Laterites and bauxites are residual products derived from a wide variety of rocks by intensive chemical weathering under strongly oxidizing and leaching conditions. Generally the main constituents of these residues are Fe, Al and Ti present in form of hydroxides and/or oxides. The content of silica depends on the thoroughness of the leaching process. The behaviour of trace and minor elements as Ga, Zn, Cu, Ni, Mn and Cr during the laterite/bauxite formation is not yet well established. Precise and accurate analyses of these elements are a prerequisite for such investigations. In this paper the standard addition method is presented and the results are compared with those obtained by calibration standardization. The accuracy and the precision of the methods as well as the distinction limit and the limit of detection for each element are given.

The rare earth elements (REE) together give guite complex X-ray emission spectra with a considerable number of overlaps at analytical energies by lines of other REE with lower atomic numbers. Where the concentration of REE is high, as in lanthanide minerals, this Interference is more difficult to rectify. Smith and Gold resolved a similar problem with lower atomic number elements in energy dispersive microprobe analysis by establishing a series of overlap coefficients. They asserted that accurate corrections were necessary because of the relatively poor overall resolution of the instrument and that these should not be limited to the major coincidences. Some of the smaller values had probably been ignored because they were considered statistically Insignificant. The mathematical matrix of Smith and Gold covered 22 elements from fluorine to barium, with intensity coefficients (other than intraelement) quoted from 0.01% to 282.1% and with ZAF corrections necessary in cases of K to L conversion. The overlap coefficients were also adjusted for matrix effects.

Modern on-stream X-ray equipment for slurry analysis is now used in hundreds of mines throughout the world. This technique provides valuable information for process control which cannot be achieved by other analytical methods. One of the most recent successful examples is the Brenda Mines, in Canada, where the ARL Process Control X-ray Quantometer (PCXQ) is used for ten different streams in a Cu-Mo concentrator.

The success of slurry analysis is highly dependent upon the analytical method selected, based on the following factors: (i) the features of the analyzed materials, (ii) the concentration ranges, (iii) percent of solids, and (iv) requirements of analytical accuracy and precision. A preliminary investigation of methods and of analytical conditions can be done on a laboratory X-ray spectrometer using solid models of slurry. Such preliminary work significantly simplifies the introduction of the new technology and helps in the preparation of proper specifications.

Development of matrix correction procedures in the field of X-ray spectral analysis leads usually to “universal” algorithms which allow use of a common calibration equation for assay of various materials with different composition and concentration ranges. In many applications, such methods can guarantee accuracy of 3-5% rel.

For X-ray analysis of automotive catalysts, one such technique is based on the theoretical correction of interelement effects. The article states that “the overall accuracy of this approach is about 10% relative error, although this may vary depending upon the particular specimen being analyzed.” However, narrow specification limits for the quality control of catalysts require the total error of analysis less than 1-1.5% rel.

Applications of some selected elemental determinations of different catalytic materials by radioisotope excited x-ray fluorescence will be presented. The analytical method, an improved sample preparation and analysis technique were presented at the last Denver Conference (1). Data on the accuracy and precision of each of the following types of catalysts are presented. 1) Platinum catalysts (0.5 - 10% Pt) on alumina and carbon supports used for dehydrogenation, hydrohalogenation, oxidation processes, etc. 2) Palladium catalysts (0.5% - 10% Pt) on alumina and carbon supports employed for alkylation, decarbonisation, hydrogenation, etc.

Distribution of cesium and silver ion-exchanged onto columns of zeolites may be measured directly and non-destructively on the column with the use of radioisotope-excited energy dispersive X-ray fluorescence. Quantitative accounting of the exchanged metals is obtained without elution.

An x-ray secondary target method for routine determination of gallium, iron, and nickel in plutonium metal is described that has significant advantages over wet chemical analysis. Coupons requiring minimal preparation for analysis are produced as a breakaway tab on the plutonium ingot. All three elements are determined on the same coupon. Gallium is determined using an arsenic secondary target followed by iron and nickel using a zinc target. The analysis times are 5 minutes for gallium and 15 minutes for the combined iron and nickel. The method of analysis was evaluated in the range of from 0.5 to 1.5% gallium. Iron was investigated over the range of 67 to 3000 ppm and nickel from 64 to 110 ppm.

The analysis of copper alloys is a classical case for X-ray fluorescence as an analytical technique. For many years, large wavelength dispersive X-ray spectroraeters have been used in process control associated with the manufacture of brasses, bronzes and coinage alloys. For the last two to three years, a new, benchtop WDX multichannel spectrometer, the Chem-X, has been successfully introduced for the same purpose.

The need for rapid and accurate analysis of lubricating oil additives is firmly established by the petroleum industry. The application of energy dispersive X-ray fluorescence spectrometry techniques to this analysis is a valuable aid in achieving reliable and fast quality control of these product types.

Due to the large concentration variations possible for the elements of interest within these materials (principally Zn, Ca, P, S at 0-20%), interelement effects can become significant. In order to correct for interelement effects, historical XRF methods have relied on a large number of standards for each element to be determined or matrix matching with standards closely bracketing the unknown sample concentrations.

One of the most important applications o£ XRF in the oil industry has been in the analysis of oil additives which range from Mg to Ba. In such a low atomic number, hydrocarbon matrix, small variations in the additives cause pronounced changes in the average atomic number of the matrix and hence give rise to strong interelement effects. In recent years a number of mathematical methods have been developed to correct the measured X-ray intensity of the analyte litie for these considerable effects. This paper describes a method for the determination of Zn,Ba,Ca,CI,S,P and Mg in oil additives using a concentration based correction model in which influence coefficients (alphas) are applied.

To ensure that workers in certain industries are not exposed to harmful levels of toxic chemicals, it is necessary to provide regular monitoring of the concentrations of chemical contaminants in the workplace air. In the United Kingdom, monitoring is normally carried out on a routine basis by the factory occupier backed up by periodic visits from the Factory Inspectorate acting on behalf of the Government. The main source of guidance for occupational hygienists in assessing conditions in a factory is the list of threshold limit values (TLVs) published annually by the American Conference of Governmental Industrial Hygienists. Threshold limit values refer to airborne concentrations of substances and represent conditions under which it is believed that nearly all workers may be repeatedly exposed day after day without adverse effect.

The Technical Support Staff is called upon for analysis of a wide variety of sample types many of which have little sample history. However, it is usually necessary to account for all elements present. For these reasons, x-ray fluorescence spectrometry (XRF) has been a useful tool. Unfortunately, XRF requires the use of a range of standards for each element, the preparation of which could become so time consuming that the advantages of XRF would soon be diluted. Consequently, the utility of the J. W. Criss fundamental parameters computer program was evaluated for samples in which only one standard per element was used and where the standard matrix did not strictly resemble the unknown matrix. Some of the results of these tests on environmental assessment samples are reported here.

Particle induced X-ray emission (PIXE) analysis is well established as a multi-element, non-destructive technique to measure the trace element content of thin samples (≤ 1 mg/cm2). The wide range of its analytical applications was shown in numerous reports at the 2nd International PIXE Conference in Lund. For thin sample analysis routinely done at the Marburg PIXE facility, we calibrated our spectrometer empirically at proton energies of 2 and 4 MeV using two separate sets of calibration standards, one purchased from MicroMatter Co., the other prepared by precipitate exchange. Both sets of standards were checked by AAS and CMP. K-shell ionization cross sections, calculated from our measured x-ray yields, agree very well with recent literature values, and will be reported elsewhere.

A model is under development at this laboratory to predict the mobilization of Th from an estimated 20,000 metric ton deposit, in place and weathering as long as 80 million years on the Morro do Ferro, a hill on the Pacos de Caldas plateau in Minas Gerais, Brazil. The significance of this deposit lies in the marked similarities, chemically, between Th and Pu with reference to the problem of Pu waste management from the nuclear fuel cycle. One of the objectives of this program is to determine the elemental composition of the Morro do Ferro soil and its relationship to the mobilization of Th from the deposit. X-ray fluorescence is used for the analysis of Th, U, Ba, La, Ce, Pr, Nd, Sm, Dy, Eu, Gd, Yb, Hf , Fe, Zr, N b , Mo and Pb. This paper describes the XRF system and methodology used for the determinations and presents the results of 72 core, soil and sediment sample analyses for the above 18 elements. Analyses of three United States Geological Survey (USGS) rock standards are also presented.

The technique of x-ray fluorescence is ideal for the analysis of metal fumes such as are produced in various welding and brazing operations. Although the constitution of welding and brazing fumes is dependent on the nature of the base metal, flux, and welding material used, they typically contain the oxides of iron, nickel, chromium, and manganese (stainless steel welding). The fumes are in the form of very small particles, usually 1.0 um, and are often agglomerated into long chains. Filter samples of these fumes, as are used in industrial hygiene monitoring, may contain 30-300 ug/filter of particles on one surface of a low-mass, 37-mm membrane filter.

An effective and practical computer algorithm has been developed for rapid and precise phase identification of polycrystalline materials by X-ray diffraction methods using the JCPDS database and/or user created standard files. The entire JCPDS file was reorganized for efficient search. Identifications are facilitated by a number of options: automatic correction of systematic errors using internal standard reflections, selectable window widths for file searching, elemental restrictions (chemical prescreening), handling preferred orientation, match without using intensities, match with 3 reflections, and others. A comprehensive algorithm for calculating a figure-of-merit (FOM) is used so that the “correct” phases can easily be identified with highest FOMs. This method has been tested extensively on a wide variety of analyses and is applicable to either a host or a minicomputer.

The operation of a new polycrystalline phase identification method using the IBM Series/1 minicomputer is described. Data of the unknown can be entered by automatic transfer of previous runs, stored data sets and manually. Full screen menu selections are provided to facilitate operations and correct entries. Typical S/M time for a multi-phase inorganic mixture containing 43 reflections using a 0.3° window averaged 11 sec per 100 standards and with simple chemical prescreening less than 4 sec including program initialization and calculations of comprehensive figures of merit. Interactive options provide graphics terminal comparison of the unknown pattern with selected standards which appear as diffractometer patterns, subtraction of identified standards from the unknown and others. Utility programs permit storing data sets for later analysis, user created files and a program to display any file standard as a diffractometer pattern.

Over the past several years there has been considerable interest in computer search/match programs for qualitative analysis of powder diffraction patterns. This interest has been stimulated by the availability of modern minicomputers supported by relatively inexpensive mass storage devices capable of containing the entire JCPDS (l) data base on line. As the traditional search/match algorithms have been reviewed for possible implementation on the slower speed and restricted memory minicomputers being supplied with today's automated diffractometers, new ideas have emerged for such algorithms. One very extensive set of new algorithms has been developed by our group and these are contained in the SANDMAN search/match/identify program which was described at this conference last year (2). Experience has shown those algorithms to be extremely effective, particularly in handling eases where the presence of systematic errors in the data has precluded the correct analysis by other computerised search/match systems.

X-ray diffraction to determine the phases present in complex specimens generally proceeds by acquiring a pattern of D-spacings and intensities from the sample, which is then compared to patterns from a series of known or standard phases. The process of searching through large data bases to identify matches is too time consuming for manual methods, and so computerised search/match programs have come into use. The magnitude of the general problem, in which perhaps 40,000 known patterns (corresponding to the present size of the JCPDS powder diffraction file) are involved, places it in the realm of large nainframe computers, which are often accessible on a time-share basis by many laboratories or researchers using XRD. On the other hand, in many practical applications, the search need only be carried out over a limited range of compounds which are expected or may possibly be present based on the known history of the material, or on its independently determined composition. In this case, the data base becomes much smaller, and can be accommodated In micro- and minicomputers. The system we report here utilizes a 64K-byte 8 -bit microcomputer (Apple 11+) with dual floppy disk drives. One 5-1/4 inch disk can hold the pattern information for fifty compounds, and can be searched for possible matches in a few minutes. The same computer can control the scanning diffractometer to acquire the pattern data, and can act as an intelligent terminal to larger time-share machines when it is necessary to access the large data bases.

The real-time x-ray powder diffractometer control system AUTO incorporates several advances in data collection and analysis. Counting procedures for selected area data collection are optimized to achieve either a preselected statistical error in minimum time or a minimum error in fixed total time. Run files are employed to greatly simplify quantitative analysis procedures and for controlling repetitive runs. External calibration curves for 20 are used to eliminate all but sample dependent aberrations to peak positions. A generalized data file structure is used to document the instrumental variables and sample parameters.