structure and dynamics of magnetically ordered systems. Inelastic x-ray scattering, the equivalent of Raman scattering in the visible region, would become feasible with a significant increase in x-ray flux. High-energy phonons and magnons, which are difficult to probe with neutron scattering, could be resonantly enhanced by exploiting the laser's tunability. The behavior of quasicrystalline and fluid-phase short-range order, and of quenched disorder in glassy and related materials, could be revealed. While some of these structural studies could be done with third-generation synchrotron sources, an x-ray laser would be required to probe the dynamical behavior of these systems. Materials in which only minute samples are available would also be amenable to study with a brighter x-ray source.

For many years, high-resolution elemental mapping—by imaging above and below K-edges of atomic transitions—and three-dimensional pictures have been actively pursued. A number of experiments have shown that these goals are difficult to achieve, and problems encountered with x-ray imaging must be considered in the context of the advances in scanning microscopies such as the tunneling, force, and near-field optical microscopes.

X-ray microscopy done in either an imaging or scanning mode has achieved a transverse resolution on the order of 300 Å. Since the numerical aperture of x-ray optics is ≤ 0.1 for the best Fresnel zone plates, the depth resolution is at least 20 times worse. X-ray shadow microscopy in which a sample is placed on top of a high-resolution photoresist such as polymethyl methacrylate has achieved around 100-Å resolution, limited primarily by the damage range of the x-rays as they penetrate into the resist. Realistically, soft x-ray microscopy in the water window could achieve 200-Å transverse resolution in the near future, but achieving dramatically higher spatial resolution will require a breakthrough.

Given the lack of high-numerical-aperture x-ray optics, three-dimensional imaging can be done either by tomography or holography. Tomography uses a set of two-dimensional projections to reconstruct a three-dimensional image. The quality of the image is heavily dependent on reconstruction algorithms. Holography uses the interference between light scattered from a sample and light from a reference beam. A true three-dimensional image can be achieved only if holograms are taken from a number of views.

A highly coherent x-ray source is a necessary condition for three-dimensional high-resolution imaging. However, it is not a sufficient condition. Atomic resolution demands a high-resolution recording medium and reading capability. The best holographic images, recorded in a high-resolution resist and read out with a scanning tunneling microscope, have achieved a resolution of 560 Å. The resolution in this work was determined by the signal-to-noise ratio of the recording medium, which limited the number of high-frequency oscillations that could be detected.