With the development of soft X-ray sources by Henke, it is now possible to extend the range of quantitative analyses by X-ray spectrometry into the very light element range (magnesium through, boron). Techniques of specimen preparation, recognized as one of the most critical aspects of quantitative X-ray analysis in the heavier elemental ranges, present new problems, especially with the physically and chemically heterogeneous materials encountered in mineral and rock analysis. This paper presents the results of new tests extended into the light and very light element range. It is shown that most of the problems previously found are intensified as lower atomic number elements are considered; conditions of the specimen surface are critical and require precise control of particle size. In addition, significant absorption by relatively heavier elemental constituents, either in a separate phase or within the same phase, causes poor precision and is difficult to predict. Thus, the combination of matrix absorption within and between discrete phases, combined with the difficulty of grinding different phases to a uniform particle size, suggests that fusion techniques are required for hest precision. Without fusion, both scatter and bias are introduced through non-linear calibrations, with each calibration dependent upon specimen composition and particle size. It is concluded that only with nearly uniform specimens, where gross deviations are to be checked (e.g., process control) can ground rock powders be used with success in light element analyses.

Lead stearate and lead lignocerate multilayer soap-film structures are used to disperse the K emission lines of boron and beryllium respectively. Data are presented showing the dependence of the peak height and half-width on the number of layers in the lignocerate structure. Spectra are presented and compared for the pure element and several compounds of each element. Both electron and X-ray excitation are used. Detection is by a thin-window flow-proportional counter.

Results are shown for O and F K emission lines from several oxide and fluoride compounds. The spectra are examined for both structure and wavelength. Excitation of the spectra is by both electrons and X-rays. The emission spectra from several oxygen- and fluorine-bearing gases and the influence of the composition of the diffracting crystal on the shape of the observed oxygen spectra are also discussed. Curved crystals of potassium acid phthalate and lead stearate, and a thin-window flow proportional counter, are used to obtain the results.

The major elements m common rocks are of low atomic number, but analyses of high precision are possible by soft X-ray spectrography if several grams of rock sample are available. The electron-probe microanalyzer is shown to complement this established method by permitting analyses of particles as small as 1 μ in diameter. This paper describes applications of these methods to the analysis of the major and minor elements of silicate, carbonate, and phosphate minerals and rocks.

Elements of particular interest are as follows : carbon in particles enclosed in carbonate rocks; oxygen, as the major constituent of the specimens; phosphorus in phosphatic nodules and apatites; manganese and iron, as colorations in fossil shells; and the group oxygen, sodium, magnesium, aluminum, silicon, potassium, calcium, and iron as complex segregations and zonations within single crystals of several mineral phases.

If the bulk composition of a rock is known, and also the chemistry of the constituent minerals, it is possible to compute quantitative minéralogie analyses of high precision. Thus, the combined use of soft X-ray spectrography and electronprobe microanalysis can provide quantitative chemical and mineralogicat information on the earth's crust on all scales from thousands of square miles (by means of appropriate sampling) down to the scale of 1 μ.

A system is described whereby X-ray fluorescence analysis of elements below atomic number 19 (potassium) is undertaken with a “windowless” demountable X-ray tube. Comparison is made with 100-μ- and 30-μ-thick beryllium windows- Deflection plates are required before or behind the sample to eliminate scattered electrons. A high-resolution X-ray diffractometer is also described by making use of a fine focal line on the target. Burn-through due to localized heating on the target is eliminated by using a thermocouple arrangement.

The primary X-ray analyzer is used for nondestructive spectrochemical analysis of solid specimens. Accelerated electron beams bombard the specimen surface directly and generate primary X-rays which are measured in a vacuum spectrometer. The method of primary X-ray spectroscopy is superior to the fluorescence X-ray spectroscopy because (1) detectable sensitivity for such light elements as magnesium and aluminum is very high, and (2) the correction of the measured value for self-absorption of X-rays by the specimen itself is low. The performance of the instrument and applications are reported.

During the past several years, a number of analyzing crystals have been prepared in the U.S. Air Force Aerospace Research Laboratories. Crystals such as alkaline acid malonates, sucrose, pentaerythritol (PET), and several others have been grown with the sole purpose of application in X-ray fluorescence analysis of silicon, aluminum, and other elements of low atomic number. Crystals from natural sources, such as quartz, mica, and gypsum, have also been procured from different parts of the world for this purpose.

Among the analyzing crystals so prepared, pentaerythritol gave the highest count rate for silicon and aluminum. However, since this crystal is organic in nature, great care must be exercised in handling this type of crystal in order to obtain constant count rates. For the analysis of silicon and aluminum, a-quartz crystal gave a somewhat low count rate, but this crystal has certain advantages over the organic crystals, and can be used for analysis of materials having high silicon and aluminum content.

For instance, when using a conventional X-ray emission vacuum spectrograph of standard make, with the latest alterations, one of the PET crystals attained a count rate of 101,870 cps for pure silicon when the instrument was operated at 60 kVP and SO m A with a negligible background count. With a constant potential attachment and operated at 60 kVP and 34 mA, 1.15,600 cps with negligible background count was obtained for silicon. The count rate for aluminum was of the same order of magnitude with somewhat higher background count.

Several sets of standards have been procured and small amounts of silicon and aluminum in these standards have been analyzed by the latest modified vacuum spectrograph. These results are under study and the limit of detection calculated. Procedures and results are described and discussed.

Art X-ray spectrographic technique requiring no calibration curves is used to determine chlorine, bromine, and lead in automotive combustion deposits. Other elements, such as zinc and barium, can probably be determined by the same X-ray technique. The inert dilution method, described by Sherman in 1957, is employed. Improvements in sample preparation have been made to make the procedure more reliable. Results are calculated by the use of a formula, which is shown to be valid even though sample and standard have composition differences which cause interelement-effect errors if calibration-curve techniques are used. Adequate agreement is obtained between the inert dilution X-ray method and chemical procedures.