Coherent X-ray Imaging (CXI)
In a spatially and temporally (i.e. monochromatic) coherent X-ray wave, all photons are in phase. To scatter such a wave from a specimen is an ideal starting point to extract information on the sample structure. In this situation, if we could measure the intensity and phase of the scattered radiation for all scattering angles, we would directly obtain a 3D image of the object under investigation. The problem, however, is that we cannot measure the phase of the x-rays, as the oscillations are too fast for any currently conceivable detector. CXI is concerned with recovering this phase information and thus generating an image of the object, ideally with diffraction limited resolution. With the increasing availability of coherent x-ray beams, these techniques are developing rapidly. Holography, coherent diffraction imaging and ptychography are schemes that are increasingly used for nanoscale x-ray imaging.
In holography for example, a reference beam is used to encode the phase via interference with the object beam1. For example, a small reference aperture next to the sample can be used to provide a reference beam. The resulting coherent diffraction pattern is a hologram, and the image of the object can be obtained by direct Fourier inversion (FT). In the figure below this is moreover done in resonance with a Co core level excitation (778 eV), providing x-ray magnetic circular dichroism (XMCD) contrast for the magnetic domain structure of the sample1. Pure magnetic domain contrast is then obtained by looking at the difference signal from opposite x-ray helicities.
Reprinted by permission from Macmillan Publishers Ltd: Nature 432, 885, copyright (2004).
Iterative phase retrieval techniques such as coherent diffraction imaging and ptychography rely on additional information to recover the phases of the scattered radiation via iterative algorithms. These boundary conditions can be quite simple (e.g. that the sample has a finite extent) and hence applicable to many classes of samples. If the phase can be recovered, spatial resolution is only limited by the maximum momentum transfer in the scattering experiment. These coherent imaging approaches can be carried out using the same contrast mechanisms known from conventional soft x-ray microscopy (elemental, chemical, magnetic etc.)
Direct, real space X-ray imaging (like STXM) with tens of nanometer resolution has been intensively used and perfected over the past few decades. However, one clear limit in achieving even better spatial resolution is set by the manufacturing quality of the X-ray optics (mirrors and lenses). If we also take into account the arrival of FEL sources, which has pushed an interest for diffractive imaging techniques and the development of extensive image reconstruction theory. Simultaneously, the computing power needed for fast, iterative image reconstruction and analysis is now a reality and within the last 15 years, lensless X-ray imaging has been able to blossom as an approach which is not affected by x-ray optics manufacturing limitations. As a result, research is currently carried out in a wide variety of fields. Pushing the spatial resolution below 10 nm and increasing information content (specific contrast; temporal resolution; 3D information; sample environment; multiplexing, etc.) are key areas.
Examples illustrating the width of this field range from the development of holographic imaging in reflection, so that even materials that can only be grown epitaxially on crystalline substrates can be imaged,2 to keyhole diffractive imaging,3,4and ptychographic tomography5, where extended objects can be investigated and the properties of the x-ray beam itself can be analyzed. So, just as light microscopy has proven to be of crucial cross-sectional importance to many areas of science and technology on the micrometer scale, now the development of coherent X-ray imaging is well on its way to offer similar significance in nanoscience.
For an insight into this burgeoning field, a good place to start is the review by Chapman and Nugent, and the references to the different CXI approaches mentioned in there.6
- S. Eisebitt et al., Nature 432, 885 (2004).
- S. Roy et al., Nat. Photonics 5, 243 (2011).
- B. Abbey et al., Nat. Phys. 4, 394 (2008).
- H. M. Quiney et al., Nat. Phys. 2, 101 (2006).
- M. Dierolf et al., Nature 467, 436 (2010).
- H. N. Chapman, and K. A. Nugent, Nat. Photonics 4, 833 (2010).
Scanning transmission x-ray microscopy (STXM)
In its simplest form STXM offers chemically specific information on nm-size areas of a thin sample. The basic technique uses a coherent, monochromatic x-ray beam that is focused through a Fresnel zone plate onto the sample. An aperture which only admits first order focused light is put between the zone plate and sample. Currently, typical focal points (and thereby spatial resolutions) are of the order of 15-50 nm in diameter, which is largely determined by the width of the outer rings of the zone plate.
By monitoring the X-ray signal transmitted through the thin specimen (from 100 nm to a micrometer or so) an image of the sample is obtained as it is raster-scanned. The contrast is hence given by the elements that do/don’t absorb x-ray photons of a specific energy. In this way C can be differentiated from O, but also C-H from C-O as they have different absorption energies – the energy resolution of these type of beamlines is around 0.1 eV.
Alternatively, staying in one spot one can vary the photon energy and record the transmitted intensity to get an x-ray absorption spectrum. Combining these two procedures will then give you a full chemical map over an area of choice.
One of the main advantages of STXM is that the sample can be mounted in air (with sufficiently short X-ray path length), in a He atmosphere, or sandwiched between two X-ray transparent silicon nitride windows. The latter approach is used to study wet samples such as eg. hydrated polymers1 or biological material, but can also be used in conjunction with gas flow2. The focus on the soft x-ray regime makes elements like carbon, nitrogen and oxygen accessible for analysis by exploiting the natural absorption contrast in the water window, rather than having to stain or heavy metal-label parts of the sample, and at a lower radiation dose then in a transmission electron microscope. Using the polarisation of the light magnetic information (circular polarisation) or bond orientation (linear) is accessible, as demonstrated in the picture above3. Most current STXM set ups also provide the possibility for fluorescence measurements (materials with elements showing limited x-ray transmission) and 3D tomography scanning of samples4. Also in the STXM, some of the CXI methods, like ptychography, can be implemented.
For an introduction to STXM, read more on Adam Hitchcock’s website: STXM-intro