Figure 2.9 PW 1540 from Philips.
Significant progress could be made by the introduction of electronic data processing, which began in the 1970s. At the beginning, it was mainly used for complex data processing, and later also for the control of instrument functions. As a result, XRF became much more powerful, not only regarding the analytical performance but also in regard to the ease of use, the sample throughput, and the ability to automate the analytics. This was of extraordinary importance for the acceptance of X-ray spectrometry as an analytical method and its wide use in quality control in industrial production. A further important step was the integration of automated sample preparation, measurement, and instrument monitoring for the improvement of analytical quality and the reduction of the subjective influence of human work.
These developments were mainly carried out by the companies mentioned above, some of which in the meantime underwent changes in ownership, but even now with the new name they are the current major players, such as Bruker (formerly Siemens), Malvern Panalytical (formerly Philips), and Thermo Fisher (formerly Bausch & Lomb). Parallel developments were also made in Japan – here mainly Rigaku has to be mentioned.
Another important development step for XRF was the availability of effective ED solid-state detectors. These Si- or Ge-based detectors, first used for γ-spectrometry, continuously achieved improved energy resolution through the development of low-noise signal electronics, the cooling of the detectors down to −200 °C, and the refinement of the manufacturing technologies. Owing to the improvements in energy resolution they were able to be used for lower radiation energies, and lastly even for the X-ray energy range.
These detectors were first used for X-ray microanalysis in electron beam instruments. Up to this time, the fluorescence intensity generated in an electron microscope was too low for WDSs. Therefore, X-ray micro-analyzers had to be built with a sufficiently high beam current. The high electron intensity however saturated the electron detectors quickly, increased the volume analyzed, and thus reduced image resolution. As a result, these instruments did not have good imaging quality and often had to be operated in parallel to electron microscopes with their good imaging function. For ED detectors, however, the fluorescence intensity in a scanning electron microscope is sufficient. This has made it possible to combine the good imaging function of electron microscopes with the analytic function of X-ray spectrometry in one instrument.
The manufacturing of these detectors required special technologies, which were only developed by a few companies, primarily in the United States by EDAX and KEVEX, and by Link in the United Kingdom. These companies also manufactured and sold the first XRF instruments with ED detectors. The detectors again had to be cooled with liquid nitrogen to −200 °C, which was complicated by the required Dewar vessel and the handling of liquid gas. One of the first-generation instruments, the KEVEX Analyst 0700, is shown in Figure 2.10.
EDSs were first mainly used for screening tasks, i.e. for qualitative analyses. In the course of time, the analytical performance of these ED instruments could be significantly improved. This was possible not only using special excitation geometries (grazing incidence, use of monochromatic or polarized radiation), but also by the improvement of the performance of the detectors (energy resolution, pulse throughput) and the improvement of the data evaluation methods. The precise determination of the line intensities, for example, by fitting or deconvolution methods, was of particular importance, since the limited resolution of ED detectors causes line overlaps, which may limit the accuracy of the analysis. On the other hand, improvements in quantification were possible by the availability of the entire spectrum including the background.
The availability of silicon drift detectors (SDDs), with their high count rate capability, gave EDS another boost, since now statistical errors could be achieved comparable to those of WDS instruments. The development of ED spectrometry was especially advanced by companies such as Spectro, Oxford, and Thermo.
A special application for X-ray spectrometry is the analysis of layered materials. Here, we have special analytical conditions, that are reflected in the measurement equipment as well as in the evaluation routines. The measurements are carried out usually on finished products, on which layers have been applied for decorative or functional purposes. This means that the samples are rarely flat and homogeneous over a large area, as it is necessary for conventional XRF. Therefore, the analysis must be carried out on small sample areas. This requires collimation of the exciting beam, thereby reducing the excitation intensity. The intensity loss associated with collimation must be compensated by large solid angles for the detection of the fluorescence radiation. Therefore, mostly ED instruments are used for coating thickness measurements. The associated loss of spectroscopic performance is acceptable, since layer systems usually contain only a few elements that are even known, since this analysis is usually carried out as quality control of the coating process.
Figure 2.10 Analyst 0700 from KEVEX.
As a result of the different analytical tasks, but primarily due to the different equipment requirements for WDS and EDS instruments, the development of coating thickness analyzers was for a long time independent from that of conventional X-ray spectrometers. The main companies were UPA, TwinCity, and CMI in the United States, Helmut Fischer in Germany, and Hitachi and Seiko in Japan. In the meantime, companies that had previously concentrated on the development of WDSs have recognized the high demand for coating thickness analyzers and therefore have also focused on this application. This development is also supported by the increased use of EDSs for the analysis of bulk samples at these companies. On the other hand, the companies, which were once primarily focused on layer analysis, are increasingly turning to applications on compact samples, for example, the analysis of precious metal in jewelry, the determination of toxic elements in consumer products or toys, i.e. samples which are finished products and therefore are not homogeneous or do not have flat surfaces. As previously mentioned, these measurements require small sample areas. Fischer and Bruker in Europe, Hitachi in Japan, and Bowman in the United States are currently the leading companies for this type of equipment.
A continuous trend in the development of X-ray spectrometers has always been the reduction of instrument size. The first commercially produced X-ray spectrometers were filling a complete room, in particular due to the elaborate electronics. Today, a table is often sufficient for setting up the instruments. Further, increase in the integration of the electronics has not only resulted in an increase in performance but also in a significant reduction of instrument footprint.
A significant step in this direction is the use of instruments that can be held in one hand. This development began shortly after the turn of the millennium with the demand in the United States for the determination of lead pigments in wall paints. This task was important because lead as a toxic element was to be identified, in order to be replaced in wall paint. The requirement was that the analysis is done on site in order to avoid the logistical effort necessary for the analysis of pieces of paint in the laboratory. The first instruments still operated with radioactive sources for excitation; they were later replaced by low power tubes. Owing to the very small distances between radiation source, sample, and detector the tube power can even less than 5 W. This not only assures a higher safety of the instruments, but also a good analytical performance and higher flexibility. The integration of powerful computing technology has further also contributed significantly to the improvement of performance. In the meantime, handheld instruments have been used for a wide range of analytical tasks. The possibility of on-site analysis has played a decisive role in the selection of analytical tasks.
Typical