Grazing-incidence X-ray analysis techniques which are commonly used for the nondestructive characterization of surfaces and thin films are reviewed. The X-ray reflectivity technicue is used to study surface uniformity and oxidation, layer thickness and density, interface roughness and diffusion, etc. The grazing-incidence in-plane diffraction technique is used to determine in-plane crystallography of epitaxial films. The grazing-incidence asymmetric-Bragg diffraction is used for surface phase identification and structural depth profiling determination of polycrystalline films. Typical examples to illustrate the types of information that can be obtained by the techniques are presented.

The structural characterization of thin films is important for research development and manufacturing of electronic, magnetic, optical, and other high-tech materials. The grazing incidence X-ray diffraction technique has bean used successfully for the determination of crystalline phases, structural-depth profiles, crystallite size, and strain, etc. of thin films with thickness's down to a few tens of Å, If the crystal structure, e.g. the distribution of atoms in the unit cell, or the crystallinity and texture (or preferred orientation) of a film is of interest, the conventional Bragg-Brentano diffractometer technique with the θ-2θ scanning geometry has been found to be appropriate.

The use of grazing incidence parallel beam x-ray diffraction (GIXRD) in the characterization of lead zirconate titanate (PZT) films is described. This tool has enabled us to depth profile the films. The transmission electron microscopy (TEM) results obtained from a cross section of one film are shown to compliment the GIXRD results. The variation in crystallographic structure versus depth in the film was the primary focus of this study.

The insults from three PZT films having Zr/Ti ratios of 25/75, 48/52, and 75/25 are given. TEM results are reported from the sample with a Zr/Ti ratio of 48/52.

Depth profile analysis of phase composition, residual strain and crystallite size in ceramic specimens can be determined by varying the information depth (ID) as defined by Delhez et al. (1988) for an x-ray opaque specimen. The information depth may be controlled by making use of the dependence of ID on Bragg angle and by selecting wavelengths which give appropriate x-ray absorption within the specimen.

Procedures are outlined for extracting depth profiles for (i) composition from intensity data and (ii) residual strain and crystallite size from peak width data. Results are described for both discrete-peak and Rietvetd pattern-fitting analysis of zirconiaalumina specimens for which preliminary results have been reported by van Riessen et al. (1990).

The evaluation and understanding of the x-ray diffraction properties of multilayer structures is critical in predicting the degree of success in their utilization in x-ray optics applications. A key material parameter affecting the diffracting efficiency of multilayer structures is the roughness of the interfaces between the deposited layers. This paper will use measured roughness to predict the integral reflection coefficient as a function of diffraction order for several multilayer structures.

We have made atomic force microscope (AFM) measurements of the detailed displacement distribution for me surface roughness of W/C and W/Si multilayer structures. This information has been introduced into a theoretical model to predict the reduction of the integral reflection coefficient induced by surface roughness. In a fashion analogous to the Debye-Waller factor, this reduction in diffracted intensity is predicted to be strongly dependent on the diffraction order. These predictions have been compared with experimental integral reflection coefficient data. The measured diffraction results indicate a greater reduction in reflection than is predicted by the AFM measurements. A possible explanation for this discrepancy is that the AFM measurements underestimate the roughness. Possible reasons for the differences between these methods will be discussed.

The pioneering work of Marra, Eisenberger and Cho (1979) demonstrated that powder diffraction from near surface and interfacial regions could be obtained using glancing incidence geometry. This has generated considerable interest in research at synchrotron sources, but also several slit attachments have been developed for commercial diffractometers. By adding a focussing mirror, we demonstrate here a simple arrangement that produces adequate intensity with a sealed tube to examine ultra thin films.

We have studied a series of laminated thin films by x-ray diffraction to correlate their structural and magnetic properties. Previous works have shown that the signal-to-noise characteristics of laminated magnetic films with non-magnetic interlayers can exceed that of single layer magnetic films. In the case where the magnetic layer is known to have low signal-to-noise performance, the signal has been observed to increase as the number of layers whereas the noise increases as the square root of the number of layers. This yields a net improvement of the signal-to-noise ratio, making the overall film more attractive as a magnetic recording medium. In this paper, we investigate films which are laminated layers of magnetic CoPtCr and non-magnetic Cr. The films were deposited with sputtering parameters that generally give low noise characteristics in single layer films. Up to four layers of CoPtCr films were made with Cr spacer layers of 2 nm. We observe some improvement in signal-to-noise characteristics and reduction in coercivity. X-ray diffraction analysis shows that the crystallographic c-axis, which corresponds to the magnetic easy axis, becomes more preferentially oriented perpendicular to the film plane with each additional layer. This change in preferred orientation is consistent with the reduced in-plane coercivity of the film. In the double CoPtCr layer with one Cr spacer layer experiment, we see that as the Cr spacer layer is increased from 0.5 to 8 nm, the c-axis of the CoPtCr again becomes more preferentially oriented out of the film plane, resulting in decreased in-plane coercivity. The media signal-to-noise improves once the Cr spacer layer is beyond 2 nm, consistent with previous observations.

This is a report of an investigation by XRD and SEM of thin coatings such as are applied by plasma spray methods for improving properties, such as are wanted in electronic devices, and for increasing durability in service. Samples were studied that had been coated under controlled conditions: Physical Vapor Deposits (PVD) of titanium nitride on steel plate and plasma spray coatings of molybdenum on steel plate. X-Ray determinations were made of x-ray lattice constants, crystallite size, lattice strain, orientation and residual stress, before and after strain was applied to the specimens.

Many analytical problems cannot be solved using a single XRD measurement. One example is the epitaxial growth of thin films of High Tc superconductors. To fully characterise such a sample it is necessary to study both the orientation (texture) and the phase composition of the layer.

Thin films of BaTiO3 and Pb(Zr53 Ti47)O3 (PZT) have been deposited on Pt and Pt-coated silicon substrates using both Nd-YAG and excimer lasers. The BaTiO3 films were prepared using heated substrates and were crystalline. The PZT films were deposited at room temperature and were amorphous; on annealing, they crystallized and gave rise to well-defined powder x-ray diffraction patterns.

To compare and correlate the properties and processing conditions of these thin films, characterizations were performed using a variety of analytical techniques including x-ray diffraction, TEM, SEM/EDX and ferroelectric and dielectric property measurements. The x-ray diffraction technique was used for identifying the various phases formed and also for analyzing the profiles of the diffraction peaks. Both the PZT films annealed below 800°C and the BaTiO3 films typically show polycrystalline x-ray diffraction patterns corresponding to a pseudo-cubic structure (i.e no peak splitting) instead of the tetragonal patterns characteristic of the target materials. It was found that for the BaTiO3 films the pseudosymmetry was due to crystallographic alignment of the longer c-axis In the substrate surface to relieve strain. In the FZT films annealed below 900°C, it is suggested that the residual surface strain and/or small crystallite size of these materials may have precluded the peak splitting; at higher annealing temperatures, the tetragonal symmetry was recovered.

A number of X-ray methods characterizing lattice defects are described. They were developed in response to a variety of challenging problems in materials science, A method based on a computer-aided rocking curve analysis, CARCA, was developed which offers a rapid mapping of the dislocation structure in an epitaxial film, as well as a tensor analysis of nonuniform elastic strains. Characterization methods were developed in an attempt at bridging, systematically, the gap between micro and macromechanics when the problem arose to clarify the distribution of elastic strains emanating from stress concentrators such as notches, cracks, holes and to elucidate strain interactions. Gradients of elastic strains were characterized by a method of local intensity measurements. For crystal with homogeneous elastic strain distribution a tensor analysis is described, based on precision measurements obtained by the backreflection divergent beam method. A direct linkage between the imaging of the micro structure by TEM and the macro-response of deformation and recovery of commercial alloys was achieved by a version of the CARCA method, designed to characterize the lattice defects in polycrystalline materials. Example applications of the methods are presented with the hope that their usefulness may find adaptations in other areas of investigation.

White beam transmission topography was used to image defect structures in gallium arsenide single crystal substrates representative of those used for microwave integrated circuit fabrication. A scanning topographic camera was developed for Beamline X-19C at the National Synchrotron Light Source. Acquiring film topographs within a particular range of scan rates and x-ray fluxes was found to produce distorted images. The distortions were shown to be thermoelastic, arising from inhomogeneous deposition of thermal energy through the absorption of residual x-rays. The problem could be circumvented using faster scans with some sacrifice in spatial resolution. Real time electrooptic topographs were recorded using x-ray fluxes about one percent of those used for the film topographs. The potential for reducing the flux requirement to even lower levels which would be accessible to flash x-ray generators is considered.

White beam synchrotron X-ray topography has been applied both to the characterization of two semiconductor heterostructures, GaAs/Si and InxGa1-xAs/GaAs strained layers, and a substrate to be used for growing semiconductor epilayers, Cd1-xZnxTe. In the case of the heterostructures, misfit dislocations were observed using depth sensitive X-ray topographic imaging in grazing incidence Bragg-Laue geometries. The X-ray penetration depth, which can be varied from several hundreds of angstroms to hundreds of micrometers by rotating about the main reflection vector, which in this specific case was (355), is governed by kinernatical theory. This is justified by comparing dislocation contrast and visibility with the extent of the calculated effective misorientalion field in comparison to the effective X-ray penetration depth. For the case of Cd1-xZnxTe, twin configurations are observed, and their analysis is presented.

The characterization of defect configurations in various perovskite-like substrate materials for high Tc superconductor epitaxial films has been conducted using white beam synchrotron X-ray topography. The substrates were found to contain crystal lattice defects such as twins, dislocations and grain boundaries. It is shown that characterization of substrates can potentially afford insight into factors controlling the properties of the high Tc superconductor tilms supported on them. This can help in the selection of optimum substrate material. Defect formation mechanisms in individual materials as well as their respective influences on the films are discussed. Comparisons between the physical and chemical properties of several potential substrate materials are presented.

Polycrystalline materials may either be compact such as metals, ceramics, or some polymers or they may be powder samples. In both cases preferred orientation of the crystallites must be taken into account. In compact materials, texture formation by solid state processes and the texture-property relationship are the main purposes for texture investigations. In powder diffraction, texture correction can increase the accuracy of the results. Also strongly textured samples may be used in order to seperate diffraction peaks which are coincident in random powders.

Texture, i.e. preferred orientation (PO), of crystallites can cause serious systematic errors in quantitative analysis of crystalline materials using x-ray powder diffraction (XRPD) data. The singleparameter model of March (1932), proposed by Dollase (1986) for use in powder diffractometry is a promising mathematical formalism for correcting PO in XRPD analysis of uniaxially-oriented specimens. O'Connor et al. (1991) successfully applied the March formula in applying preferred orientation corrections for gibbsites, Al(OH)3, using Rietveld pattern-fitting and a line ratio method in which corrections are determined according to the intensity ratios of selected lines. The paper gives an appraisal of the general applicability of the methods considered by Li and O'Connor with particular reference to powder diffraction data for gibbsite, molybdite (MoO3), calcite (CaCO3) and kaolinite specimens. It is shown that some caution should be exercised when using the March formula to describe PO in Rietveld analysis.

Texture (of a polycrystalline sample) means the orientation distribution of the many crystallites in the sample with respect to a coordinate system on the sample, for example the rolling direction. Texture is responsible for many macroscopic properties of materials, for example in metals and alloys in the process of rolling or hot-pressing. It carries importance in many everyday production processes. Also in areas of geological science the texture of the sample, e.g. of the rocks should be known in order to explain geological processes.

Texture, by definition, is a description of the spatial arrangement of the polycrystalline grains in solids. For metals, the grain distribution is commonly presented in a pole figure in which the pole density of a set of crystallographic planes is plotted using standard stereographic projection techniques. In this paper, the pole figure is plotted such that the center of the pole figure coincides with the plate normal. A vertical line drawn in the equatorial projection corresponds to the rolling direction, while a horizontal line corresponds to the transverse direction. Consequently, the spatial arrangement of the grains in a given sample may be determined from an analysis of the pole figure.

Accurate diffraction intensity measurement from powder samples with a modern automatic x-ray diffractometer is nowadays a routine work. Most of the difficulties are caused because the samples are not the ideal ones. Especially, the effect of preferred orientation (or texture) may drastically distort the ratios of intensities in the collected data. To diminish these effects more careful preparation of the samples is needed. This is, however, often tedious and its success is not certain.

For pole figure measurement the flat sample has to be tilted which leads to peak broadening and peak overlap in peak-reach diffraction spectra. Using a linear position sensitive detector /PSD/, complete diffraction spectra can be obtained in each sample orientation. Peak separation can then be done by Gaussian fitting. When measuring with a PSD, each diffraction peak corresponds to a particular diffraction vector. This has to be taken into consideration by a coordinate transformation. The PSD-method is the only practical method of texture measurement for materials with complex diffraction spectra.