The beamline ID19 is mainly devoted to X-ray imaging. Across a specimen, the various implemented techniques detect the spatial variations of features that affect the X-ray beam, viz. absorption, scattering, or the optical path-length.

Bragg diffraction imaging

Methods for imaging defects in single crystals using Bragg diffraction ("topographic" methods) were pioneered half a century ago, and gained momentum in the late 50’s and 60's, on the basis of laboratory X-ray generators. Diffraction imaging maps, in real space (as opposed to reciprocal one), the spatial variations of the direction and/or amplitude of the beam diffracted by a single crystal. These variations are usually associated with distortion and/or distortion gradients related to the presence of defects, domains, phases, ... which can be, in this way, visualized.

Diffraction imaging is non destructive and rather sensitive: a wide range of distortions (10-2-10-7) associated with an angle variation Dq and/or a lattice parameter variation Dd/d, can be observed. The diffracted beam is recorded either on X-ray films or with a CCD based detector. The film or nuclear plate resolutions range from about 1 to several µm (over a very large field of view) and that of digital cameras is ? 1 - 30 µm (over a more restricted field of view, which is proportional to the required resolution and the number of pixels available).

The availability of third generation synchrotron radiation facilities provides new experimental possibilities in diffraction imaging. Some of them are straightforward: the availability of high energy (50-120 keV) photons allows the investigation of heavy and/or bulky samples in transmission, and the photon flux makes real-time experiments on the 10-2-10 s time scale feasible. This time scale can be dramatically reduced in stroboscopic experiments, performed either with an external chopper or using the pulsed structure of the beam. Another feature of modern SR facilities, the small angular size of the source as seen from a point of the sample, or effective divergence of the incident beam, a = s / L (~10-6), where s is the source dimension (s ~ 0.1 mm) and L the source-specimen distance, has very important consequences: details in the topographs can be observed with acceptable blurring at a sample-to-detector distance of the order of a metre, the geometrical resolution remaining below one micrometer. This allows the use of the crystal-to-detector distance as a new parameter to characterize domains and defects. The fact that a is small is also associated with a high degree of coherence of the X-ray beam.

A variety of investigations were performed. Using the white beam directly, it was possible to visualize in real time the effect of an external parameter (stress on the movement of dislocations, applied magnetic field on a phase transition, applied electric field on a unidimensional ionic conductor, …) as well as to perform stroboscopic investigations of periodic effects, such as acoustic waves. The growth of grains, without a priori knowledge of their orientation, can be followed in-situ, either from the melt, or in solid-state recrystallization. In the monochromatic beam version, fine aspects of the lattice distortion were investigated in semiconductor materials. A recent development takes advantage of the coherent, spherical wave, characteristics of the beam, which allows an original investigation of the strain field around defects in nearly perfect crystals [9].

X-ray computed microtomography

X-ray computed microtomography (µCT) consists in recording a number of projections from an object, with different angle of views, and reconstructing from these projections a 3D image with the help of an adapted algorithm. The ESRF allows improvements and new possibilities for µCT, because of :

  • a high photon flux in a homogeneous, parallel, monochromatic, and highly coherent beam (polychromatic, divergent and incoherent on laboratory setups);
  • recording of 'phase images' obtained by adjusting the sample to detector distance ('propagation technique');
  • adapted detector (FRELON camera, already described);
  • implementation of efficient reconstruction procedures.

The parallel and monochromatic beam allows to record absorption images and reconstruct the volume with an improved spatial resolution (up to 1 µm) of bones, foams, building materials, metallic alloys, rocks ... The present evolutions are :

  • to perform quantitative processings taking benefit of the monochormaticity of the beam (e.g. bone densitometry);
  • the implementation of 'local tomography', i.e. the high resolution reconstruction of a region of interest within a matrix, which is only reconstructed at a low resolution. This is a crucial point for many applications where it is not possible to extract a small sample from its matrix.

The use of coherence led to the development of 'phase' tomography, which dramatically extends the possibilities and application range of the method. Phase tomography is a unique technique to investigate features like holes in bulk quasicrystalline grains [10] composite materials where the various components display very similar X-ray attenuation, or damage assessment in the bulk of a non transparent material. The images are often recorded in the 'edge detection' regime, a case where the usual filtered back-projection algorithm used for absorption tomography appears to be acceptable approximation.

Another, more quantitative, approach is the retrieval, from a set of images recorded at several distances, of the optical phase. Algorithms, developed for electron microscopy, were successfully adapted to the X-ray case. The phase retrieval, combined with tomography (‘holotomography’ technique [12]), leads to high quality quantitative 3D results.

High resolution diffractometry

High resolution X-ray diffractometry is the measurement of rocking curves ("classical" meaning) and two-dimensional reciprocal space maps (extended meaning). From a general scattering point of view, the reflectivity curves and reciprocal space maps near the origin of the reciprocal space may be regarded as special cases of this technique. High resolution X-ray diffractometry is usually an ‘integral’ technique: the detector records the intensity of a beam originating from a certain volume of the sample, for different angular positions (or reciprocal space positions), providing intensity curves and maps. An integration over the illuminated volume is therefore performed, and we have no direct access to the direct space co-ordinates like in X-ray topography.

The importance of rocking curve measurements is based upon two fundamental properties :

1) The details of rocking curves are sensitive to the strains and strain gradients in the specimen.
 
2) For a given structural model, the rocking curve may be computed with high accuracy using fundamental X-ray scattering theory.

 

These measurements are nowadays widely applied, e.g. to the quality control of multiple-layer semiconductors. The routine analysis of such rocking curves gives the composition of ternary (or quaternary) epilayers, periods of superlattices and thickness of layers, whilst more advanced analysis can provide a complete strain and composition profile as a function of depth. The recording of two-dimensional reciprocal-space maps allows to obtain more detailed information on thin layers and the surface and interface conditions.

The new features for high-resolution X-ray diffractometry at the ESRF are :

  • a) in-situ measurements with high (stroboscopic measurements) and medium (time scale of seconds and minutes) time resolution.

    b) the possibility to use a sample environment under high resolution conditions.

    c) the combination of high-resolution diffractometry with topographic measurements.

    d) the extensive use of two-dimensional detectors (FRELON camera ‘local’ high-resolution diffractometry)

These new features allowed to investigate, for instance :

  • the propagation of surface acoustic waves excited on GaAs [13] or LiNbO3,
  • the dynamics of phase formation in thin yttrium-hydride films [14],
  • the in-situ characterization of the lattice strain distribution in broad area high power semiconductor laser devices,
  • the combined topographical imaging and local rocking curve measurements (position and width) of entire semiconductor wafers for quality control [15].