In recent years X-ray computed tomography (CT) has become an important method for generating 3D volumetric data from specimens in medicine and in non-destructive material testing (NDT). The quite simple principle of a CT setup offers the realization of different magnifications according to the given problem (Fig. 1). With a state of the art CT system spatial resolutions in the range of 1 µm are easily achievable.
Such resolutions, however, are not sufficient for some problems, e.g. the analysis of porosity in limestone. The measured porosity depends on the resolution of the imaging system due to the fact that pores with sizes below the resolution limit are simply dismissed and not taken into account when calculating porosity (Fig. 2).
Methods that offer higher resolution are of course available, such as Scanning Electron Microscopy (SEM) or Focused Ion Beam Tomography. These methods, however, can only give information about the specimen’s surface (SEM) or destroy the specimen in the process of scanning layer by layer. If the use of a non-destructive X-ray method is still necessary, it is possible to use X-ray optics. These come with a difficult handling and high initial cost. Another possibility is to push the limits of geometric magnification in conventional cone beam CT down to higher resolutions.
The limiting factors for the resolution of a cone beam CT setup are especially the stability and precise repeatability of the positioning system and the physical size of the X-ray source spot (Fig. 1) . This is the main focus of our work at the Chair of X-ray Microscopy, especially the development of transmission and reflection target to achieve very small source spots.
For the X-ray Microscope XRM I located at the Fraunhofer Development Center EZRT in Fürth (Fig. 3) we develop thin-film transmission targets that are optimized for either high brilliance, i.e. a high flux per spot size, or a high integral X-ray flux. For these optimizations the processes that contribute to X-ray generation inside the target have to be understood as well as scattering processes. To calculate optimal target thicknesses these processes are modeled in an event-based Monte Carlo-simulation toolbox. Thus, the influence of layer thickness on X-ray intensity and spot size can be studied (Fig. 4). Together with the Nanotechnology Service at Würzburg Üniversity we are working on micro-structured targets to limit the X-ray source spot by the physical extent of the target.
The X-ray Microscope XRM II in Würzburg makes use of the fact that the actual target size limits the X-ray source spot. Wires of suitable material (usually tungsten or molybdenum) undergo certain electro-chemical and micro-mechanical treatments so that the result is a very fine tip with radii down to 100 nm (Fig. 5). The use of these needle-like targets has the advantage that the use of different targets for different tasks (High power/high resolution) is no longer necessary, because the extent of the X-ray spot can be influenced by the site of the electron focal-spot.
Development of suitable resolution tests
For the characterization of the CT set-up and correct determination of the resolution in both projections and 3D imaging we are developing resolution test phantoms together with the Nanotechnology Service. With these tests the resolution of a system can be determined over several orders of magnitude.