 40 to 50 nsec flash x-ray device to a shock front which is produced by the detonation of a high explosive placed in contact with a sample. Diffracted radiation is usually recorded on very sensitive film protected by a blast cassette. Thus far we have subjected lithium fluoride, aluminum, and carbon to pressures in the range of 100 to 300 kbar. Either powder or single crystal samples can be used. The principal difficulties of this experiment are the lack of sufficient intensity and the synchronization of the x-ray pulse to the shock front.

Three novel methods of x-ray spectrometry have been developed in recent years at Texas Tech University. These are:

• 1. Three crystal spectrometer

• 2. Two curved crystal spectrometer

• 3. Spherically bent crystal spectrometer.

In this paper the new design features, and experimental results will be discussed to indicate the usefulness of the new instruments. The three crystal spectrometer is a modified two crystal instrument. A third crystal is used to analyze the output of the two crystal spectrometer. The first two crystals are operated as a standard two crystal spectrometer. The third crystal is swept through the spectrum transmitted by the first two crystals for each setting of the first two crystals. The peak intensity of the third crystal sweep corresponds to the energy setting of the two crystal spectrometer, and is the intensity used to plot the spectral lines. The two curved crystal spectrometer utilizes two transmission spectrographs with radii having a 2:1 ratio in series, the crystal with the smaller radius being set so that its focal point falls on the Rowland circle of the larger radius crystal, This instrument has a very low background intensity and is suitable for precision scattering and diffraetion work. The spherically bent crystal spectrometer makes use of high light gathering power and high orders of reflection to allow high resolution studies of weak spectral lines. It also has the advantage of ease of alignment and operation.

Energy yields in the generafion of monochromatic X-rays by electron bombardment are poor and the utilization of X-rays is even more wasteful, particularly in X-ray diffraction. Only a small stereoangle of radiation issuing from a flat target passes through the collimator of a camera.

An end-window X-ray tube using a conical cavity as a target is more economical, since there is a finite probability of several excitation acts by one electron in the cavity. A major part of the radiation emerges from the tube in a stereoangle of ≈ 0.02 instead of 2π steradian. This and the proximity of the tube focus to the sample are favorable for X-ray diffraction, as will be demonstrated in several examples. The geometry and performance of the tube are also expected to provide a superior replacement for the weak isotopic sources normally used in energy dispersive analysis.

Several versions of the tube, covered by U.S. Patent 3,584,219, H. K. Herglotz and C. D. Reilly, are described. Of particular interesf for X-ray diffraction is a double cavity tube with 2 exit ports. The Iow power consumption of the tubes (35 kV, 1 mA or less) requires only a small power supply. All equipmenf for XRD work including batteries can be accommodated in a suitcase and therefore lends itself to field use by mineralogists, for example, or in materials testing.

The manual X-ray spectrometer, PW-1410, has been upgraded to a level of automatic operation through the addition of a hardwired Angle Mode Programmer, a Motor Controller and by interfacing a dedicated microcomputer. It is the purpose of this communication to describe the system together with data reduction software developed for the microcomputer. This paper will also discuss quantitative analysis techniques using matrix mass absorption coefficients measured with a new absorption cell attachment.

The use of energy dispersive analysis of x-rays (EDAX method) is now well entrenched in the electron column field(l), where more scanning electron microscopes have been fitted with EDAX instrumentation than all of the conventional (wavelength-dispersive spectrometer) microprobes ever made. The principle advantage of the EDAX approach for the SEM user is the efficiency of detection, which permits its use at the low power levels of the SEM. In addition, the simultaneous analysis of the entire spectrum and the lack of focusing restrictions that permits analysis of rough samples are important advantages.

Reported here is a computer-coupled energy dispersive x-ray system capable of performing rapid elemental analysis and phase identification employing fluorescence and energy powder pattern analyses, The system consists of a Si(Li) detector, commercial diffractometer, vacuum specimen chamber with fluorescsnce geometry and a multichannel analyzer interfaced to a laboratory computer.

The continuous radiation is used to excite both fluorescent x-rays and Bragg reflections. The diffractometer spectrum obtained with a fixed 2θ setting gives the crystalline diffraction pattern plus the detectable emission lines. With the fluorescence geometry, the system is capable of detecting the characteristic x-rays of the elements down to Na. The fluorescent spectrum can be used to separate the diffraction peaks from the emission lines in the diffractometer spectrum. Furthermore, the fluorescent spectrum can be used for qualitative or/and quantitative elemental analysis.

Software was developed for element recognition, to isolate the diffraction pattern from emission lines, d-spacing calculations, as well as matrix corrections and mass fraction determination of elements in quantitative analysis.

We have recently described an automated powder diffractometer, the APD-3500, in which a small (4K x 18 bit) minicomputer is utilized for hardware control and data processing. Among the software programs developed for this system is a “Peak Search“ routine which is designed to record a diffraction pattern with optimum, use of scanning time and to give output directly in J.C.P.D.S. Powder Data File Format. Optimum use of scanning time is achieved by a double-pass System in which a first (rapid) pass is made to establish significant data.

A second routine is a “Line Profile” routine which gives the choice of rapid slewing or slow Step scanning of the goniometer between selected angular ranges. The routine thus allows the controlled stepping of the goniometer between selected angular ranges giving incremental and integral, background corrected counts. An automatic internal and/or external calibration capability makes this particular mode ideal for quantitative analysis.

A comparison of the relative merits of the energy dispersive derector-pulse height analyzer, scintillation detector-graphite monochromator, and proportional detector-pulse height analyzer combinations.

Typical energy dispersive detectors are not configured for maximum efficiency on the diffractometer. Being only on the order of 3 mm diameter, a good deal of the available information is not collected by the detector. This is especially true with the Wide optics found in modern diffractometers. The energy dispersive detector incorporated into this system is optimized for the x-ray diffractometer. Its detection area is a 1.25 X 0.25 inch rectangle. The resolution is only sufficient to remove the Kβ portion of the spectrum.

Conventional diffractometer techniques incorporate either a scintillation detector-crystal monochromator, or a proportional detector-pulse height analyser combination. The question posed is “what are the advantages in signal to noise ratio and pulse height distribution of the energy dispersive-pulse height analyzer over the more conventional arrangements.”

Tomographic systems are able to produce cross sectional planar images of three dimensional volumes because of the relative motion of the source, film and the volume under examination. Analysis shows that the image produced is a result of a three dimensional linear filtering process where the filter characteristics are determined by the scan geometry (3). If, instead of integrating continuously on a single film, a set of N radiographs are recorded, each corresponding to a point along the scan trajectory, then a simple filter can be defined to reconstruct the entire three dimensional structure from this data. In this case, the transfer function exhibits repetitive peaks whose spacing is determined by N and whose width is determined by the total scan length. The number of views required to produce the same “blurring” as the continuous case can then be determined by the Nyquist criteria(3).

An optical projection system based on circular geometry for producing three dimensional medical images has been fabricated and tested. The technique can be generalized to any geometry and to all x-ray applications where plane-by-plane examination of a structure would prove beneficial.

The system to be described includes hardware and software for the on-line computer control of the X-ray diffraction measurement of residual stress. This determination involves accurately measuring the angles at which a back-reflection line is diffracted, first by diffracting planes parallel to the sample surface, and then by planes at an angle (ψ) to the sample surface. The residual stress is calculated from the difference in the two measured diffraetion angles. The procedure executed by the computer consists of locating the peaks, selecting three angles for collection of X-ray counts, correcting the measured counts, fitting the equi-angular intensity measurements to a three-point parabola, calculating the peak angles, calculating the residual stress from the measured angles and typing a report. This automation has eliminated the tedium of the manual X-ray data accumulation and of the residual stress calculation. The online control has also permitted improvements in the technique not practicable with the manually performed measurement of residual stress.

Double bending fatigue tests were performed on cold worked 1100 aluminum, with a maximum bending stress varying from 100 to 50% of the yield stress. Variation of residual macrostress from the cold worked state during the fatigue tests was measured by the photographic x-ray diffraction method. A general decrease of this macrostress was observed. Line profile analysis was used to study the change in the mean effective domain size and microstrain. A rate dependence of the increase in the effective domain size with respect to the maximum bending stress was observed. Microhardness and tensile tests were also made on the specimens during the fatigue testing and the results are correlated with the x-ray Parameters.

Two kinds of improved techniques are described for precise measurement of lattice parameters with an accuracy of 1 ∼7x10-5, in both of which a differenoe of Bragg angles between two X-ray beams such as K α1 and K β1 is measured.

In the first technique, which was applied to relatively imperfect crystals such as gallium arsenide, gallium arsenide phosphide, and gallium phosphide, two different diffraction planes were utilized for two radiations such as CuK α1 and CuK β1 in order to reduce the difference of the Bragg angles to the order of several minutes of arc.

In the second technique, which was applied to a nearly perfect crystal such as silicon, a double crystal arrangement of parallel setting was used, where lattice parameter of the specimen was relatively measured in comparison with the one of the first crystal which was known beforehand.

The angular dependent factors which affect the intensities of X-ray diffraction line profiles, as observed on a Standard X-ray diffractometer, are well established. They are the Lorentz factor, the polarization factor, the atomic scattering factor, and the temperature factor. For residual stress analysis, in which measurements are made at non-zero omega (psi) angles, an absorption factor must be added to this list. It is usual, however, in residual stress analysis, to omit two of these factors: the atomic scattering factor and the temperature factor. The omission of these two factors appreciably alters the change in the combined correction factor with change of diffraction angle. The proper angular dependent corrections are in the opposite sense to those customarily employed. The effect of the additional factors on measurements of residual stress are discussed.

The magnetic crystal disrortion of weakly ferromagnetic α-Fe2O3 was investigated by x-ray diffraction techniques. Here crystal distortion is taken as the temperature dependent changes of lattice constants and thermal expansion coefficients. Moreover, the oxygen position parameter and the carbon-oxygen distance of MnCO3 were determined.

The lattice constants and thermal expansion coefficients of α-Fe2O3 were measured from room temperature down to 243°K. The crystal distortion, as measured by the changes in lattice constants, thermal expansion coefficients and axial ratio, was found to be highly anisotropic. The co hexagonal lattice constant was influenced very slightly by magnetic distortion; it changed only by 0.01 percent between room temperature and the Morin temperature of 254°K. On the other hand, the ao lattice constant changes by 0.11 percent between room temperature and the Morin temperature. The thermal expansion coefficients of the lattice constants showed a similar contrast. The co coefficient was found to be independent of temperature from room temperature down to the Morin temperature. However, in the same temperature range, the ao coefficient showed an anomalous increase with decreasing temperature. In addition, the ao coefficient showed an infinite discontinuity at the Morin temperature.

The change in the axial ratio with temperature suggests that the net weak ferromagnetic moment of α-Fe2O3 reaches a maximum at 275°K.

The oxygen position parameter, x, in MnCO3 as determined from two reflections has a value of 0.2702 ± 0.001. The carbon-oxygen distance as calculated from the lattice constants and the oxygen position parameter is 1.29 ±0.002 Å. This value is another confirmation of the Pauling theory of the resonating carbonate structure.

The results of the third year of observations on the effects of self-irradiation on the lattice of 238(80%)PuO2 indicate a small but definite trend toward a reduction in the lattice parameter of this material. The lattice parameter increased from 5.41408±6 Å at 5 weeks to 5.41636±38 at 114 weeks and then decreased to 5.41364±23 at 157 weeks after the start of the experiment. A similar trend is noted in the half-widths of the diffraction lines. For the 622 reflection the half-width increased from 0.283±14°20 at 5 weeks to 0.711±90 at 114 weeks and then decreased to 0.660±89 at 157 weeks.