SPIE Digital Library Get updates from SPIE Newsroom
  • Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:
SPIE Photonics West 2017 | Register Today

SPIE Defense + Commercial Sensing 2017 | Call for Papers

Journal of Medical Imaging | Learn more


Print PageEmail PageView PDF


The XEUS x-ray telescope: solving the space-borne x-ray optics dilemma

Future x-ray astronomy telescopes in space will need vastly-increased collecting areas while maintaining resolution in the arcsecond class.  A novel optics fabrication technique can solve this problem for the XEUS mission.
7 March 2006, SPIE Newsroom. DOI: 10.1117/2.1200601.0073

Since the first imaging x-ray astronomy telescope was launched in 1978 on the Einstein Observatory, 1 the capabilities of telescopes have evolved enormously with substantially increased spatial resolutions and collecting areas. Because x-ray astronomy has to be performed in space, practical considerations such as mass, volume, and in-orbit deployment are paramount. The next generation of telescopes for x-ray astronomy needs to provide significantly-improved sensitivities compared with those provided by current x-ray observatories. This increase in capability will allow detailed spectroscopic x-ray investigations of objects that currently are at the limit of detectability, such as the first massive black holes at redshifts of 5–10.

There is no material that efficiently refracts x-rays, so reflection is used for astrophysical applications. Reflection occurs efficiently over a range of photon energies only if the x-rays are reflected with a small grazing angle (typically <1°). The preferred geometry is the Wolter-I configuration,2 comprising parabolic and hyperbolic surfaces. The telescopes on-board Chandra 3 and XMM-Newton 4 have driven technology in different directions. Chandra provides high-spatial resolution imaging (∼1" HEW) at the expense of collecting area, while for XMM-Newton, the imaging capability (15" HEW) was relaxed and the collecting area maximized.

The relatively large collecting area of XMM-Newton was achieved by dramatically increasing the nesting of electroformed-nickel mirror shells. Each of three mirror modules has a mass of 350kg and provides an area of 0.15m2 at 1keV. The mass of the Chandra and the combined three XMM-Newton mirrors are similar, while the mirror area of XMM-Newton is approximately six times that of Chandra, but with a spatial resolution roughly a factor 20 lower. These state-of-the-art telescopes have closely-nested shells manufactured by direct machining and super-polishing of individual shells, or by replication from super-polished mandrels. In both approaches, the shells were individually integrated and aligned within the telescope assembly. For future astrophysics missions, such inherently-heavy optics is extremely costly to manufacture and launch.

A future sensitive broadband mission, such as XEUS, 5 which is currently under study by the European Space Agency, could fulfil a wide range of investigations proposed by the astronomical community. However, the implementation of such a mission requires the development of novel, high-performance, lightweight mirror technology. We offer a new approach to assembling an x-ray telescope using modular Hhigh-performance pore optics (HPO). The main science goals of XEUS impose a requirement for an effective collection area of several square meters at 1keV, while a spatial resolution of 2-5" half-energy width (HEW) is needed to avoid source confusion at these extremely low flux levels. 6–7 An extremely high-performance x-ray-reflecting surface is needed to provide these capabilities, and optics based on current technologies cannot meet these requirements (see Figure 1).

Figure 1. Area densities versus resolving power for a number of cosmic x-ray telescopes. Three different types of mirror technology have been used so far: thin foils, thin replicated shells, and thick monolithic shells. Foil optics have not provided arc-second-type resolutions, but are light-weight. Monolithic optics with large apertures are too heavy for space applications, and replicated optics are challenged by problems of both mass and angular resolving power. Strikingly, the performance of all the telescopes launched in the last 25 years lies close to the dashed line indicated. The position in this figure of the optics required for the next generation of x-ray astrophysics missions, such as the X-ray Early Universe Spectroscopy (XEUS), indicates the challenge necessary to meet the very demanding science requirements.

The Wolter-I geometry can be approximated using conical surfaces, where closely-packed shells, whose pitch is smaller than the optics point spread function (PSF), require a single figure of rotation only. The system can be miniaturized with very thin (100μm) cylindrically-curved plates. For example, with a 50m focal length, a 2" resolution can be achieved with shells stacked closer than 1mm, but only ∼40mm in length. In this design the optics modules are much more compact and lighter due to the thin mirror shells. High spatial resolution can be achieved, provided the mirror shells remain stiff and retain the appropriate shape. This advanced modular design is what we call high-performance pore optics.

Self-supporting modules are produced using a structure of thin ribs spaced between the shells, forming the pores of the basic unit. The shells must have a high figure accuracy, and surfaces must be extraordinarily smooth (surface roughness of <0.3nm rms) to avoid scattering and degradation of the PSF. Silicon wafer technology developments, such as chemo-mechanical polishing, have resulted in the production of atomically flat wafers with extremely low surface roughness, and these wafers have the required surface properties for mirror shells. Gapless three-dimensional attachment can be achieved using compact sensor technology.

With the modular approach, the current generation of 30cm silicon wafers does not limit the telescope size, since a large telescope can be assembled out of HPO modules based on Si mirror plates, as outlined in Figure 2. Processed silicon wafer components are stacked on top of one other while being bent in the azimuthal direction to form a single monolithic unit that is intrinsically very stiff. A number of these sub-petal units are integrated, aligned, and fixed to form the major component of the mirror petal. We have produced the basic element of such optics, an HPO sub-petal unit consisting of a stacked structure of tens of Si mirror plates.

Figure 2. The segmented optics entrance aperture hierarchy. 1: A structured silicon plate, made from a 30cm round wafer, forms the basic building block of the sub-petal. The typical size of a plate is 6.5cm square. Plates can be optionally coated with Au, Pt, or multi-layers on one side to improve x-ray reflectivity. 2: Ribbed plates are successively assembled until a high-performance pore optic (HPO) is formed. 3: Two of the HPOs are attached and accurately aligned to form a segment of a Wolter I optic (not shown here). 4: Several HPOs are joined and co-aligned to form a petal. The large telescope aperture is then populated by such petals, which are subsequently actively aligned in orbit.

The starting element of the stack is a figured, polished, silicon former on which the modular HPO is built prior to integration and fixation within the full petal structure (see Figure 2). Since the reflectance angle increases with the radius of curvature, a change in grazing incidence angle of about 1" needs to be compensated for in each mirror plate within the stack. A special industry standard treatment of the wafers has been developed and validated that allows the highly accurate implementation of such a wedge and its fixation

After assembly, the performance of the prototype optics is routinely measured using a synchrotron radiation facility. The rms roughness of the mirror material was confirmed by standard reflectometry to be ∼0.2nm. Pencil-beam testing of selected areas of a wafer at an energy of 3keV revealed that single pores have angular resolutions of <2" HEW. Extended beam illumination over many pores on scales of a few cm2 provides resolution approaching the XEUS requirement, limited to date by the initial mandrel quality. The HEW of the obtained PSF was found to be 5" when corrected for the intrinsic beam width of the same order (Figure 3).

Figure 3. Prototype modular silicon x-ray optics (top left) assembled on top of a polished silicon former. The mirror elements are bonded to each other, held in place by van der Waals forces only, and can be disassembled if necessary. Annealing of the HPO brings it in to a non-reversible assembled state. The shape of the silicon mirror plate is controlled during assembly by interferometric analysis (bottom left). The HPO point spread function (PSF) is re-constructed from the measurement of 480 single reflections of 3keV X-rays on an integrated area of ∼ 2cm2 , as indicated in the figure. The predicted telescope function, as given in the right part of the figure, is of slightly better performance, or HEW, since it accumulates more intensity in its center. Note the low level of scattering in the wings of the PSF.

An example of a prototype HPO sub-petal is shown in figure Figure 3. To account for the double reflection of Wolter I optics, two HPOs are combined to form a tandem pair. These are then integrated into a larger petal structure forming a segment of the aperture of a large diameter light-weight x-ray mirror. Such a mirror will combine high-resolution x-ray optics with a huge collecting area. For an increase in mass of only a factor of three (from 1000 to 3000 kg), a mirror of 6× 6m2 in size can be manufactured that provides more than fifty times the collecting area of the largest x-ray mirror flown to date (XMM-Newton), while achieving significant improvement in the required spatial resolution of a few arc seconds. HPO technology provides the single critical element in the construction of a feasible mission and has wider applications in other areas of high-energy astrophysics, such as a large area collector for timing studies

David Lumb, Marcos Bavdaz, Tone Peacock and Arvind Parmar
ESA Office of Science Advanced Concepts and Payload Technology
Noordwijk, Netherlands
David Lumb is an Instrument Scientist in the European Space Agency Advanced Concepts and Science Payloads Office, investigating novel payloads for future science missions. He was Calibration Scientist for the XMM-Newton project, and formerly developed CCDs for X-ray astronomy applications
Marcos Bavdaz is head of the technologies section of the Advanced Concepts and Science Payloads Office of ESA
Tone Peacock is Head of the Advanced Concepts and Science Payloads Office of the European Space Agency
Stefan Kraft, Marco Beijersbergen, and Maximillien Collon
Cosine Research
Leiden, Netherlands

1. R. Giacconi, ,
Astrophys. J.,
Vol: 230, pp. 540, 1979.
2. H. Wolter,
Ann. der Phys. series, 6,
Vol: 10, pp. 94, 1952.
3. M. C. Weisskopf, H. D. Tananbaum, L. P. V. Speybroeck, S. L. O'Dell,
Proc. SPIE 4012,
pp. 2, 2000.
4. F. Jansen,
Astron. Astrophys.,
Vol: 365, no. L1, 2001.
5. M. J. L. Turner,
ASP Conf. Proc.,
Vol: 251, pp. 230-233, 2001.
6. D. M. Alexander,
Astron. J.,
Vol: 126, pp. 539, 2003.
7. T. Miyaji, R. Griffiths,
Astrophys. J.,
Vol: 564, no. L5, 2002.