Laterally graded W/B<sub>4</sub>C multilayers were conceived for the focusing of hard X rays at 3<sup>rd</sup> generation synchrotron sources. They were deposited using a differential sputter coating technique. The multilayer mirror was bent to the correct shape on a dynamical bending device applying automated alignment routines. During experiments on the ESRF beamline ID19 the undulator source was focused vertically to a 41 nm (FWHM) wide line using a photon energy of 24 keV. The measured line width can be attributed to the finite source size, to diffraction effects, and to slope errors of the mirror. The potential impact of beam penetration into the multilayer will be discussed.
This paper presents the first series of round-robin metrology measurements of x-ray mirrors organized at the Advanced Photon Source (APS) in the USA, the European Synchrotron Radiation Facility in France, and the Super Photon Ring (SPring-8) (in a collaboration with Osaka University,) in Japan. This work is part of the three institutions' three-way agreement to promote a direct exchange of research information and experience amongst their specialists. The purpose of the metrology round robin is to compare the performance and limitations of the instrumentation used at the optical metrology laboratories of these facilities and to set the basis for establishing guidelines and procedures to accurately perform the measurements. The optics used in the measurements were selected to reflect typical, as well as state of the art, in mirror fabrication. The first series of the round robin measurements focuses on flat and cylindrical mirrors with varying sizes and quality. Three mirrors (two flats and one cylinder) were successively measured using long trace profilers. Although the three facilities' LTPs are of different design, the measurements were found to be in excellent agreement. The maximum discrepancy of the rms slope error values is 0.1 μrad, that of the rms shape error was 3 nm, and they all relate to the measurement of the cylindrical mirror. The next round-robin measurements will deal with elliptical and spherical optics.
The ESRF optics metrology laboratory was created 15 years ago. Various measurement devices have been progressively installed and the present status of available equipment will be briefly presented. Since the beginning of the first beamline construction, all X-ray mirrors have been tested before their installation. Most of the mirrors are mounted on mechanical bending systems, and it is mandatory to characterize optical elements under working conditions and to calibrate the systems before their installation on a beamline. These calibrations are now part of the acceptance tests whenever a system is delivered. Optics tests carried out on the Long Trace Profiler (LTP) will be described, with particular emphasis on the special configuration developed for mirrors facing down. Measurement reproducibility and accuracy achieved with the LTP will be discussed.
The emerging micro focusing needs at ESRF have promoted the development of Kirkpatrick-Baez systems. Precise metrology plays an important role to control the mirror clamping using interferometry techniques and to predict the performance of the system using the LTP. The automatic shaping procedure will be described.
One of the factors influencing the focus size in diffractive-refractive optics is the quality of diffracting surface. If the surface is uneven, then the diffraction at each point of the surface is a combination of an asymmetric and inclined diffraction (general asymmetric diffraction). This somewhat deviates and spreads the diffracted beam. The integration over the surface hit by an incident beam gives the angular spread of the diffracted beam. It is shown that in some cases (highly asymmetric, highly inclined cut) the etched surface may create the spread of the diffracted beam, such that it causes a significant broadening of the focus. In this case a mechanical-chemical polishing is necessary.
This paper describes the mechanical design of Kirkpatrick-Baez (KB) mirror systems that have been developed at the ESRF over several years. These very compact and stable though flexible focusing devices for synchrotron x-ray beams are based on bending an initially flat, superpolished plate, which permits to vary the focusing conditions. Nowadays a whole family of mechanical benders exists at the ESRF that allows us to choose the most adapted system according to the properties defined by the experiment such as the energy and energy range, the focusing parameters such as magnification and focusing distance and the substrate coatings, <i>i.e. </i>single-layer mirrors or multilayers. The geometrical characteristics of these KB systems can be chosen in terms of focal distances ranging from 0.1 to 3 m and circular or elliptical bending radii from 20 to 1000m. Mirror substrates such as silicon or pyrex, single-layer or multilayer coatings require different motorisations and deformation systems. The very challenging requirements for mechanical resolution and sensitivity have led to the development of several generations of micro-motors. The ESRF has built a special multi-purpose micropusher that provides the required resolution and linearity, a thrust up to 80 N and finally a good position latching. Issues such as mounting interfaces, stress in the bent mirror and the dynamic bender, local mirror deformation and curvature stability had to be addressed and were solved. The ESRF has developed mirror clamping technologies controlled by mechanical and optical metrologies. The dynamic stability and reproducibility requirements to achieve a spot size variable from sub-micron to tens of microns required by various beamlines necessitate a very high degree of stiffness.
ID09 is a dual-purpose beamline dedicated to time-resolved and high-pressure experiments. The time-resolved experiments use a high-speed chopper to isolate single pulses of x-rays. The chopper is installed near the sample (focal spot) and the shortest opening time depends on the height of the tunnel in the chopper, i.e. the sharpness of the vertical focus. In the 16-bunch mode, the opening window of the chopper has to be smaller than 0.352 μs in order to isolate single pulses of x-rays. This requires reducing the height of the tunnel to 0.143 mm. To ensure a reasonable transmission though the tunnel, we have designed a very precise toroidal mirror that focuses the beam 22.4 m downstream with a magnification M = 0.677. The 1.0 m long silicon mirror is curved by gravity into a nearly perfect toroid with a meridional radius of 9.9 km. The curvature is fine-tuned by a stepper motor that pushes via a spring from below the mirror. The overall figure error from the gravity sag and the corrective force is less than 0.3 μrad. The polishing error is 0.7 μrad (rms) averaged over the central 450 mm of the 1000 mm long mirror. The measured size of the polychromatic focus is 0.100 mm × 0.070 mm (h x v) in agreement with the prediction from the ESRF long trace profiler data. The small focal spot, which integrates the full central cone of the U17 undulator, is the result of very high optical quality, curvature fine-tuning, strain-free mount, vibration free cooling and careful alignment.
The mirror bender installed at the SLS protein crystallography beamline is designed to be capable of adjusting the vertical phase space of the undulator to the acceptance of protein crystals, i.e. to produce micro-beams as well as essentially parallel beams. The two-moment bender is based on the flexural hinge design pioneered at the ESRF but adapted to high-vacuum by making use of in-vacuum motors and high resolution worm-gears. Special care was taken in the design and fabrication of the clamps and the application of the clamping torque. The Rh-coated fused silica substrate (Zeiss) has a free length of 400 mm and a thickness of 30 mm. Metrology tests at ESRF indicate the high quality of the mirror and the bender. Over the useful length of 350 mm the meridional slope error was found to be 0.17-0.3 μrad (rms) prior to, and less than 0.5 μrad after clamping to the bender. In practice this allows the full central cone of an in-vacuum undulator to be focussed to 7 μm at an image distance of 7.1 m and to 2.1 μm at 1.75 m, corresponding to effective slope errors of less than 0.25 μrad. The bending is very reproducible and is well described by an interaction matrix. Finally, the long range tails were measured in the context of the generation of fs-pulses by means of bunch slicing. Their level cannot be attributed to the surface roughness of 2.9 Å (rms) but rather to scattering from other beamline components such as Be-windows.
We describe the fabrication and testing of a novel type of tunable transmission hard x-ray optics. The diffractive elements are generated by electron beam lithography and chemical wet etching of <110> oriented silicon substrates. Structures with widths down to 100 nm and extreme aspect ratios were obtained using this method. By tilting the lenses with respect to the x-ray beam, the effective path through the phase shifting structures can be varied. This makes it possible to optimize the diffraction efficiency for a wide range of photon energies, and to obtain effective aspect ratios not accessible with untilted optics. The diffraction efficiency of a Fresnel lens was measured for various energies between 8 keV and 29 keV. Values close to the theoretical limit (approx. 35%) were obtained. The described technique provides focusing in one direction only. For two-dimensional focusing, two linear lenses with different focal lengths and orthogonal orientations can be placed along the optical axis. Depending on the coherence properties of the source, such an arrangement can improve the resolution and flux compared to a single circular zone plate. The wet etching technique is also applied to the fabrication of linear gratings with pseudo-random pitch, which will be used as one-dimensional decoherers to adapt the coherence of a synchrotron beam in a defined way. Linear gratings with uniform line density can be used as beam splitters for applications such as holography or interferometry.
We describe Kirkpatrick-Baez (KB) reflecting mirror systems that have been developed at the European Synchrotron Radiation Facility (ESRF). They are intended to be used mainly in the hard x-ray domain from 10 KeV to 30 KeV for microfluorescence, microdiffraction and projection microscopy applications. At 19 KeV a full width at half maximum (FWHM) spot size of 200x600 nanometers has been measured and with an estimated irradiance gain of 3.5x10<SUP>5</SUP>. The alignment and bending processes of the system are automated based on the wavefront information obtained by sequentially scanning slits and reading a position-sensitive device located in the focal plane. The sub-microradian sensitivity of this method allows us to predict the spot size and ot provide a metrology map of the surfaces for future improvements of the performances. A novel device based on specular reflection by a micromachined platinum mirror has been used to determine the spot size with an equivalent slit size of less than 100 nanometers. Projection phase images of submicron structures are presented which clearly show both the high potential and also the present limitations of the system. First microfluorescence images obtained at 20.6 KeV are shown. Finally, a roadmap towards diffraction-limited performance with metal and multilayer surfaces is presented.
The properties of optical components are described that are required by synchrotron x-ray scattering techniques needing to fully exploit the coherence properties of third- generation sources. For example, phase imaging and microfocusing applications require reflecting surfaces with errors much smaller than a nanometer up to spatial frequencies in the mm<SUP>-1</SUP> range while keeping roughness on the sub-Angstrom level. The two major challenges being addressed in this paper are the accuracy of the metrology and the technology of the machining based on the metrology for deterministic corrections. The low emittance and the stability of the synchrotron source combined with x-ray position sensitive detectors are suitable to measure the reference wavefront with picometer accuracy. The deformation of this wavefront by a reflecting mirror leads to surface error maps whose accuracy corresponds to the conditions for coherence preservation. These maps are used as input for an ion-beam milling machine correcting these height errors on grazing incidence mirrors. First results are presented and limits of these techniques including possibilities of multilayer corrections are discussed.
Hard x-ray focusing devices based on laterally graded multilayers are key elements to fully exploit the advantages of third generation synchrotron sources. We have developed a design method to produce laterally graded multilayers using sputter deposition techniques. The multilayers are adapted to the given application by the proper choice of layer materials, d spacing, and the partition of the two constituent materials. The optimization of all relevant parameters yields an ab initio estimation of the desired layer thickness gradient. The performance and the accuracy of this method are demonstrated. The experimental lateral thickness errors could be reduced below 0.5% RMS over a total length of 300 mm. Reflectivity measurements at different energies are in good agreement with theoretical simulations. During focusing experiments at 13 keV a spot size of 1 micrometer and a gain in flux of 1000 were achieved.
In this paper we describe an x-ray long trace profiler that takes an x-ray synchrotron beam as a wavefront reference. According to results of experiments conducted on the Optics Beamline at the ESRF, this instrument allows us to measure surface slope errors with precision and accuracy better than 25 nrad (rms) and 50 nrad (rms), respectively, with a lateral resolution of 5 mm in the meridional and less than 1 mm in the sagittal direction. A very similar technique was developed to figure in situ mirrors mounted on mechanical benders into a stigmatic shape for microfocusing purposes. Micron spot sizes were achieved without difficulty and submicron precision should be possible. The technique is particularly useful if energy tunability is needed. The emphasis has been put on automation and speed of the measurement.
The very low emittance of new synchrotron sources and the increasing number of micro-focusing applications make the production of highly stable and well defined beams increasingly necessary. The use of flexible mirrors whose curvature can be changed while maintaining a correct figure appeared to be a very attractive solution. For over two years, the ESRF has been developing a new approach which consists of making an X-ray mirror from an active material such as piezoelectric ceramics. With respect to conventional bender this concept, already used in astronomical and laser applications, has the advantages to be mechanics free, very compact and relatively cost effective. This paper presents the status of the ESRF's developments in this field. First, theoretical and technical descriptions of the system are given. Experimental tests of various configurations confirmed the potential of this concept. For example, two 150 mm long bimorph mirrors set into a Kirkpatrick-Baez geometry gave a focused spot of 10 micrometers (vertical) X 20 micrometers (horizontal). Finally, the developments of in-situ control systems (strain gauges, optical devices), necessary to fully exploit the capabilities of these active optics, are discussed.
We present the x-ray performance on an ESRF synchrotron beamline of a focusing device based on the dynamical bending of a flat silicon plate coated with a 2.5 m d-spacing W/Si multilayer. The mirror was shaped by trial and error to a cylindrical ellipse using an optical profilometer. In a first experiment the device was bent to a 71-m radius to account for the demagnification factor and the energy of operation. With a monochromatic incident beam set at 9 keV a vertical spot size of 4.5 micrometers was obtained at 41 m from the source and 1 m from the multilayer, which agreed closely with theoretical expectations. Such good agreement was due to small residual slope error with respect to the ellipse: around 2 (mu) rad over a 150-mm length for radii greater than 50 m. Moreover, as the divergence of the incident beam was larger than the rocking curve width, less than 25% of the mirror could contribute, reducing the distortion to an even lower figure. With the same geometrical parameters the device exposed to the white beam (monochromator removed) lead to a vertical spot size of about 7 micrometers . Here the whole mirror surface cold reflect, which caused more distortion to the incoming beam but also gave rise to a much higher intensity. The gain in flux obtained with a gradient of d-spacing along the mirror surface is discussed. Finally, results with a Kirpatrick-Baez arrangement let expect in the near future a flux gain greater than 10<SUP>4</SUP> with a 10 micrometers by 10 micrometers focal spot.
Beamline optical components and their dynamic holding and bending systems are qualified at the ESRF in the optical metrology laboratory, which has been moved recently into the ESRF experimental hall. Software and hardware enhancements of the five instruments used to perform these ex situ characterizations are described. In situ beamline mirror metrology using a wavefront analyzer has been introduced in ID24 and ID20 beamlines. The design of the analyzer, its architecture with respect to the mirror chambers and its use in active optics control with mechanical and bimorph benders are described in this paper.
X-ray beams delivered by third generation synchrotron radiation machines have a high degree of coherence that allows to use different coherent imaging techniques such as phase contrast microscopy, holography, and phase-contrast tomography in the high-energy domain. The question arises to what extent the existing optical elements such as mirrors and multilayers are capable of preserving this high spatial coherence. A theory of partially coherent x-ray scattering by a slightly modulated mirror surface under total reflection is developed. This gives estimations for intensity contrast as a function of the surface modulation. X-ray topography with coherent light is proposed and applied for mirror diagnostics. X-ray interferometry in an in-line holography setup is used to describe the coherence properties of the x-ray beam including optical components like mirror and multilayers. It is shown that significant improvements of the polishing process are still needed for the long mirrors while small state-of-the-art mirrors (less than 100 nm) are nearly adequate.
In a previous paper, we described the experimental set-up and requirements used to study an XMM mandrel by x-ray angle-resolved scattering (ARS). We presented first results and compared them to micro-profilometry data. Here we complete the description of the experimental method and the data analysis of the x-ray ARS studies. We point out several pitfalls and propose solutions to avoid them. We emphasize the need to span a wide intensity dynamical range and the importance to separate the intensities form the specular and the scattered beams. This separation is of particular interest for estimating the rms-roughness from the power spectral density, modeled by a power-law of the spatial frequency. We then compare the results for the roughness with those obtained from profiler measurements. In a second part, the figure measurements of the XMM mandrel are described and analyzed in detail. They have been carried out with both an x-ray pencil beam and an optical long trace profiler. In particular, much attention has been given to the determination of the angle between the two sections of the Wolter I optical configuration and to the effect of the mandrel mounting supports. The PSD was completed with the low-frequency results. Finally, the surface data from the earlier experiment on the Ni-coated normally-polished paraboloid and the superpolished hyperboloid were used to predict the image quality of a Wolter I type optics having the same surface characteristics. The influence of different surface finishes on the image point spread function of a grazing incidence mirror is discussed.
To adress the problems encountered when measuring subinicrons CD patterns with white light microscopy, an instrument
using the ultraviolet spectrum around 280 urn has been designed and tested.
The technologies developped for dealing simultaneously with ultraviolet, visible and infrared light are pnsented.
The main advantage lies in the absorbance ofphotonsists, leading to simple optical proffles.
Theontical results obtained with a modelling approach are iepoited along with experimental results characterized with
respect to SEM cross sections.