This PDF file contains the front matter associated with SPIE Proceedings Volume 9965, including the Title Page, Copyright information, Table of Contents, and Conference Committee listing.

A two-stage adaptive optical system using four piezoelectric deformable mirrors was constructed at SPring-8 to form collimated X-ray beams. The deformable mirrors were finely deformed to target shapes (elliptical for the upstream mirrors and parabolic for the downstream mirrors) based on shape data measured with the X-ray pencil beam scanning method. Ultraprecise control of the mirror shapes enables us to obtain various collimated beams with different beam sizes of 314 μm (358 μm) and 127 μm (65 μm) in the horizontal (vertical) directions, respectively, with parallelism accuracy of ~ 1 μrad rms.

We present an X-ray mirror bender that includes multiple spring actuators that introduce a controlled deformation of the mirror substrate capable of correcting residual figure errors on the mirror, below one nanometer. For usual mirror dimensions, this requires applying correcting forces with resolution and stability in the order of 0.01 N, and a range up to 20 N, depending on the initial figure error of the mirror. To obtain the required stability, the actuators need to compensate intrinsic mechanical instabilities, such as thermal drifts or the limited repeatability of parts that move during the adjustment of the figure. The concept we propose uses weak springs that allow reducing all these effects below noticeable values. Additional considerations on friction and parasitic components of the force are accounted. The system also includes two independent bending actuators with a larger force range to generate the mean elliptic figure of the mirror. Metrology tests of the performances of the system show that the correctors are repeatable within 0.01 N, and reach much higher resolution. A prototype of the bender has been used to correct the figure error of a 500 mm long mirror below one nanometer (root mean square). The agreement to the predicted figure is better than 0.08 nm rms.

The European XFEL will generate extremely short and intense X-ray laser pulses of high coherence and nearly diffraction-limited divergence. Guiding these X-rays beams over a distance of more than 1 km to the experiments requires an extreme precision in pointing stability of beamline components like mirrors and gratings and also a control of the divergence of the beam. The specifications of the X-ray mirrors that will be able to transport, distribute and focus the beam are quite challenging. The European XFEL mirrors for the beam transport are 950 mm long and the optical surface specifications are 2 nm Peak-To-Valley. Some of the mirrors will have bending capabilities in order to focus the beam in the right position and with nanometer accuracy. This is implemented using a mechanical bender that will ensure stability of the optics in the nanometer range and will also offer the possibility to correct for mechanical or temperature drifts.

We present here the characterization of a mechanical bender that was done using two instruments, a Large Aperture Fizeau interferometer and a system of three capacitive sensors. The bender is designed in a way that the mirror is hold with clamps on both ends and a symmetric torque is applied on the clamps, inducing a cylindrical shape on the mirror surface. Several long-term stability measurements were done, as well as the characterization of bending capabilities. The parameters retrieved from the measurements are the sagitta and therefore the radius of curvature for different bending positions. The behavior of the variation of the shape of the mirror was also studied. The information gathered from our measurements will be used to optimize the final design of the bender.

The success of the LCLS led to an interest across a number of disciplines in the scientific community including physics, chemistry, biology, and material science. Fueled by this success, SLAC National Accelerator Laboratory is developing a new high repetition rate free electron laser, LCLS-II, a superconducting linear accelerator capable of a repetition rate up to 1 MHz. Undulators will be optimized for 200 to 1300 eV soft X-rays, and for 1000 to 5000 eV hard X-rays. To absorb spontaneous radiation, higher harmonic energies and deflect the x-ray beam to various end stations, the transport and diagnostics system includes grazing incidence plane mirrors on both the soft and Hard X-ray beamline. To deliver the FEL beam with minimal power loss and wavefront distortion, we need mirrors of height errors below 1nm rms in operational conditions. We need to mitigate the thermal load effects due to the high repetition rate. The absorbed thermal profile is highly dependent on the beam divergence, and this is a function of the photon energy. To address this complexity, we developed a mirror cradle with variable length cooling and first order curve correction. Mirror figure error is minimized using variable length water-cooling through a gallium-indium eutectic bath. Curve correction is achieved with an off-axis bender that will be described in details. We present the design features, mechanical analysis and results from optical and mechanical tests of a prototype assembly, with particular regards to the figure sensitivity to bender corrections.

LCLS-2 is a high repetition rate (up to 1 MHz) superconducting FEL and the soft x-ray branch will operate from 0.2 to 1.3 keV. Over this energy range, there is a large variation in beam divergence and therefore, a large variation in the beam footprint on the optics. This poses a significant problem as it creates thermal gradients across the tangential axis of the mirror, which, in turn, creates non-cylindrical deformations that cannot be corrected using a single actuator mechanical bender. To minimize power loss and preserve the wave front, the optics requires sub-nanometer RMS height errors and sub-microradian slope errors. One of the key components of the beam transport in the SXR beamline is the bendable focusing mirror system, operated in a Kirkpatrick-Baez Configuration. For the first time in the Synchrotron or FEL world, the large bending needed to focus the beam will be coupled with a cooling system on the same mirror assembly, since the majority of the FEL power is delivered through every optic leading up to the sample. To test such a concept, we have developed a mirror bender system to be used as a multipurpose optic. The system has been very accurately modeled in FEA. This, along with very good repeatability of the bending mechanism, makes it ideal for use as a metrology tool for calibrating instruments as well as to test the novel cooling/bending concept. The bender design and the tests carried out on it will be presented.

In order to advance significantly scientific objectives, future x-ray astronomy missions will likely call for x-ray telescopes with large aperture areas (≈ 3 m²) and fine angular resolution (≈ 1²). Achieving such performance is programmatically and technologically challenging due to the mass and envelope constraints of space-borne telescopes and to the need for densely nested grazing-incidence optics. Such an x-ray telescope will require precision fabrication, alignment, mounting, and assembly of large areas (≈ 600 m2) of lightweight (≈ 2 kg/m² areal density) high-quality mirrors, at an acceptable cost (≈ 1 M$/m² of mirror surface area). This paper reviews relevant programmatic and technological issues, as well as possible approaches for addressing these issues-including direct fabrication of monocrystalline silicon mirrors, active (in-space adjustable) figure correction of replicated mirrors, static post-fabrication correction using ion implantation, differential erosion or deposition, and coating-stress manipulation of thin substrates.

A technique to obtain lightweight and high-resolution focusing mirror segments for large aperture X-ray telescopes is the hot slumping of thin glass foils. In this approach, already successfully experimented to manufacture the optics of the NuSTAR X-ray telescope, thin glasses are formed at high temperature onto a precisely figured mould. The formed glass foils are subsequently stacked onto a stiff backplane with a common axis and focus to form an XOU (X-ray Optical Unit), to be later integrated in the telescope optic structure. In this process, the low thickness of the glass foils guarantees a low specific mass and a very low obstruction of the effective area. However, thin glasses are subject to deformations that may arise at any stage of the production process, thereby degrading the angular resolution. To solve this problem, several groups are working on the possibility to correct the mirror profile post-manufacturing, using piezoelectric elements exerting a tangential strain on the non-optical side of the glass mirrors. In this paper we show the results of the approach we have adopted, based on the application of piezoceramic patches on the backside of thin glass foils, previously formed by hot slumping. The voltage signals are supplied to the piezoelectric elements by a system of electrodes deposited on the same side of the mirror via a photolithographic process. Finally, the matrix of voltages to be used to correct the mirror shape can be determined in X-rays illumination by detection of the intra-focal image and consequent reconstruction of the longitudinal profile. We describe the production of some active mirrors with different arrangements of piezoelectric elements and the X-ray tests performed at the XACT X-ray facility to determine the optimal actuator geometry.

Use of thin glass modular optics is a technology currently under study to build light, low cost, large area X-ray telescopes for high energy astrophysics space missions. The angular resolution of such telescopes is limited by local deviations from the ideal shape of the mirrors. One possible strategy to improve it consists in actively correcting the mirror profile by gluing thin ceramic piezo-electric actuators on the back of the glasses. A large number of actuators, however, requires several electrical connections to drive them with the different needed voltages. We have developed a process for depositing conductive paths directly on the back of non-planar thin foil mirrors by means of a photolithographic process, combined with metal thin film evaporation and selective removal. We have also designed and built a modular multichannel electronic driver with each module capable of driving simultaneously up to 16 actuators with a very low power consumption. Here we present our electrical interconnections technology and the solutions adopted in the implementation of the electronics.

Thin glass modular mirrors are a viable solution to build future X-ray telescopes with high angular resolution and large collecting area. In our laboratories, we shape thin glass foils by hot slumping and we apply pressure to assist the replication of a cylindrical mould figure; this technology is coupled with an integration process able to damp low frequency errors and produces optics in the Wolter I configuration, typical for the X-ray telescopes. From the point of view of the hot slumping process, the efforts were focused in reducing low-, mid- and high- frequency errors of the formed Eagle glass foils. Some of our slumped glass foils were used for the development of active X-ray optics, where piezoelectric actuators are used to correct the slumped glass foil deviations from the ideal shape. In particular, they were used for the Adjustable X-raY optics for astrOnoMy project (AXYOM) developed in Italy, and the X-ray Surveyor mission, as developed at the Smithsonian Astrophysical Observatory / Center for Astrophysics (SAO/CfA) in USA. In this paper we describe the optimisation of the hot slumping process, comparing the results with the requirements of the considered active optics projects. Finally, since the present configuration of the Pennsylvania State University (PSU) coating equipment is limited to 100 x 100 mm², the slumped glass foils used for the SAO project were cut from 200 x 200 mm² to 100 x 100 mm², and a low-frequency change was observed. A characterisation of the profile change upon cutting is presented.

In this paper we review the progress and current status of thermal forming activities at SAO, highlighting the most relevant technical problems and the way to solve them. These activities are devoted to the realization of mirror substrates for the X-ray surveyor mission concept, an observatory with Chandra-like angular resolution and 30 times more effective area or larger. The technology under development at SAO is based on the deposition of piezoelectric material on the back of the substrates. About 8000 mirror segments, with initial quality of 10 arcseconds or better are required for the telescope.

Here we report progress toward the fabrication of adaptive or active Si X-ray mirrors via a two step process. The first step is to curve a Si flat and then coat it with Terfenol-D that will allow the shape control via the application of a magnetic field. The goal is to create a mirror whose local (a few mm-length scale) slope can be changed and left for several hours or more. The current work described here was done in on Si to demonstrate the ability to produce the initial curvature, and in parallel, work to on magnetically hard NiCo 5 cm x 5 cm square plus on a glass sample. The glass sample was used a proto-type to model making changes in two different locations on a mirror. The NiCo sample was used to show that a magnetic field can be retained in a magnetically hard substrate such that the magnetically induced stress in the Terfenol-D was able to maintain a deformation for as long as time permitted to make the measurement which was 71 hours.

Deformable, piezo bimorph mirrors are often used to expand X-ray beams to a continuous range of sizes. However, optical polishing errors present on all X-ray mirrors introduce striations into the reflected beam. To counteract them, reentrant surface modifications with alternating concave and convex curvature have been proposed and applied to mirrors of fixed shape or bimorph mirrors. For the latter, a new method of constructing re-entrant surface modifications on segments of unequal length is described. This allows the re-entrant modification required for a desired beam size at the focal point to be matched to the bimorph mirror’s polishing errors, thus reducing the voltage variations. Optical profilometry using the Diamond-NOM showed that a 5-segment and a 7-segment modification could be suitably applied to a deformable bimorph mirror. X-ray tests showed that striations caused by the 5-segment modification in the beam at the focus are concentrated at the beam edges, while the beam center is left clear. This is in contrast to simple defocusing, in which a strong side shoulder appears. The 7-segment modification produces a pattern of evenly spaced striations. The intensity spikes seen with the re-entrant modifications are caused chiefly by the finite curvature of the mirror at the turning points. The question of whether deformable bimorph mirrors with different piezo response functions could sharpen the curvature changes will be investigated. Optimal modifications of continuous curvature, which could more realistically be applied, will be sought.

FERMI is the first seeded EUV-SXR free electron laser (FEL) user facility, and it is operated at Elettra Sincrotrone Trieste. Two of the four already operating beamlines, namely LDM (Low Density Matter) and DiProI (Diffraction and Projection Imaging), use a Kirkpatrick-Baez (K-B) active X-ray optics system for focusing the FEL pulses onto the target under investigation. A wafefront sensor is used as diagnostic for the characterization of the focused spot and for the optimization of the parameters of these active optical systems as well. The aim of this work is, first, to describe in detail the optimization procedure using the wavefront sensor through the minimization of the Zernike coefficients, and second, report on the final results obtained on the K-B optical system at the DiProI endstation. The wavefront sensor, mounted out of focus behind the DiProI chamber, allows to compute the intensity distribution of the FEL beam, typically a mix between several modes resulting in a ”noisy hyper-Gaussian” intensity profile, and the wavefront residual from ideal propagation shape and after tilt correction. Combining these two measures we can obtain the electric field of the wave out of focus. Propagating back the electric field we reconstruct the focal spot in far field approximation. In this way the sensor works as a diagnostic reconstructing the focal spot. On the other hand, after modelling the electric field with a Zernike polynomial it is easy and fast to optimize the mirror bending and the optical system angles by minimizing the aberrations, quantified in terms of Zernike coefficients. Since each coefficient corresponds to a single parameter, they can be minimized one at the time. Online wavefront measurements have made possible the optimization of the bending acting on the mirror curvature, and of the (pitch and roll) angle positions of the K-B system. From the wavefront measurements we have inferred a focal spot for DiProI of 5.5 μm x 6.2 μm at 32 nm wavelength, confirmed by the PMMA ablation imprints. The experimental results were compared with the predictions from simulations obtained using the WISE code, starting from the characterization of the actual mirror surface metrology. The results from simulations were found to be in agreement with the experimental measurements.

Our experiments at beamline 5.3.1 of the Advanced Light Source feature a 45-cm long x-ray deformable mirror (XDM). We describe the experiment and present recent results in two areas. First, we directly image the 3 keV x-ray beam and demonstrate customized shaping of its intensity in the near field. Detailed physics simulations of the experiment agree very well with actual measurements. Second, we use a grating interferometer to measure known figure errors applied to the surface of the XDM. A relative height change on the XDM of 2.5 nm RMS is measured at an SNR of eight in single measurement. A provisional error budget analysis indicates that uncalibrated errors in the system are by far the largest component.

After years of development from a concept to early experimental stage, X-ray Deformable Mirrors (XDMs) are used in many synchrotron/free-electron laser facilities as a standard x-ray optics tool. XDM is becoming an integral part of the present and future large x-ray and EUV projects and will be essential in exploiting the full potential of the new sources currently under construction. The main objective of using XDMs is to correct wavefront errors or to enable variable focus beam sizes at the sample. Due to the coupling among the N actuators of a DM, it is usually necessary to perform a calibration or training process to drive the DM into the target shape. Commonly, in order to optimize the actuators settings to minimize slope/height errors, an initial measurement need to be collected, with all actuators set to 0, and then either N or 2N measurements are necessary learn each actuator behavior sequentially. In total, it means that N+1 or 2N+1 scans are required to perform this learning process. When the actuators number N is important and the actuator response or the necessary metrology is slow then this learning process can be time consuming. In this work, we present a fast and accurate method to drive an x-ray active bimorph mirror to a target shape with only 3 or 4 measurements. Instead of sequentially measuring and calculating the influence functions of all actuators and then predicting the voltages needed for any desired shape, the metrology data are directly used to “guide” the mirror from its current status towards the particular target slope/height via iterative compensations. The feedback for the iteration process is the discrepancy in curvature calculated by using B-spline fitting of the measured height/slope data. In this paper, the feasibility of this simple and effective approach is demonstrated with experiments.

MSFC has a long history of developing full-shell grazing-incidence x-ray optics for both narrow (pointed) and wide field (surveying) applications. The concept presented in this paper shows the potential to use active optics to switch between narrow and wide-field geometries, while maintaining large effective area and high angular resolution. In addition, active optics has the potential to reduce errors due to mounting and manufacturing lightweight optics. The design presented corrects low spatial frequency error and has significantly fewer actuators than other concepts presented thus far in the field of active x-ray optics. Using a finite element model, influence functions are calculated using active components on a full-shell grazing-incidence optic. Next, the ability of the active optic to effect a change of optical prescription and to correct for errors due to manufacturing and mounting is modeled.

We report recent results of the performance measurement of our X-ray telescope with adaptive optics. The telescope is designed to use the 13.5nm EUV with the Mo/Si multilayers, making a normal incident optics. The primary mirror is 80mm in its diameter and the focal length of 2m. The deformable mirror is controlled by measuring a wave-front of an optical laser. Effects of a difference between the light paths from the reference and from an object are examined. The angular resolution is measured with optical light and we confirm almost diffraction limited resolution as well as its appropriate function as adaptive optics.