Smart X-ray Optics
The Smart X-ray optics project is a collaboration of University College London, Kings
College London, University of Edinburgh, University of Leicester, Daresbury Laboratory,
Gray Cancer Institute and Birmingham University. This is a 4 year project funded under
the Basic Technology programme run by the UK Research Councils.
We are proposing a revolution in the capability to exploit X-rays. Current X-ray optics
cannot take advantage of the corresponding very small wavelengths, especially in the
case of large scale optical devices. While the wavelengths of X-rays are several orders of
magnitude smaller than that of visible light, X-ray optical systems are still typically outperformed by their visible-light counterparts. Applications for X-ray technologies are
found in many disciplines from medicine to astronomy, and the current state-of-the art
poses limitations in areas as diverse as exploring and treating cancer cells, and answering
some of the fundamental questions about the Universe itself. We propose to build upon
the results of an earlier Basic Technology proof of concept project, to demonstrate a
diverse set of major advancements in the development and application of X-ray optics to
scientific research in a wide range of disciplines. Cross-programme common aspects
identified in the proof of concept project include:
modelling by finite element analysis, ray tracing and wavefront propagation
the reduction of surface roughness and figure errors
optical testing to compare with modelling
actuators for adaptive and active optics
exemplar applications
The objectives of the project are:
To develop new types of X-ray focusing optics and a versatile tabletop X-ray source
that together will generate micrometer beams of X-rays. Such a system will be used to
irradiate precisely parts of living cells, tissues and indeed living organisms. This, a highly
sought-after technique in radiation biology, also has potential applications in
microelectronic manufacturing.
To develop a prototype X-ray telescope with sub arc second resolution and potentially
large filling factor whose optical form is actively controllable.
X-ray optics and X-ray sources have many applications in the scientific domain, either
directly, e.g., in X-ray telescopes, or indirectly, e.g., in the next-but-one generation of
integrated circuits which will require ever finer feature sizes, ultimately deliverable only
by short wavelength technologies. Improved optics will permit the efficient utilisation of
smaller X-ray sources, providing many of the capabilities of synchrotrons, but in a
laboratory environment and at a fraction of the infrastructure cost. X- ray microscopy
employing synchrotron sources is used in the fields of biological and medical sciences,
earth and planetary sciences, environmental and soil sciences, material and surface
sciences and archaeology. Improved X-ray optics will facilitate the development of
stereoscopic X-ray microscopy, providing 3-D images. Work with X-ray microprobes
may also be extended into three dimensions allowing controlled irradiation of volume
materials, specifically biological cells and tissue.
The two specific application areas highlighted in this application will impact cell
radiation studies and astronomy. Low and medium energy (up to 20 keV) bright X-ray
sources, not currently available commercially, that can generate micrometre and submicrometre focused X-ray beams will allow for exciting new radiobiological studies,
targeting individual cells (and even parts of cells) in single cell studies, in tissue and in
living organisms where penetration is important. The lack of suitable optics for X-ray
astronomy can be seen most starkly in the performance of the two large X-ray
observatories presently in orbit. In one (the NASA Chandra Observatory) the angular
resolution is 0.5 arcsecond which is obtained using 4 stacked zerodur reflectors, with an
outer mirror diameter of 120cm, which provides a geometrical sensitive area of 1145cm2,
an aperture filling factor of only 0.1. In the XMM-Newton observatory, 5 arcsecond
resolution is achieved using 3 modules of 58 stacked reflectors with an outer foil
diameter of 70cm. Each module has a geometrical area of 1750cm2 and an aperture filling
factor of 0.45. Other foil telescopes have been developed for X-ray astronomy with even
greater numbers of stacked foils, e.g. BBXRT, Astro-D, and XSPEC with 101, 120 and
143 foils respectively but with generally poorer imaging performance than the XMM
telescope. The dramatic difference between Chandra and XMM comes about because at
present it is not possible to fabricate high performance, thin X-ray mirror shells despite a
vast investment in the UK and Europe. Even just achieving the collecting area of XMM
at a resolution of Chandra would, for example, enable astrophysicists to detect the hot Xray emitting winds from starburst galaxies in the Hubble deep field, and so directly
measure the ‘feedback’ that has been proposed as the mechanism that regulates galaxy
formation.
Background to the project
Traditional X-ray optics for terrestrial applications suffer from
poor efficiencies (zone plates and, in some cases, multilayer mirrors);
poor spatial resolution capabilities (grazing incidence & multilayer mirrors, mono- &
poly-capillaries, and compound refractive lenses);
large aberrations (grazing incidence mirrors);
strong chromatic aberration (zone plates); and
low angular and wavelength bandpasses (zone plates and periodic multilayer mirrors).
Although zone plates feature strongly in this list, they are often the optic of choice as they
provide by far the best spatial resolution of any type of X-ray optic to date. However,
compared to optics used at longer wavelengths they are extremely poor, with f-numbers
of at best ~200. This programme intends to overcome these deficiencies by developing
new active/adaptive X-ray optics. One of the technologies proposed is microstructured
optical arrays (MOAs), which use grazing incidence reflection through consecutive
aligned arrays of channels. Similar in concept to polycapillary and microchannel plate
optics, they are more flexible. Bending the arrays allows variable focal length, while
flexing parts of them provide adaptive or active systems. Custom configurations can be
designed, using ray tracing and finite element analysis, for sub-keV to several-keV
applications. The other main area of technological development proposed is in the field of
large high throughput X-ray optics. Bimorph mirrors have been used over many years in
astronomical adaptive optics systems. This type of mirror has already been used as a
focusing device in grazing incidence X-ray systems, but only in essentially a one
dimensional single strip system. A consequence of grazing incidence is that the
geometrical area of optics as projected onto the aperture is small compared with its
physical dimension, a problem that becomes increasingly difficult as the energy increases
and the critical grazing incidence angle falls. For this reason it is often necessary to
concentrically stack reflectors close together to provide a high throughput. This in turn
leads to a second problem associated with stiffness: very stiff optics tend to be more
massive and therefore provide a low filling factor with a stack, i.e., the geometrical area
of the stack is much greater than the subtended geometrical area of the mirror surfaces.
Using simple spherical optics, which can be manufactured to a very high precision, a
resolution of 0.15 arcsec has been achieved, but severe field distortion means that their
use is problematic for any practical application. We believe that through active optics
based on piezo-electrically deformed thin shells we can significantly improve on this
figure and lead the way to a new era in X-ray astronomy. The imaging performance of
such foil telescopes depends upon diffraction (negligible in our context), surface
roughness and alignment errors. Surface roughness can be further divided into microroughness and mid-spatial frequency structure with the foil. While micro-roughness
cannot be addressed by active optics the scale size of the mid-spatial frequency structure
(1–5mm for Al foils) could be addressed with a very significant improvement in
performance. Similarly the very small alignment errors involved could also be
ameliorated.
Work at UCL
An actively controlled prototype X-ray telescope of dimensions 2030 cm will be
developed based on an initial hyperbola-parabola thin shell design whose figure will be
improved through embedded piezoelectric actuation. The aim is to deliver a near
diffraction limited spot size. A baseline design will involve thin nickel (1mm) mirror
substrates produced by electroforming on Zerodur or stainless steel moulds. These shells
will then be integrated with thin (<500 μm) piezoelectric unimorph actuators.
Modelling work will build on the experience and results gained in the proof of concept
research programme and modelling work developed at University of Leicester. To
approach the diffraction limit the optics must manipulate the X-ray wavefronts with
atomic precision and the adjustment, alignment and stability required of the optical
elements is very demanding. We will develop software to design and simulate diffraction
limited X-ray optics. Finite element analysis modelling of the prototype system will be
developed that will include the mirror substrate and piezo-actuator physics. This will be
used to test the efficacy of different surface configurations and the sensitivity of the
performance to tolerances in surface figure and alignment. When used in conjunction
with surface metrology we will be able to predict the expected performance of the optics
thus providing a baseline against which we can assess measured performances. Many
components of the software have already been developed and utilized in several space
instrumentation projects, the ROSAT WFCXUV telescope, the XMM-Newton telescope,
the JET-X/Swift telescope and most recently the very large area XEUS X-ray mirror.
However, none of these has approached the diffraction limit. The design and performance
prediction software developed will be used for both the micro and large optics proposed
within this programme.
Mirror fabrication: Mirror moulds will be manufactured at UCL using existing optical
fabrication and testing facilities including those at the new UCL optical fabrication
research centre at the OpTIC technium in North Wales. The moulds will be gold coated
by vacuum deposition and then electroformed nickel coated to produce the nickel
substrates. These mirrors will then be integrated with the piezoelectric actuators
developed at University of Birmingham. The mirrors will then be mounted in a test rig
and connected to the actuator drive electronics.
Laboratory testing: Testing of these types of X-ray optics can be carried out in two
complementary ways: pencil beam testing, in which a very narrow and highly collimated
X-ray beam is scanned across the surfaces as a metrological probe, or full aperture testing
in which the optic is tested in its operational configuration. The former is best performed
at a synchrotron facility where the X-ray flux available is large. If the optic has a large
aperture then the latter is much easier using a long-beam test facility. We propose to use
the 27m beam facility at the University of Leicester to conduct full aperture testing of
medium sized optics. The source, pipe and test tank have recently been refurbished using
SRIF funding and the test tank is installed within a cleanroom designed for the handling
and integration of X-ray optics. Space qualification tests including thermal cycling,
radiation, vibration and thermal vacuum will also be conducted on the completed system
at MSSL