Optimized vacuum spectrographic measurement of low-energy fluorescence has been found to yield counting rates and peak-to-background ratios which are enough to permit the extension of fluorescence analysis for elementary chemistry into the light-element range—sodium through boron. This is accomplished with an efficient, demountable ultrasoft X-ray source, with close coupling among source, crystal, and detector, with KAP and multilayered stearate analyzers, and with, optimized flow-propordonal counting. Specific methods for achieving peak-to-background ratios on practical samples containing these light elements are presented. The extension of these methods of light-element analysis with the use of curved long-spaced crystals for X-Ray macroprobe and electron microprobe measurements is discussed. The design and construction of multilayered soap film “crystals” for long-wavelength X-ray analysis is described.

Instrumentation for soft X-ray research is briefly described. Preliminary results are shown for several dispersing systems which have been found useful in the 10 to 60 Å region. Results from single crystals of Group 1A salts of phthalic acid are discussed and intensity measurements are reported. Other organic single crystals, such as long-chain derivatives of adipic acid are evaluated, along with films of organic compounds prepared by the Blodgètt-Langmuir technique. The efficiencies of the dispersing elements mentioned above are compared with commercial crystals such as mica, gypsum, ADP, and EDDT. Performance of detectors for soft X-rays is shown for fiow-proportional and Geiger counters, a solid state silicon barrier detector, and a windowless resistance strip magnetic electron multiplier. The relative efficiencies of several photocathode materials for the latter are discussed. The design for a scintillation counter sensitive over a wide energy range is discussed and preliminary results using the counter in the region 1-15 Å are shown.

Two aluminized gratings with different blaze angles, 1° and 7.55′, and a barium stearate monolayer crystal were compared for first-order X-ray line spectra from oxygen (23.707 Å) to beryllium (113 Å). The peak intensities for the gratings were found to be greater than for the monolayer crystal when the same detector, source, and excitation conditions were used. Line-to-background ratios and halfwidths were comparable.

The fluorescent excitation of long-wavelength X-ray spectra is reviewed with respect to X-ray tube target element, inherent filtration, and optimum kilovoltage. A demountable X-ray tube vacuum spectrograph designed for the determination of the light elements is described. Operation of this instrument with both secondary and combined primary—secondary excitation is evaluated. Examples from the literature are cited to show the feasibility of direct electron excitation of longwavelength spectra.

Fluorescent X-ray spectrometers bave a high inherent reproducibility. It is possible to achieve results, on a routine basis, which have a coefficient of variation of less than 0.1%. Given extremely good chemical standards and the absence of interelemental effects, these readings can be interpreted into percentage concentrations with the same order of accuracy. Results, with this accuracy of analyses of copper in a straight brass will be given. With samples of greater complexity and more elements the interelement effects are large and have to be tackled. A correction method has been established and will be explained in detail.

As an example of the method it was decided to tackle the determination of chromium at high concentrations (7-32 %) in high-alloy steels containing additions in the following ranges: Si, 0.2-2.4%; Al, 0,07-5%; Cu, 1,3-3.6%; W, 0.25-3%; V, 0.1-0.3%; Nb, 0.1-0.3%; Mo, 0.1-4.4%; Ti, 0.05-1%; Mn, 0.35-10%; Ni, 0.2-24%. These very wide alloy specifications were a severe test on the accuracy of the chromium determination & the correction for interelement effects.

Recently, interest in the determination of selenium in trace amounts has been greatly intensified because of the nutritional aspects of this element. It has been reported that selenium in the amount of 13 μg in the form of sodium selenite in 100 g of feed has an effect similar to that of vitamin E. In the field of semiconductors, the detection and determination of trace amounts of selenium in arsenic, antimony, and small single crystals of solid solution of cadmium selenide and sulfide are of considerable importance in semiconductor performance.

In the Aerospace Research Laboratories, 4,5 diamino-6-tbiopyi-imidine has been successfully adopted as a reagent for the spectrophotometric determination of selenium. The reaction of 4,5 diamino-6-thiopyrimidine and tetxavalent selenium produces a yellow color with the formation of elemental selenium. It is possible to determine elemental selenium by collecting it in a thin layer. The selenium deposited in this layer may then be determined by an X-ray fluorescence method. A procedure of this nature has the advantage of eliminating the matrix effects commonly encountered in X-ray fluorescence. Furthermore, the slow generation of selenium affords a convenient means of detection and confirmation of this element by the use of X-ray diffraction procedures. By this technique selenium is first converted to its tetravalent state and is then reacted with 4,5 diamino-6-thiopyrimidine. On standing, the selenium is reduced to a red precipitate of elemental selenium which can be dissolved in carbon disulfide. Finally, the selenium can be converted into its hexagonal structure by annealing at 205-207°C.

Mineralogical effects, which relate to the occurrence of an element in different forms of chemical combination, often are a problem to the X-ray analyst since these forms usually differ in X-ray sensitivity. An example of this is cited in connection with the analysis of sylvite concentrates for potassium. An evaluation is made of mineralogical effects and a quantitative relationship between X-ray intensity and mineral form and particle size is derived. If the particle size of a minor constituent is reduced sufficiently the mineralogical effect disappears. Target materials for X-ray sources are found to have only minor effects on relative intensities of different compounds of an element. Finally, it is concluded that the advantages of higher intensities gained through the use of target materials close in atomic number to the material being analyzed far outweigh particle size effects which are shown to be relatively small.

During the normal operation of conventional receiving-type electron tubes, certain metallic elements, such as barium, strontium, nickel, and manganese, sublime from the hot cathode and deposit on cooler parts of the tube. This process gradually impairs the performance of the tube and may eventually limit its useful life. This paper describes some applications of a variety of X-ray spectrometric techniques to the qualitative and quantitative analysis of these sublimates and to the mapping of their distribution on various surfaces in the tubes.

All work was done on a standard commercial X-ray spectrometer, but specially designed accessories are described for mounting and rotating small parts in the primary X-ray beam and for confining the beam to these parts. Procedures are given for analysis of microgram deposits on electron-tube cathodes, grids, plates, micas, and bulbs by both nondestructive and filter-paper-disk techniques. Procedures are given for point-by-point mapping of the distribution of sublimates over the surfaces of these parts by use of a commercial “X-ray probe” selectedarea accessory in conjunction with a curved crystal to give increased intensity. Procedures are given for various techniques for calibrating all the analytical methods, including preparation of thin films of known composition.

The results of some typical electron tube studies are described and illustrated. The approximate sensitivity of the methods for determination of total sublimate is, in micrograms per part, 0.1 for barium, strontium, and manganese and 0.05 for nickel. The approximate sensitivity for mapping sublimate distribution with a 1-mm aperture is, in micrograms per square centimeter, 1 for barium, strontium, and manganese and 0.5 for nickel. Although the techniques described are applied specifically to sublimates in electron tubes, they are also readily applicable to analysis of specks and films on other small parts and to thin-film studies in general

A method is presented for the quantitative determination of thorium in monazite by fluorescent X-ray speetrometry. Strontium is used as an internal standard, and sample preparation consists of mechanically blending the finelyground (325-mesh) sample with reagent grade strontium carbonate. The prepared sample blend is packed into a sample cup and irradiated at 50 kV and 27 mA. A lithium fluoride analyzing crystal is used, and the intensity of the Th Lα1 and Sr Kα peaks are scaled. Intensities are corrected fov background. Calibration is achieved with a single sample of monazite of known thorium content. National Bureau of Standards monazite sample number 2601 (9.65% ThO2) was used in this work. The standard monazite is blended with strontium carbonate, irradiated, and scaled under the same conditions as the sample. The thorium content of the sample is then calculated by the equation: %ThO2 = (R sample) (%ThO2 standard)/R standard, where R = Thcounts/sec/Srcounts/sec. Replicate results agreed within 0.2% ThO2 on six samples of monazite which contained from 4.4% to 6.8% ThO2. Independent results by a second laboratory show no significant difference in precision.

The method was used to determine the thorium content gf several rare-earth materials which contained approximately 0.5% to 1.0% TriOs. The results agree closely with values obtained by gamma-ray spectrometry.

A technique has been developed for quantitative analysis of the major constituents in small samples (less than 0.1 g) of corrosion products found On various missile parts. The technique was developed primarily to aid in interpretation of X-ray diffraction patterns of multicomponent corrosion products. The corrosion sample is mixed with KBr and the mixture is pressed into a disk in the same fashion commonly used in infrared work. The Br Kβ1 line serves as an internal standard. Intensities of the element's Kα line and the internal standard are established by scanning through the appropriate Bragg angle and recording peak heights on a strip-chart recorder. A xenon-filled proportional detector and a pulse-height analyzer were used.

The percent of an element is determined by references to calibration curves which relate intensity ratios to weight ratios for the 3-d transition metals. The oxides of the metals were used in the preparation of the calibration curves. The weight ratio vs. intensity ratio relationship is linear over the weight ratio range 0.01 to 0.11. Data were collected using an air path and a helium path; the higher intensity ratios obtained with, the helium path, and the dependence of intensity on atomic number are illustrated. A comparison is made between the intensity ratios in a KBr matrix and in a NaBr matrix. The technique developed requires about 15 min sample preparation time.

An X-ray fluorescence method is described for the analysis of niobotamalate concentrates, bringing speed and accuracy to determinations that are difficult chemically. Many of the problems inherent in the X-ray fluorescence analysis of powdered samples are eliminated by fusion of the sample in a mixture of Li2B4O7 and La2O3. Absolute amounts can be determined without reliance on chemically analyzed standard concentrates. Standards in a wide range of concentrations are readily prepared from pure chemicals. La2O3 plays a dual role, aiding in the fusion of highly refractory oxides and minimizing absorption differences between samples. In addition to niobium and tantalum, elements commonly present in niobate-tantalate ore concentrates, such as titanium, iron, tin, and manganese, are readily determined.

The fusion mixture consists of 80 mg of sample, 120 mg of La2O3 and 800 mg of Li2B4O7. The fusion is made in graphite crucibles at 1100°C for 15 min. The cooled bead is ground in a mixer-grinder and pressed into a pellet. X-ray fluorescence measurements yield linear calibration curves for each of the elements over a wide range of concentration.

The use of electronic discrimination at a fixed goniometer position for an X-ray spectrographic determination is a conventional means of improving the signal-tonoise ratio and minimizing interferences due to multiple-order diffractions of radiation with shorter wavelengths. The same advantage can be gained with a moving goniometer in a scanning procedure if adjustment of the baseline of the electronic discriminator is properly coordinated with the movement of the goniometer. Å study of optimum baseline levels for various wavelengths in the range of 0.3 to 3.0 Å indicated a hyperbolical correlation of optimum baseline voltage vs. 2θ angle. This experimental correlation is the basis for synchronization of electronic discrimination with movement of the goniometer during the scanning procedure. A suitable mechanism for coordinated baseline adjustment was developed by Technical Equipment Corporation, Denver, Colorado, and improved results are obtained for trace analysis with a scanning technique routinely used for qualitative, semiquantitative determinations for elements with atomic numbers 22 and higher.

Many variables influence the performance of X-ray fluorescence analysis with regard to speed, precision, accuracy, limits of detection and the elements to be determined. Some of these variables, such as path transmittance, fluorescent yield, and detector efficiency, can be considered constant; however, the primary X-ray beam cannot be assumed a constant. The intensity distribution vs. wavelength function of the X-ray tube remains the main variable with which the X-ray spectrochemist must concern himself in providing optimum analytical performance and speed.

This distribution is dependent upon target material, window material and thickness, and tube voltage. Though the overall intensity is dependent upon the atomic number of the target material, the wavelengths of the characteristic line spectrum of the target material can provide additional enhancement for elements which absorb such wavelengths readily.

Some of these variables in tube design parameters have been studied to provide comparison performance data on the analysis of a variety of base materials for a number of elements of interest. These data and their interpretation are presented to show how improved fluorescence analysis can be achieved by modifications in X-ray tube design and manufacture.

A new concept in a dual-function diffractometer-spectrometer has been developed. With this new concept, it is unnecessary to disassemble any part of the instrument when changing from diffraction to air-helium spectroscopic applications and vice versa. Therefore, the alignment of the instrument is not disturbed and changeover is accomplished in a matter of seconds.

The diffractoraeter is the standard biplane diffractometer, having an angular range of −55 to +164° 20 and a full 360° ω. Without interfering with the entire angular range of the diffractometer, the spectrometer section allows an angular range to 154° 2θ, This uninterrupted freedom of angular motion for both functions is accomplished by the design concept of using different frames of reference for the 0-180° 2θ plane of the diffractometer and the 0-180° 2θ plane of the spectrometer. In one case, the motion of 16 toward higher positive angles is counterclockwise and in the other case, it is clockwise. The frames of reference are displaced by 96°.

The helium attachment for low-atomic-mimber elements ensures complete replacement of the air with helium in a flushing time of 15 sec. After changing samples, the helium atmosphere returns to equilibrium within 5 sec.

A ten station programmer also has been developed, enabling one to automatically program oscillating scans and stepped scans of diffraction profiles as well as sequential emission analyses. The angular position is programmed in increments of 0.01° and is reproducible within 0.005°.

During the sequential mode of operation, 20 is driven to the next position at a rate of 40 deg/min. In addition, stepping increments as small as 1/120° and scanning speeds of ⅛, ¼, ½, 1, and 2 deg/min are available. In all of the above modes of operation, data are collected and printed as a function of preset time, preset count, or angular position.

In this paper, typical performance characteristics and applications are presented.

Among the advantages of the X-ray spectrograph is the economy realised by minimizing the time of analysis. In the past few years automated features have been developed and applied to reduce analysis time further and to minimize operator errors. These have included a programmed goniometer, an automatic crystal changer, a digitized printer or automatic typewriter, and recently an automatic sample loader coupled with a programmer. With this last attachment, unattended analysis of solid metal samples has been achieved. The precision of data obtained automatically compares favorably with similar data obtained manually.