SPIE Membership 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 2018 | Call for Papers




Print PageEmail Page

Optical Design & Engineering

X-ray optics releases new potential in mature technology

From OE Reports Number 185 - May 1999
30 May 1999, SPIE Newsroom. DOI: 10.1117/2.6199905.0006

Study of the x-ray spectrum, its sources, detectors, and uses, is by no means new. For example, virtually every school boy and girl has experienced the benefit of a clinical x-ray to reveal broken bonesor to alleviate the fears of a parent. Scientists have put this technology to even greater uses, using x-ray radiation to explore organic compounds and crystal structures using x-ray diffractometry and x-ray fluorescence.

Although x rays are not new, the past several years has been marked by innovative development in new materials and approaches to x-ray optics, lifting x-ray exploration to new levels of importance by increasing the capabilities of a range of widely used systems from the scanning electron microscope to mammograms and clinical x rays.

Polycapillary lenses make great filters

At the center of polycapillary, nested cone and multilayer graded x-ray optics lies the critical angle. For most materials, the index of refraction of x rays is slightly less than that of air. Unlike the optical spectrum, x rays' short wavelength means that they do not refract very much through normal optical materials, but rather are absorbed or pass right through with only small changes in propogation. However, if the incident grazing angle is less than the critical angle as determined by the atomic weight of a given material (low-atomic-number materials generally have higher critical angles), the x-ray will "bounce" off the surface of the material, slightly changing its path. X-ray optics, therefore, seek to nudge the radiation in the right direction as opposed to the much simpler visible optics that can grab, bend, and literally twist light into any shape necessary.

A polycapillary lens, also called a Kumakhov lens after its Russian inventor, is basically an array of hollow optical fibers. Like optical fibers, capillary fibers direct low-incident angle x rays that act like a ping-pong ball tossed down a pipe. Researchers at X-Ray Optical Systems, Inc.(Albany, NY) among others, have built on Kumakhov's work, making arrays of borosilicate hollow fibers, each containing hundreds or thousands of individual channels, and bending them so that the x rays strike the walls at less than the critical angle, but enough so that the radiation is redirected to a new path. Arrays of these capillaries can actually redirect the x rays to a focal point.

Figure 1. Marshall Joy of NASA inspects a prototype capillary X-ray optic, built by X-Ray Optical Systems, comprising 2,500 fibers (inset); each fiber contains about 750 capillaries. This scanning electron micrograph shows a single fiber approximately 0.5-mm wide. (Photos courtesy of NASA/Marshall and X-ray Optical Systems)

Borosilicate glass is similar to the material used to clad visible and IR optical fibers. According to Carolyn MacDonald of the University of SUNY (Albany, NY), although metals allow for higher critical angles, creating hollow tubes with the requisite internal surface smoothness is much more difficult. The variance of the internal walls has to be controlled to within 10s of angstroms, which excludes the use of most metal with conventional manufacturing techniques. Other considerations enforced by the small critical angles and the desire to capture as much of the radiation as possible is the fiber's internal diameter. According to David Gibson, CEO of X-Ray Optical Systems, their x-ray optics can contain hundreds or thousands of fibers, each containing hundreds of channels varying between 8 and 50 µm (Figure 1).

X-ray Optical Systems offers two types of polycapillary designs: multifiber and monolithic. Multifiber components are mainly used for collimating diverse x-ray sources. External screens hold the fibers with their hundreds of channels. Internal diameters stay constant in this configuration, although they can be bent to 'focus' the photons as dictated by the application. Company officials say these optics, with parallel beam footprints ranging from 1 to 25 cm2, can produce beam spot sizes of 400 µm full-width half-maximum. The monolithic devices are smaller. Comprised of fibers fused together, these lenses do not require the screen external structure. They are often tapered for additional focusing power, providing spot sizes of 20 to 200 µm.

Although these optics hold promise for imaging applications, current uses are limited to spectroscopic, diffraction measurement and other applications in semiconductor wafer inspection and other research and development areas, according to Gibson. MacDonald is developing the ground work for capillary lenses that will increase the contrast of mammograms and other clinical x-ray imaging applications because of their inherent filtering capabilities1. Because these optics only guide radiation that is very close to the optical axis, they exclude Compton scattered photons (as high as 90 percent of transmitted radiation during a mammogram) to approximately 1 percent. Conventional antiscattering grids allow 20 percent of scattered x rays to pass through to the imaging film at 20 KeV.

Nested cones focus the stars

The ability of Polycapillary optics to focus x-ray radiation above 2 KeV may be its most beneficial characteristic. NASA, working with X-Ray Optical Systems and the University of SUNY at Albany, hopes that capillary optics will extend the spectroscopic sensitivity of future telescopes like the Chandra Telescope (previously the Advanced X-ray Astrophysics Facility [AXAF]) beyond its 10 KeV limitation out to 80 KeV. While this certainly bodes well for polycapillary x-ray concentrators, several options exist for x-ray applications with energies at or below 2 KeV.

One of those options is nested cones, first developed by NASA for astronomic observation. Gibson's company has developed polished nickel nested cones for focusing collimated x-ray radiation, or for collimating x-ray light from a diverse source. These optics, made from nickel or heavier elements such as gold, are parabolic in shape. The focal point of the parabola is removed, allowing light close to the optical axis to pass through untouched. Photons off axis, but still below the critical angle, are reflected inward, creating a parallel beam. By combining many concentric cones, the optic is able to focus (or collimate depending on the source location in respect to the optic) a greater percentage of the radiation.

In addition to astronomy, collimated x-ray beams have a wide variety of uses including x-ray collection during scanning electron microscopy fluorescence and wavelength dispersive spectroscopy (WDS). In WDS, a collimated beam is the difference between a large, expensive experimental set-up and a more cost-efficient exercise. During WDS, x-ray photons transmitted or reflected by an object are directed at a rotating diffracting crystal. As the crystal is rotated, it diffracts x rays of a particular energy, providing spectroscopic information on the sample under study. According to Gibson, when the beam is collimated, the system can use a flat crystal, which reduces the computational burden. However, if the beam is dispersive, a curved crystal has to be used to capture a larger amount of the radiation and the "geometries become much more complex," he said.

Similar arguments apply to x-ray fluorescence in material science. This technique is often used in silicon wafer inspection for contaminates. By focusing a lower energy x-ray beam on a wafer, the system can avoid exciting the silicon layer while revealing the contaminants underneath.

Intensity improvements in multilayer mirrors

Creating parallel beams from divergent laboratory x-ray sources has also been among the chief tasks of graded multilayer reflective mirrors, also called Göbel mirrors after Herbert Göbel of Siemens (Muenchen, Germany). For this application, layers of different reflecting materials are deposited on a curved surface (parabolic for collimating, elliptical for creating an intense line for diffractometry, and planar multilayer mirrors for divergent sources).

Recent developments by Göbel and his associates have improved the intensity and spectral purity of the multilayer mirror while adding to the stability of the underlying structures. In the past, Göbel mirrors have used alternating layers of heavy materials such as tungsten, molybdenum and nickel, interchanged with layers of lighter elements such as carbon, silicon and boron carbide. Earlier this year, Siemens announced intensity improvements gained by switching W/Si and W/B4C layers with WSi2/Si, Ni/Mg and Ni/B4C.

Göbel's research found that layers of WSi2/Si showed better smoothness and layer conformity than W/Si, even for short periods of more than 100 layer pairs2. As a result, the mirrors maintained the same high levels of reflectivity while showing better spectral purity. By changing the layers to Ni-based materials, the mirrors showed an intensity gain of 16 compared with gains of 6 of previous work. Göbel estimates that this comes close to the theoretical limit of 20 to 25.

Siemens and the Institute of Materials Research (Geesthacht, Germany) were also able to improve the stability of the mirror by improving the uniformity of the underlying substrate. Conventional multilayer mirrors are created by bending the underlying silicon substrates into the optical shape after deposition, or by pressing the deposited silicon wafer into a negative shape and then glueing the mirror onto a stiff backing. According to Göbel and his associates, this presents several problems, such as imperfections created by dust on the negative, resulting in surface imperfections on the mirror that reduce reflectivity. The glue also degrades over time as a result of aging and x-ray irradiation.

A new technique uses prefabricated quartz substrates. Mechanical and ion-beam polishing shapes the quartz into the required parabolic figure, resulting in a reduction in surface errors by about a factor of 10 compared with bent and glued mirrors.

Refraction in x rays

Although optics that depend on total external reflectance such as polycapillaries, nested cone and multilayer mirrors are highly transmissive for the correct incident angle, attention is gaining for refractive x-ray optics for imaging applications. Because x rays are absorbed easily in most materials, refractive optics were once considered a pointless pursuit. However, by using low-atomic-weight materials with carefully controlled parabolic shapes, Bruno Lengeler at the Institute of Technology of the State of Nordrhein-Westfalen (RWTH; Aachen, Germany) and his associates have demonstrated refractive optics for imaging and microscopic analysis applications.

"You can use them just like you use an optical lens. They're handy, robust, and withstand wide-beam or undulator beam sources with no problem," said Lengeler. Contrary to visible optics, x-ray refractive focusing optics are concave because x rays have a slightly lower index of refraction when passing through solids compared with air. (It's the opposite for visible light and glass optics).

Figure 2. Refractive optics enable this X-ray micrograph of a free-standing gold mesh (above) and a Fresnel zone plate (below), a structure of 85 concentric gold rings, whose width decreases towards the outside. All the gold rings are 1.15 µm thick. The outermost ring has a width of 0.3 µm.

By placing several of these optics into a row, Lengeler has imaged features around 0.3 µm (Figure 2). Previous designs that were basically a series of holes drilled into metal exhibited significant spherical aberration. The new design, however, exhibits only slight chromatic aberration up to 50 KeV, which Lengeler said can be corrected by adjusting or removing some of the elements in the cylindrical, multi-element lens. Although these lenses suffer in transmission -- passing between 1 and 40 percent of the radiation through, depending on the material -- they can significantly reduce the focal length of many x-ray imaging systems from several 10s of meters down to 0.5 m.

Lengeler's most recent results are expected to appear in the Journal of Applied Physics in the near future. The paper will reveal experimental results using aluminum lenses with an outer rim diameter of 1 mm and inner diameters at the peak of the parabola of 20 µm. The entire 40-lens element is about the size of a matchbox, he said. Future experiments could involve the use of lower atomic-weight materials such as beryllium and boron. When considering materials, he said, one must pay close attention to homogeneity, because holes and gaps in the material result in unwanted scattering.

Regardless of their approach, new approaches to x-ray optics is giving greater utility to a technology that is more than a century old. Perhaps the advances of x-ray optics themselves may not give great cause for reflection within the scientific community, however, it seems likely that the 'effects' of the x-ray optics 'cause' will have a major impact. Smaller, affordable x-ray optics will likely make a previously unmanageable area of the spectrum readily available to many more laboratories and researchers. And as research expands in crystal and protein analysis leading to new understandings of disease, and new methods are constructed to verify next generation microchip designs, x-ray optics are likely to be eye to eye with startling new discoveries.


1. Beam Collimation, Focusing, Filtering and Imaging with Polycapillary X-Ray and Neutron Optics, C.C. Abreu and C.A. MacDonald, Center for X-Ray Optics, University at Albany, State University of New York, Albany, NY 12222, In Press, Physica Medica.

2. Improved Graded Multilayer Mirrors For XRD Applications, C. Michaelsen, P. Ricardo and D. Anders at Institute of Materials Research, GKSS Research Center, 21502 Geesthacht, Germany; M. Schuster, J. Schilling and H. Göbel, Siemens AG, ZT MF 7, Otto-Hahn-Ring 6, 81739 Munchen, Germany.

R. Winn Hardin
R. Winn Hardin is a science and technology writer in Fairbury, NE.