The Linac Coherent Light Source (LCLS) is a free-electron (FEL) now in the advanced planning stage at the Stanford Linear Accelerator Center. This machine will employ the process of self-amplified spontaneous emission (SASE) to generate extremely bright pulses of hard x-ray radiation, at wavelengths d own to about 1.5A. Being the first laser to operate in this spectral region, it will produce radiation with properties that are radically different in several respects from that produced by any other x-ray source, including today's 3rd generation synchrotron sources. Using this radiation to best scientific advantage will place unprecedented demands on the x-ray optics. Most importantly, the LCLS will be a pulsed radiation source, with extremely intense, short pulses and a low duty cycle. It will be possible and desirable to collect a complete experimental data set with a single radiation pulse. This will require optics which manipulate a single pulse so as to cover the full range of parameter space to be studied. Also, because of the very high peak power, sub-picosecond pulse length, and very high coherence of the FEL pulse, optics must be robust and able to preserve the short pulse length, and able to preserve spatial coherence. Finally, since it seems likely that the LCLS FEL radiation will be right enough to promote nonlinear x-ray processes, there will be a need for optics which emphasize particular nonlinear effects.

The presented set of parameters for SASE X-FEL beamlines are based on configurations as given in the DESY-TESLA conceptional design report. The present range of photon wavelengths is delivered by several undulator designs and two electron energy ranges from 1-25GeV and 20-30GeV. A wavelength of 1A (12.39keV) is the design goal for TESLA X- FEL undulator beamlines. The possibility to extend the wavelength range to a shorter wavelength of 0.86A (14.4keV), essential for the Mossbauer community, seems feasible. Possible setups of optical elements in typical photon beampaths are shown. Properties and necessary enhancements of these elements and of calculations are discussed.

At the Deutsches Elektronensynchrotron DESY the conceptual design of a linear collider with integrated X-ray free electron lasers (X-FEL) has been presented. The definition of the technical layout of the X-FEL multi user facility has started. Such a facility is different from well-known synchrotron radiation facilities like electron storage rings. An electron beam switch-yard will allow the operation of up to 10 beamlines simultaneously. Dump-systems stop the electron beams after passing the undulators. The experimental hall with about 30 experiments will have a size of about 190 x 34 m². The beamlines for the SASE FELs will have a length between 200 and 400m. The switchyard tunnel system, the buildings of the experimental area and the radiation safety concept are discussed.

The specifications of presently proposed x-ray free electron lasers (FELs) are for machines that will provide x-ray pulses as short as 100 fs with a photon energy as high as 12.3keV. Since the pulse will contain as much as 13 mJ of energy, these devices will present the experimenter with an opportunity to expose matter to an unprecedented x-ray energy density. This high concentration of energetic x-rays presents both a promising frontier in energy-matter interaction, as well as a technological crevasse to be crossed by the experimenter attempting to use the FEL beam. We shall look at three possible problems confronting the experimenter: (1) synchronization of a detector, laser pulse, etc., to the FEL pulse; (2) radiation damage to the target sample; and (3) the presence of an electromagnetic pulse that could damage sensitive electronics located in the experimental area.

Although the realisation of femtosecond X-ray free electron laser (FEL) X-ray pulses is still some time away, X-ray diffraction experiments within the sub-picosecond domain are already being performed using both synchrotron and laser- plasma based X-ray sources. Within this paper we summarise the current status of some of these experiments which, to date, have mainly concentrated on observing non-thermal melt and coherent phonons in laser-irradiated semiconductors. Furthermore, with the advent of FEL sources, X-ray pulse lengths may soon be sufficiently short that the finite response time of monochromators may themselves place fundamental limits on achievable temporal resolution. A brief review of time-dependent X-ray diffraction relevant to such effects is presented.

The emission from plasmas created with fs-lasers provides sub-picosecond x-ray pulses in the keV-range. Intense emission of K(alpha) lines as well as quasi continuum x-rays can be used for time-resolved diffraction and spectroscopy, i.e. to study lattice or atomic dynamics with sub-picosecond resolution by using a laser pump x-ray probe technique. The x-ray yield and x-ray pulse duration of the laser plasma source depend on the laser parameters and the target design, such as intensity, laser wavelength, pulse duration and prepulse level. To accumulate as many photons as possible of the isotropic source an efficient large aperture optic has to be used to select an x-ray line or a wavelength range and focus the radiation onto the sample. It is shown that the use of toroidally bent crystals provides the possibility to refocus 10^-4 of the photons emitted in the whole solid angel to spot size of around 80 micrometers with a temporal broadening below 100 fs. Combinations of bent focusing crystals with a flat sample crystal for fast x-ray diffraction application are discussed. Experiments showing the temporal response of laser heated crystals are presented and compared with theoretical simulations based on Takagi-Taupin theory.

Methods and results of synchrontron radiation wavefront calculations in the framework of scalar diffraction theory are discussed. First, wavefront pecularities of undulator and bending magnet radiation are considered in the single- electron approximation, in comparison with a point source. The consideration includes the effects related to finite electron beam emittance in storage ring sources. Finally, examples of SASE wavefront computation are presented.

Compound refractive lenses (CRL) for hard x-rays are genuine imaging devices like glass lenses for visible light. They are ideally suited for both full field and scanning microscopy with hard x-rays in the range from 2 to 100keV. They are robust and can withstand the heat load of the white beam of an ESRF undulator source. In full field microscopy, resolutions down to 300nm have been achieved so far using aluminium lenses. Resolutions below 100nm are expected for beryllium lenses currently under development. For scanning microbeam techniques, a monochromatic microbeam of 550nm by 5.5micrometers with a 1.1 10¹⁰ph/s (gain 1120) has been achieved with aluminium lenses at a third generation undulator source. For beryllium as lens material, a flux up to two orders of magnitude higher is expected. At planned FEL beamlines, the source size and distance from the source are favorable to microbeams produced by compound refractive lenses, and a diffraction limited microbeam is expected both horizontally and vertically. For beryllium lenses the diffraction limit can be below 100nm. A typical FEL beam size of approximately 1mm at the experiments hutch ideally matches the aperture of compound refractive lenses. Estimates of the heat load on the CRL as well as expected photon fluxes and micro beams sizes are given.

X-ray free-electron lasers (XFELs) designed to operate at approximately 1A wavelengths are currently being proposed by several laboratories as the basis for the next (4th) generation of synchrotron radiation sources. The unique radiation properties of these proposed sources, which include 200 fs pulse duration and peak beam brilliance in excess of 10³³ photon (2 .1%-bw mrad² mm²), offer the possibility of ultrafast time-resolved experiments, perhaps down to 10- fs resolution levels using pulse compression or slicing techniques. Motivated by such potential applications, this paper addresses the relevant instrumentation issue of perfect crystal dynamical diffraction of ultrashort x-ray pulses when the pulse lengths become comparable to the extinction length scales. The basic calculations reported here show the transient time-dependent diffraction from perfect crystals excited by plane-wave delta-function electromagentic impulses. Time responses have been calculated for 8 keV photon energy, for reflected and transmitted beams in both Bragg and Laue cases. Interesting diffraction effects arise, and their implications for XFEL optics are discussed.

With the upcome of shorter and shorter time scales, time dependence of multiple diffraction effects will play a fundamental role in X-ray optics. Here we report on experimental X-ray photon storage in backscattering geometry between two silicon crystal slices cut from a monolithic ingot. The slices are 150 mm apart and wedge shaped to vary the diffracting thickness between 50 micrometers and 500 micrometers . A photon energy of 15.816 keV fulfills the condition for the 888 Bragg reflection. We used the dedicated backscattering beamline ID28 at ESRF which delivers a highly monochromatic beam equal to the natural width of the reflection considered. In Bragg condition, each crystalline boundary of the cavity has a probability of photon transmission and reflection, the ratio depending on the crystal thickness. Once a photon is transmitted by the first slice, it can be reflected by the second crystal and so on. A fast avalanche detector positioned behind the cavity detects the photons as a function of time with respect to the synchrotron bunches. Thus, photons that exit in direct transmission, or after N multiple forth and back bounces are separated by N times one nanosecond. Up to 14 reflections could be observed. The experiment demonstrates not only feasibility of photon storage in crystal cavity which may be relevant in the X-ray optics for a free electron laser but it also points towards the importance of time domain, where pulses shorter than the diffracting volume may be deformed and shaped considerably due to multiple scattering.

The SLAC Linac Coherent Light Source (LCLS), an X-ray Free- Electron Laser (XRFEL) designed to operate over a fundamental energy range of 1 - 8.5 keV, is expected to produce ultra-short pulse lengths down to approximately 200 fs. Even though this represents an enormous decrease with respect to currently available high-brightness X-ray sources, it is believed that for a number of proposed LCLS applications (e.g., imaging or structural studies of molecular clusters with highly focused pulses) it will become necessary to reduce the pulse duration even further, possibly by as much as 1-2 orders of magnitude. Of the various compressive or chopping (i.e., slicing) optical techniques considered for shortening the pulse, the focus of one of our recent studies has been on a recently proposed slicing scheme based on the interaction of a longitudinally chirped LCLS pulse with a specially designed multilayer. The chopping mechanism is the selective reflection of only that sub-interval of the pulse that fulfills the multilayer Bragg condition. Of particular interest are the reflection efficiency and the distortion induced in the temporal fine structure of the reflected pulse, both of which can be critical to the efficacy of the scheme for a given application. Here we present results of selected parameter studies of different multilayer-based beam chopper systems using codes and LCLS radiation models developed at SSRL and LLNL.

A method for X-ray interferometric investigation of the deformation field of crystal imperfections by means of double and triple interferometers is proposed. It is shown experimentally that using double and triple interferometers one can detect segregation lines, displacement lines, as well as the Moire patterns of imperfections of different types.

An International Workshop on Metrology for X-ray and Neutron Optics has been held March 16-17,2000, at the Advanced Photon Source, Argonne National Laboratory near Chicago, Illinois (USA). The workshop gathered engineers and scientist from both the U.S. and around the world to evaluate the metrology instrumentation and methods used to characterize surface figure and finish for long grazing incidence optics used in beamlines at synchrotron radiation sources. This two-day workshop was motivated by the rapid evolution in the performance of x-ray and neutron sources along with requirements in optics figure and finish. More specifically, the performance of future light sources, such as free-electron laser (FEL)-based x-ray sources, is being pushed to new limits in term of both brilliance and coherence. As a consequence, tolerances on surface figure and finish of the next generation of optics are expected to become tighter. The timing of the workshop provided an excellent opportunity to study the problem, evaluate the state of the art in metrology instrumentation, and stimulate innovation on future metrology instruments and techniques to be used to characterize these optics.

Refractive lenses have been used successfully to focus incoherent x-ray emission in the wavelength range from 2 to .5A with focal lengths on the order of one meter. A stack of N lens elements is employed to reduce the focal length by the factor N over a single element, and such a lens is terms a Compound Refractive Lens (CRL). Contrary to intuition, misalignment of parabolic lens elements doesn't alter the focusing properties and results in only a small reduction in transmission. Coherent x-ray sources are being developed with wavelengths of 1-1.5A and source diameters of 50- 80micrometers , and the CRL is ideally suited to produce a small, intense image. Chromatic aberration increase the size of the image and so it is important to provide chromatic correction to minimize the image dimensions. Pulse broadening due to the dispersion of the lens material is negligible. Intensity gain is in the range from 10⁵ to 10+$6), where gain is defined as the intensity ratio in an image plane with and without the lens in place. Maximum image intensity is obtained when the CRL is placed a distance of 100 to 200 m from the source, and the typical diameter of the focused spot is about one micron.