XRF can be a powerful tool for quantitative elemental analysis-it can also be a big headache. The problem is, of course, that to perform a proper quantitative analysis one needs to first develop a good matrix correction model to convert the measured line intensities to concentrations. For empirical models, aside from the inherent difficulties in optimizing a set of interelement correction coefficients, the commercially available computer controlled spectrometers have in the past either provided rather poor regression analysis programs with their software or none at all. The reason for this has been the rather limited core size of the online computer and the lack of fast efficient mass storage facilities such as floppy disks.

It might seem at first glance that quantitative models relating intensity to concentration should be identical for energy - or wavelength - dispersive fluorescence analysis. In both cases the interelement effects that complicate the use of simple linear calibration curves occur in the specimen, at which time the X-rays are not yet aware of which kind of detector will be used to count them. This can be true in some cases, but not in general.

Recent developments in analytical techniques and software have allowed the accurate quantitative determinations of both the major and minor elements in stainless steels by energy dispersive x-ray fluorescence. The successful analysis of 300 and 400 series stainless steel is reported utilizing this technique. The analysis of this type of material represents one of the most severe tests of the method due to numerous peak overlaps and interelement effects such as absorption and enhancement.

Sixteen standards of ASTM 300 series and ten 400 series were prepared by polishing on a 220 grit aluminum oxide belt and subsequently washing the surface in absolute methanol. Analyses were performed with an EG&G ORTEC 6110 Tube Excited Fluorescence Analyzer utilizing a dual anode (Rh/W) x-ray tube. Peak deconvolutions and interelement corrections were made with a 16K PDP-11/05 computer utilizing the program FLINT (1). Utilization of spectral deconvolutions and interelement corrections yields a relative accuracy of approximately IX of the concentrations of the major elements.

Alpha-particle induced X-ray emission studies of steel samples have provided quantitative, non-destructive determinations of the carbon content in surface layers and in the bulk. The analyses were performed using the ALPHA-X™ technique of alpha-induced X-ray emission. Absolute carbon surface densities were determined directly from the ratio of the carbon X-ray count-rate for the unknown layers to that for standard free-standing carbon films. Assessment of the bulk detection limit was made using a steel standard with 0.5 weight-percent carbon. The threesigma minimum detection limits for twenty minute analyses were 0.1μg carbon/cm2 (equivalent to≃5 Å) in the layers and 1000 ppm carbon in the bulk.

Using corrections applied to the XRE analysis of copper ore flotation materials, two methods of determining the ratio of solid to liquid content in slurries are described. Both are based on the use of coherently scattered radiation. In the first method the intensity of Cu, Pb and Fe is normalized using a coefficient defined as a ratio of scattered radiation from the sample and pure diluent. In the second, the regression equations are applied. Satisfactory results have been obtained for solid contents ranging from 10 to 40%.

Quantitative assay of body components in the living human is a subject of keen interest. Measurement of such fractions on superficial view seems a simple analysis problem. In fact, the opposite is true. The ‘specimen’ is not amenable to the usual destructive analytical procedures. Due to its importance, a continued effort has been mounted to assay the most basic of such components. This paper and bibliography traces the recent applications of penetrating radiation for such analysis.

Some of the fractions whose measurement has been attempted are; bone mineral, adiposity (lipid to lean ratio), body hydration, tissue density, voids and spaces (i.e. lung). The importance of quantitative analysis of such gross components is apparent when taking mineralization as example.

The Digital Equipment Corporation (DEC) operating system RSX-11M has recently been installed on a DEC PDP 11/34 computer which is used for the control of, and to acquire and process data from, three X-ray diffractometers, one X-ray fluorescence analysis unit and an electron microprobe. The RSX-11M system replaced the modified DEC 1 - 8 User BASIC previously employed, thus replacing an operating system which was known in detail only to its writer by a system which is supported by the computer manufacturer. There are three major advantages in the use of RSX-11M over 1 - 8 User BASIC: an improved handling of program scheduling, the integration of the software driver for the computer - X-ray electronics interface into the operating system without a major modification of the latter, and the ability of RSX-11M to undertake concurrent execution of instrument control, data acquisition, and data reduction. The 11/34 - RSX-11M system has been implemented to use FORTRAN; a BASIC interpreter has, however, been added which allows users to interact on-line with the computer. A command interpreter which can accept a command line from a terminal has been included.

The determination of optimal excitation conditions for energy dispersive x-ray fluorescence is particularly critical for multielement analysis covering a wide range (viz. 10 or 20 keV) of the spectrum. Functions that quantitatively describe the spectral quality are used as objective functions in pattern search algorithms. It is shown that the filters can be arranged in a definite order, at least with respect to the energy of the K-absorption edge of the tube and can therefore be employed as a dimension in the optimization process. Of the algorithms that were compared, the Nelder-Mead and Routh-Swartz-Denton versions of the sequential simplex search gave the best results if the excitation voltage and the current could be controlled in small increments. If the optimization includes dimensions with a few discrete stages (e.g. filters) the fixed size simplex proved to be of greatest value. The functions can be weighted to reflect special interest in one or more elements. Conditions for increasing the counting time and terminating the search are discussed.

Energy dispersive x-ray fluorescence is an established and versatile tool for measuring major and trace elements in virtually any kind of solid or liquid specimen. The usefulness of the method could he extended even further if the time of analysis for multicomponent samples could he reduced. In other words, we desire to analyze a wide range of elements with detection limits at least as good as obtained when the excitation conditions are optimized for a specific element or narrow range of elements. A major impediment to achieving this goal when analyzing bulk, low-Z materials is the scatter of source radiation into the detector by the specimen being analyzed. The adverse effects of the scattered radiation are its contribution to the background signal (i.e., “noise“) and its overwhelming contribution to the limited counting rate of the system electronics.

Mercuric iodide (HgI2) with high atomic numbers of 80 and 53 for the components, and wide bandgap (2.1 eV) has been considered a potentially useful material for gamma ray spectroscopy for the last number of years. Considerable effort by numerous groups has been directed toward purification and growth of large single crystals of this material. Even with continuing improvement, however, the material is still characterized by good electron transport properties and only modest to poor hole transport behavior. Interesting results nevertheless have been reported in mid-range gamma ray spectroscopy using “up to millimeter thick sections of HgI2. One question always asked is: has the fundamental of lattice scattering etc. been reached in increasing hole transport behavior in this material? There is interesting speculation at this time that this may not be so.

Chlorinated poly(vinyl chloride) (CPVC) is similar in many ways to its thermoplastic parent poly(vinyl chloride) (PVC), Chlorination imparts superior high temperature properties, which make CPVC preferable for many and varied applications.

Our primary purpose in studying CPVC with x-ray photoelectron spectroscopy (XPS) was to see whether we could gain insight into how chlorine substitutes into the PVC molecule. We also hoped we would obtain insights into the chlorination reaction itself, and how variations would affect the final product.