Wilhelm Conrad Röntgen was born on 27-03-1845, 150 years ago, in Lennep- Remscheid in Westphalen, Germany, from a German father and a Dutch mother. He remained an only child. When he was 3 years old, the family moved to Apeldoom in the Netherlands. His parental home exists and will be the Röntgen bar. After going to school in Apeldoorn, he went to a technical school in Utrecht, where he finished in 1865. He enrolled in the University of Utrecht in the Netherlands, but found that he did not have the right certificates to pass the examinations. He then went to Zurich to the newly stifted Eidgenössische Technische Hochschule, where restrictions were less severe. In 18869, he got his doctorate on ‘Thermodynamics of gases.’ He was a brilliant student with a wide variety of interests.

While both X-ray Fluorescence Analysis and X-ray Powder Diffractometry have their roots back in the earlier half of this century it wasn't until the 1960s that the two techniques became widely accepted. The growth in the application of X-ray methods for materials analysis grew rapidly between 1960 and 1970, then gained another major leap forward in the early ‘70s with the introduction of mincomputers. The introduction of Si(Li) detectors in the late’ 60s and early ‘70s added a further dimension to the available instrumentation. This paper reviews the growth in the field of X-ray materials analysis and highlights the major mile-stones in intrumentation and techniques.

Software has come a long way from the days of plug board wiring and multi-step computations which were characteristic of the 1950s, In the early days, hardware development led software development partly because the evolution of the hardware was so rapid and partly because the software was primarily developed for specific units. In those days the major cost was in the hardware and the software development generally involved only salary for relatively few individuals. Modern computer costs are now within the reach of the individual, and the money that goes into programming and programs is now the major cost.

This paper reviews some of the steps in the evolution of software as it affected the analysis of powder diffraction data. Many “new” programs are adaptations of earlier routines to a different computer platform, but there are many truly new programs implementing inproved concepts for extracting crystal information from the digitized diffraction data.

In the 100 years since the landmark work of Wilhelm Conrad Röntgen, much progress has been made in X-ray science and applications. X-ray Fluorescence Analysis (XRF) essentially had its beginning not long after that development, when, in 1913, Moseley demonstrated the relationship between the atomic number of an element and the wavelength of its X-ray spectra. As early as the 1930's, workers were analyzing materials for trace elements. By the time of the 1940's, X-ray spectroscopic instruments were becoming commercially available, and analysts were beginning to appreciate the simplicity of the X-ray spectra, and thus the potential for a significant impact on elemental analysis.

Soon after Von Laue's experiment of X-ray diffraction in 1912, a similar experiment was conducted in Japan by Prof. Terada at University of Tokyo. He made direct observation of Laue spots from rock-sault on a fluorescent screen. In 1914, Prof. Nishikawa also at University of Tokyo photographed Laue spots of spinel. The X-ray generator employed was one designed for medical use. The first X-ray diffractometer in Japan was fabricated in the 1920s at the Institute of Physical and Chemical Research, It was designed with reference to Bragg's spectrometer using an ionization chamber as the X-ray detector.

In 1932, X-ray generators and cameras were manufactured by Rigaku Denki Mfg., the predecessor of the present Rigaku Corporation, Shimazu Mfg. and some other few companies. X-ray tubes at the time were Coolidge type demountable tubes.

To facilitate orientation in time, some selected events will be briefly presented. Approximately five hundred years ago, Columbus discovered America. One hundred years ago, on November 8th, 1895, Wilhelm Conrad R6ntgen discovered the X-rays which in the German language are called after him. In 1912 Max von Laue conducted the first X-ray diffraction experiment. In 1892 Richard Seifert Sr. founded the Electrotechnical Plant in Hamburg. After World War I (1914-1918) the company founder gradually handed the firm over to his son Richard Seifert Jr. After his son had completed studies in physics and electrical engineering he conducted pioneering experiments on the application of X-rays in science and technology. From the very beginning, X-ray equipment was produced in the three fields of medicine, science and technology. It was only ten years after World War II (1939-1945) that the line of medical equipment was discontinued and the daughter [1] as a member of the third generation gradually took over executive management tasks.

Periodic multilayers are ideally suited as high-reflectivity and wide-bandwidth Bragg reflectors. Their period can be matched laterally to the incidence angle so that for all points on the reflector, Bragg reflection is obtained for the same wavelength. Three major types of laterally graded multilayer optics were appJied to X-ray diffraction: (i) Parabolically curved multilayer mirrors were used to convert divergent radiation emerging from an X-ray source into a parallel beam. The parallel beam was applied in powder diffraction, grazing incidence diffraction, reflectometry, high-resolution diffraction, and protein crystallography, (ii) Elliptically curved multilayer mirrors focused the divergent radiation from the source into a line on the sample or detector. The high brilliance and small dimension of the focused beam make this mirror type suited for transmission diffractometry of capillary and fiber specimens, (iii) Planar multilayer mirrors were employed in divergent-beam optics. In Bragg-Brentano diffractometers, this mirror type can serve as a compact incident-beam monochromator for removing Kβ lines and Bremsstrahlung.

Large area, parallel beams of x-rays are potentially useful in many diffraction, imaging and other x-ray analysis applications. Diffraction, in typical experiments, uses only a small portion of the incident beam that is within a limited “rocking curve” of the crystal capable of diffracting, although the whole crystal may be illuminated by the incident beam. Signal intensity can be increased if the ”whole” crystal can be made to diffract simultaneously. Similarly for imaging, improved results also are possible if divergent beams are replaced with nearly parallel ones. Production of parallel beams of large areal extent has not been simple. Often this meant that the source had to be placed at great distances from the sample, reducing the incident intensity. Sometimes, asymmetric cuts of crystals can be used to increase beam cross-sectional areas.

Production of parallel beams of large area! extent has not been simple. Often this meant that the source had to be placed at great distances from the sample, reducing the incident intensity. Sometimes, asymmetric cuts of crystals can be used to increase beam cross-sectional areas.

Simulation and experimental work that compare the performance of straight and tapered monocapillaries when used with laboratory x-ray sources are reported. Detailed simulations for various taper profiles give several important conclusions for optimizing the design of a tapered monocapillary. Several tapered monocapillaries were prepared. With a 16W x-ray source, beam intensities of 4×105 photon/sec/μm2 and 3×105photon/sec/μm2 of Cu Kα x rays were obtained from the tapered monocapillaries for output diameters of 8μm and 3.5μm, respectively. These intensities are 1.4 and 1.5 times that obtained from straight capillaries with the same output beam sizes at the experimental set-up optimized for a straight capillary. In addition to the gain in x-ray flux, the tapered monocapillaries produce output beams with significantly reduced high energy bremsstrahlung radiation and increased flux stability with respect to shifts of the x-ray source spot.

Grazing incidence diffraction (GID) is a powerful tool for the structural characterization of thin films. Unlike traditional Bragg-Brentano geometry (divergent X-ray beam, focusing geometry), GID experiments are enhanced by a parallel beam of high intensity. Classical conditioning of the X-ray beam is done by using small slits on the primary beam side and a long Soller slit in combination with a flat crystal (e.g. graphite, lithium fluoride or germanium) or an energy dispersive detector on the secondary beam side. However, new alternatives for beam conditioners are becoming available which promise increased performance. X-ray beam optics using either planar or graded parabolically curved multilayer mirrors of high reflectivity have been constructed for the primary beam side as well as for the secondary beam side.

We present a further improvement of our experimental system for the generation of hard x-ray pulses in the range 1-10 ps, A laser system based on a cw mode locked Nd:YLF laser is utilized for the emission of deep UV pulses at 193 nm with duration 1.5 ps and repetition rate of 300 Hz, This UV radiation is used to pump an x-ray diode which produces CuKa x-ray pulses with a duration shorter than 10 ps at 300 Hz.

Micron or submicron spatial resolution for x-ray microfluorescence analysis has been pursued using various x-ray point focusing methods,1-4 such as transmission zone plates, tapered capillaries, and Bragg Fresnel lenses. Point focusing using bent crystals does not seem to be a desirable means for achieving a microbeam size, although it is a traditional technique which utilizes perfectly curved crystal surfaces to focus x rays emanating from the source. It is known that point focusing can only be achieved by bending a crystal two dimensionally, which normally introduces imperfections to the crystal since the crystal is an undevelopable surface. Unavoidably, the introduced imperfections will broaden the reflection curve (rocking curve) of the crystal and, in turn, broadens the focal spot size.

X-ray focusing based on Bragg reflection at curved crystals allows collection of a large solid angle of incident radiation, monochromatization of this radiation, and condensation of the beam reflected at the crystal into a small spatial cross-section in a pre-selected focal plane. Thus, for the Bragg reflected radiation, one can achieve higher intensities than for the radiation passing directly to the same small area in the focal plane. In that case one can profit considerably from X-ray focusing in an EDXRF arrangement. The 00 2 reflection at Highly Oriented Pyrolytic Graphite (HOPG) crystals offers a very high intensity of the Bragg reflected beam for a wide range of photon energies. Furthermore, curvature radii smaller than 10 mm can be achieved for HOPG crystals ensuring efficient X-ray focusing in EDXRF applications. For the trace analysis of very small amounts of specimen material deposited on small areas of thin-filter backings, HOPG based X-ray focusing may be used to achieve a very high intensity of monochromatic excitation radiation.

To measure X-ray fluorescence radiation of low Z elements especially requires a very high excitation intensity with an energy as close as possible to the absorption edges. Furthermore, a monochromatization of the excitation radiation and a cut-off of higher energies is desirable in order to reduce the background in the spectral region of interest.

A cylindrical graded multilayer was designed and tested to concentrate and to monochromatize the radiation of a thin window Mo fine focus tube. The variation of the dspacing is tuned to reflect the Mo La-radiation of the fine focus tube. The physical dimensions of the cylindrical shape are optimized for concentrating the Mo La tube radiation onto the specimen with a spot size of about 3 mm. This corresponds to the size of the specimens from a time dependent personal aerosol sampler used.

The multilayer is characterized with respect to its reflecting, focusing, and monochromatizing properties. Two examples of low Z element analysis are given.

We have developed an algorithm for calculating the x-ray tube continuum based on the eqidistribution proposed by Love and Scott, extended the description of white and characteristic radiation given by Wiederschwinger et al for the energy range 10 to 50 keV to the low energy range from 5 to 30 keV, and compared the results from this algorithm to those responses obtained from algorithms using the absorption correction of Pochou and Pichoir, Philibert, Sewell and Pella. The comparison to other models showed a significandy better performance by our model.

A method for measuring spectral distributions of x-ray tubes directly with a solid-state detector is presented. Different anode materials (chromium, molybdenum, rhodium and tungsten) have been measured. Results for various applied voltages and take-off angles have been obtained.

A high-power X-ray generator equipped with a lanthanum hexaboride cathode has been developed for X-ray absorption fine structure experiments. A high tube-current of more than 1,000 mA can be provided when operated at low tube-voltage of less than 20 kV. In addition, the focal width is narrow enough (less than 0.1 mm) to ensure good energy resolution. Extremely intense monochromatic X-rays (106 ∼ 107 counts/(sec.mm2) at the sample position), which are completely free from higher order harmonics and tungsten contamination lines, are available, when a Johansson-type spectrometer is employed. The filament life has been significantly prolonged by the high vacuum specification of the tube.

A fluorescent and diffracted X-ray spectrometer with an X-ray sub-micron beam was developed for the analysis of stress, crystal structure, and contamination in micro-regions of ULSIs. A fine glass capillary with a parabolic cross-sectional inner wall surface was made to form a micro X-ray beam. A micro-focus X-ray generator with a membrane-type target was also developed for use with the capillary. The equipment was used to analyze the crystal phases of Ti silicide fine lines and the strains in Al interconnections.

The resistance of Ti-silicide lines increases the line width narrows under about 0.5 μrn. This is because of the growth of crystal phases of TiSi2(C49) and TiSi(B27) with high resistivity ﹛60 to 100 μΩ-cm).

Hillocks of Al grow at line edges when Al layers are compressed by Si substrates and surrounding insulating layers. On the other hand, when the Al is extended by these layers, voids form in the AI lines.

A polycrystalline thin film diffractometer for asymmetric diffraction at a fixed incident angle using parallel beam and high-resolution-parallel slits has been developed. Well collimated CuK α1 incident beam was obtained by coupling of a rotating anode generator with a channel-cut crystal monochromator. The diffracted beam was analyzed with parallel slits with 0.17° angular aperture. The observed profiles of CeO2 standard powder sample were analyzed by using an individual profile fitting technique. The minimum full-width at half-maximum (FWHM) was 0.15° in the 20 range of 40° ∼ 70°, and the variation of FWHM with 20 was almost independent of the incident angle. Intensity in asymmetric diffraction was enhanced compared to that in the symmetric 0-20 scan, and its variation was well in accordance with the theory. The use of the parallel slits compromizes the intensity and resolution for polycrystalline thin film diffraction in the present setup, but gives the high quality of diffraction data for structural study.

In modern technology, thin-layered materials with layer thickness in nanometer ranges have been utilized for various advanced components such as integrated circuits, magnetic heads and disks, X-ray mirrors and coated window glasses. For the analysis of such materials, powerful probes, fluorescence(TXRF)1), diffraction(TXRD) 2-4) and reflectivity(GIXR) 5-7), formed by X-rays in conjunction with total reflection phenomena can provide important information on element composition, crystalline structure, layer thickness, electron density and interfacial roughness.