The reduction of the background in x-ray fluorescent spectra is achieved by total reflection of the primary x-rays at a plane, smooth surface of a suitable reflector material. This effect essentially reduces the background caused by scattering of the primary photons from the substrate, thus improving the lower limits of detection in x-ray fluorescence analysis (XRF). Introducing additional to the reflector-substrate another reflector in the primary beam, this affects the spectral distribution of the exciting radiation (“high energy cutoff”). The result is an improved background, where detection limits of picogram amounts or concentrations in the ppb range are attainable for medium Z elements with energy dispersive detectors.

Most users of X-Ray Fluorescence analysis (XRF) are accustomed to thinking of it as a bulk analysis method, and certainty in many cases it does report the representative average composition of the sample. But whereas the trace analysis of lead in gasoline may extend to a depth of several cm., that for a light element in a moderately absorbing matrix may take place only within a few micrometers of the surface, placing severe demands on the specimen preparation method to produce a surface layer representative of the bulk material. XRF is rarely used specifically to analyze surfaces, except for some coating thickness measurements. Nevertheless, many of the same laboratories that rely heavily on XRF are called upon to perform surface analysis for various purposes, and therefore this invited survey paper witt seek to present an overview of some of the major techniques which are available.

Modern geoehemical studies have evolved significantly over the last few decades. Exploration programs have become more ambitious and have expanded to cover entire nations, requiring many tens of thousands of analyses. With the increased demand for chemical analyses, many laboratories have opted for automated, rapid, high-technology instruments, rather than adding a larger number of skilled personnel. It is not clear whether the advance of automated instrumentation in the laboratory predated the demand for large numbers of analyses, but with the increased availability of computers, even basic research problems have come to employ larger numbers of samples. Because modern geochemical laboratories can seldom have all new analytical equipment, from the beginning it is clear that priorities must be assigned and choices made when selecting each new instrument.

Mass attenuation coefficients (MACs) are frequently estimated over a range of wavelengths in x-ray spectrometry from the intensity of the Compton peak IC associated with a prominent tube line. The MAC μλ at wavelength λ is estimated from the MAC at the Compton wavelength λC with the approximations μλ α μC and μC α 1/IC. Systematic errors may introduce absorption edge bias (AEB) effects into the results, caused by sample components vith absorption edges between λc and λ. A procedure is described which eliminates AEB effects by measuring IC using emission radiation from a primary beam filter.

Several correction algorithms (Lachance & Traill (L-T), Rasberry & Heinrich (R-H), Lachance & Claisse (L-C) and Claisse & Quintin (C-Q) have been inter compared. All these algorithms assume that the effects of the matrixelements can be added in a linear way. Tertian, however, has shown that this is perfectly valid only when absorption is the only interelement effect to be taken into account. When enhancement is involved, corrections for crossed-effects are necessary. We have also evaluated the application of crossed-effect correction. Calculations were done, using theoretically calculated relative intensities. Influence coefficients were calculated by a regression method.

For the analysis of solid samples, such as alloys, by means of XRF, one normally assumes that either the samples are completely homogeneous, or that heterogeneity effects can be neglected or cancelled out. The authors have studied this effect by means of Monte Carlo Simulations. Two particular geometries have been studied. For the first simulations the sample consisted of alternating lamellae of different composition, perpendicular to the irradiated surface. Later, the more realistic distribution of a spherical dispersed phase in a matrix was simulated.

Fundamental parameters methods of calculating elemental concentrations from x-ray fluorescence (XRF) data put stringent demands on the accuracy of XRF intensity measurement. Two significant but rather unobvious sources of intensity error are inaccurate deadtime correction in energy-dispersive x-ray fluorescence (EDX) systems and errors caused by certain combinations of sample thickness and geometry. A method of evaluating the accuracy of an EDX system's deadtime compensation is given. In our system, a nearly constant error was found over a wide range of average x-ray photon energy. Accurate deadtime correction was obtained by adding a mathematical correction to the partial compensation by the built-in system electronics. Problems related to sample thickness and geometry can result when the effective x-ray path length is longer than the geometrically possible path length (high energy analyte radiation and light element matrix). In such samples, a known sample thickness less than the geometry-limited thickness must be used for accurate intensities. A method for determining when this is necessary has been devised. A useful empirical correction for absorption by Mylar windows is also given.

Various X-ray fluorescence spectrometers are now commercially available, and these spectrometers are classified into two categories, i.e., energy dispersive and wavelength dispersive (Table 1). Energy dispersive instruments are of low resolution. Wavelength dispersive instruments are often referred to as high resolution. However, commercially available wavelength dispersive instruments are usually equipped with a one-crystal dispersion unit. Therefore, to be more precise, these instruments could be called medium resolution apparatuses. There are other types of spectrometers which are two- or three-crysral spectrometers, and which are known to have very high resolving power. High resolution X-ray fluorescence spectrometers, however, have rarely been constructed. This is because this type of spectrometer needs a very precise and complicated scanning mechanism, and also because the intensity of X-ray fluorescence is often lost.

The potential advantages of synchrotron sources for x-ray fluorescence analysis have been discussed by several authors. These advantages include high incident flux, tunable excitation energies using crystal monochromators, and reduction of scattered background due to polarization effects. Minimum detectable limits have both been calculated and measured, and have demonstrated the improvements that can be achieved vising synchrotron sources. In this paper we present results obtained using tunable, monochromatic excitation and a high resolution lithium-drifted silicon, Si(Li), spectrometer for the quantitative analysis of a variety of sample types. Our experiments were designed to investigate the advantages and limitations of tunable monochromatic excitation with respect to optimum sensitivity, accuracy, and elemental selectivity in energy dispersive x-ray fluorescence analysis.

Trace element analyses by energy dispersive X-ray fluorescence measurements were made using synchrotron radiation from a dedicated electron storage ring at the Photon Factory in Japan. The continuum or the monochromatic beam was used for excitation. A crystal monochromator or two types of mirror systems were used for monochromatic excitation. A high signal to background ratio was attained with the crystal monochromator, while the highest absolute detectability was achieved with the mirror system. The minimum detection limita obtained from thin samples are of the order of 0.1 ppm or less than 0.1 pg. Furthermore the signal to background ratio was significantly improved by using an X-ray mirror as a sample support in which, external total reflection of exciting X-rays occured.

Polarized x-rays are used to excite samples of any kind and shape to emit characteristic radiation. In the appropriate geometry, where source-sample-ED detector are in any orthogonal position to each other, the exciting poLarized x-rays will be practically not scattered from sample and substrate into the detector. This reduces the background considerably and hence improves the lower limits of detection. The production of intensive polarized x-rays is done by using a single crystal-where Bragg reflection occurs at an angle 2 θ = 90° instead of amorphous low Z scatterers. The result is a linear polarized and monochromatic beam. The use of curved crystals instead of plane single crystals further increases the intensity of the exciting radiation. The lower limits of detection attainable with the recently constructed compact polarizer device are in the sub ppm range or in absolute amounts around 150 pg for medium Z elements.

The principle to utilize total reflection of the primary X-rays in fluorescence analysis is known since a lot of years. Nevertheless, analytical chemistry did not profit from the inherent advantages of the method for a long time. The main reason for this failure was the lack of instruments which were easy to use in practice. A few years ago, however, the development of a proper mechanical design and of adapted sample preparation techniques led to commercially available spectrometers which throughout fulfill the demands for routine applications. Since then the utilization of Total-Reflection X-Ray Fluorescence Analysis (TXRF) has increased rapidly. The scope of work is meanwhile widespread over environmental research and monitoring, mineralogy, mineral exploration, oceanography, biology, medicine and biochemistry. Accordingly, numerous matrices have been handled.

Many X-ray spectrometers are now equipped with dedicated computers, which provide automated data collection (either Energy or Wavelength dispersive), and perform quantitative corrections to derive concentrations from intensities. Some of these systems are also provided with software to assist in qualitative analysis, or identification of the peaks and elements present.

These programs include simple aids to the user (various graphic display modes, peak identification markers for KLM lines, and so forth), spectrum deconvolution (to reveal hidden peaks), as well as complete spectrum scanning algorithms that attempt to report the entire list of elements present. It is not the purpose of this paper to compare or critique the performance of these various routines, but rather to illuminate them by providing some insight into how human thought processes appear to perform qualitative analysis.

There are currently no plans for the commercial reprocessing of nuclear fuel in the U.S. This is not the case in Japan, England, France, Germany, or the USSR. The U.S. does, however, include the reprocessing of spent nuclear fuel as a part of its defense programs. The Office of Safeguards and Security of DOE's Defense Programs has funded the development and optimization of numerous nondestructive analysis techniques including K-edge x-ray fluorescence analysis. This paper reviews some of the implementation efforts resulting from R&D supported by that office.

In 1979, the concept of using K-edge x-ray fluorescence analysis (K XRFA) for the analysis of actinide solution concentrations was first presented. K XRFA using small radioactive Co-57 sources has been shown to be a practical way to measure actinide solution concentrations in offline, at-line, or on-line configurations. The experimental methodology is elegant and the hardware is simple.

Recently, we have developed a new radioisotope excited technique for x-ray fluorescence which employs “direct” beta particle excitation to produce characteristic x-rays for elemental analysis. We have also studied this technique with various other isotopes with different decay modes. But in all these previous works, we have only studied the x-ray energy region below the Bak lines (≤ 35 KeV), because of the poor efficiency of the Si(Li) semiconductor at higher energies.

In this work we have employed a planar Ge detector, which has about a 100% efficiency in the range of about 20 to 100 KeV. We have also compared this detector with a HgI2 detector, even though its resolution and efficiency is far less than the planar Ge detector.

In any analytical method, one of the first considerations in obtaining precise and accurate results is the stability of the instrument. Progressive refinement in design and manufacturing techniques over many years has led to short-term stability of X-ray generation, transmission and detection in XRF spectrometers of a magnitude such that, in favourable cases, the random process of emission (and to some extent, absorption) is the major source of error; superimposed on this, however, are other variations, some of them periodic.

Many processes modify the intensity of X-rays recorded by a spectrometer's sealer and timer: besides those produced by the instrument itself there are potential external perturbations perhaps reverting to electricity generating stations or even further. More important effects may be caused by local features such as ambient temperature and pressure, and in the case of improper setting of machine conditions or of existence of a fault, enhancement may be observed.

Traditionally users of x-ray spectrometers have been forced to optimize their instrument for light element analysis and heavy element analysis through the choice of either a light or heavy element anode. The advent of the rhodium tube brought some relief to this situation, but some compromise was still required. The recent addition of the scandium side-window tube to the range has provided twice the excitation efficiency of the chromium tube, whilst providing an equivalent heavy element performance.

In the past there have been attempts to produce x-ray tubes combining the properties of both light and heavy element tubes. An example of this is the dual-target tube which has two preferentially excited targets (anodes) of different materials constructed side-by-side.