For more than a decade, cosine has been developing silicon pore optics (SPO), lightweight modular X-ray optics made of stacks of bent and directly bonded silicon mirror plates. This technology, which has been selected by ESA to realize the optics of ATHENA, can also be used to fabricate soft gamma-ray Laue lenses where Bragg diffraction through the bulk silicon is exploited, rather than grazing incidence reflection. Silicon Laue Components (SiLCs) are made of stacks of curved, polished, wedged silicon plates, allowing the concentration of radiation in both radial and azimuthal directions. This greatly increases the focusing properties of a Laue lens since the size of the focal spot is no longer determined by the size of the individual single crystals, but by the accuracy of the applied curvature. After a successful proof of concept in 2013, establishing the huge potential of this technology, a new project has been launched in Spring 2017 at cosine to further develop and test this technique. Here we present the latest advances of the second generation of SiLCs made from even thinner silicon plates stacked by a robot with dedicated tools in a class-100 clean room environment.
Continuing improvement of Silicon Pore Optics (SPO) calls for regular extension and validation of the tools used to model and predict their X-ray performance. In this paper we present an updated geometrical model for the SPO optics and describe how we make use of the surface metrology collected during each of the SPO manufacturing runs. The new geometrical model affords the user a finer degree of control on the mechanical details of the SPO stacks, while a standard interface has been developed to make use of any type of metrology that can return changes in the local surface normal of the reflecting surfaces. Comparisons between the predicted and actual performance of samples optics will be shown and discussed.
While predictions based on the metrology (local slope errors and detailed geometrical details) play an essential role in controlling the development of the manufacturing processes, X-ray characterization remains the ultimate indication of the actual performance of Silicon Pore Optics (SPO). For this reason SPO stacks and mirror modules are routinely characterized at PTB’s X-ray Pencil Beam Facility at BESSY II. Obtaining standard X-ray results quickly, right after the production of X-ray optics is essential to making sure that X-ray results can inform decisions taken in the lab. We describe the data analysis pipeline in operations at cosine, and how it allows us to go from stack production to full X-ray characterization in 24 hours.
Silicon Pore Optics (SPO), developed at cosine with the European Space Agency (ESA) and several academic and industrial partners, provides lightweight, yet stiff, high-resolution x-ray optics. This technology enables ATHENA to reach an unprecedentedly large effective area in the 0.2 - 12 keV band with an angular resolution better than 5''. After developing the technology for 50 m and 20 m focal length, this year has witnessed the first 12 m focal length mirror modules being produced. The technology development is also gaining momentum with three different radii under study: mirror modules for the inner radii (Rmin = 250 mm), outer radii (Rmax = 1500 mm) and middle radii (Rmid = 737 mm) are being developed in parallel.
The Advanced Telescope for High-Energy Astrophysics, Athena, selected as the European Space Agency's second large-mission, is based on the novel Silicon Pore Optics X-ray mirror technology. DTU Space has been working for several years on the development of multilayer coatings on the Silicon Pore Optics in an effort to optimize the throughput of the Athena optics. A linearly graded Ir/B4C multilayer has been deposited on the mirrors, via the direct current magnetron sputtering technique, at DTU Space. This specific multilayer, has through simulations, been demonstrated to produce the highest reflectivity at 6 keV, which is a goal for the scientific objectives of the mission. A critical aspect of the coating process concerns the use of photolithography techniques upon which we will present the most recent developments in particular related to the cleanliness of the plates. Experiments regarding the lift-off and stacking of the mirrors have been performed and the results obtained will be presented. Furthermore, characterization of the deposited thin-films was performed with X-ray reflectometry at DTU Space and in the laboratory of the Physikalisch-Technische Bundesanstalt at the synchrotron radiation facility BESSY II.
As part of the ongoing effort to optimize the throughput of the Athena optics we have produced mirrors with a state-of-the-art cleaning process. We report on the studies related to the importance of the photolithographic process. Pre-coating characterization of the mirrors has shown and still shows photoresist remnants on the SiO2- rib bonding zones, which influences the quality of the metallic coating and ultimately the mirror performance. The size of the photoresist remnants is on the order of 10 nm which is about half the thickness of final metallic coating. An improved photoresist process has been developed including cleaning with O2 plasma in order to remove the remaining photoresist remnants prior to coating. Surface roughness results indicate that the SiO2-rib bonding zones are as clean as before the photolithography process is performed.
Ni-based multilayers are a possible solution to extend the upper energy range of hard X-ray focusing telescopes
currently limited at ≈79:4 keV by the Pt-K absorption edge. In this study 10 bilayers multilayers with a
constant bilayer thickness were coated with the DC magnetron sputtering facility at DTU Space, characterized
at 8 keV using X-ray reectometry and fitted using the IMD software. Ni/C multilayers were found to have a
mean interface roughness ≈ 1:5 times lower than Ni/B4C multilayers. Reactive sputtering with ≈ 76% of Ar
and ≈ 24% of N2 reduced the mean interface roughness by a factor of ≈ 1:7. It also increased the coating rate
of C by a factor of ≈ 3:1 and lead to a coating process going ≈ 1:6 times faster. Honeycomb collimation proved
to limit the increase in mean interface roughness when the bilayer thickness increases at the price of a coating
process going ≈ 1:9 times longer than with separator plates. Finally a Ni/C 150 bilayers depth-graded mutilayer
was coated with reactive sputtering and honeycomb collimation and then characterized from 10 keV to 150 keV.
It showed 10% reectance up to 85 keV.