Tabletop X-rays: miniaturising the synchrotron

Scientists have developed a compact X-ray source that could rival the power of large-scale synchrotrons

X-rays are among the best tools for finding out about the properties of matter. The high-energy radiation’s ability to pass through many types of matter, revealing hidden details; the way it refracts through the planes of crystals to unlock their complex structures; its ability to pump energy into systems: all of these have been used by scientists – and increasingly by engineers – for almost a century.

But generating X-rays of high enough intensity takes large equipment. Although engineers are achieving highly promising results from the bright X-ray source of the Diamond synchrotron in Harwell, this is only possible because of the sheer size of the device: 500m in circumference, and costing more than a quarter of a billion pounds. But this could be set to change, thanks to work by an international consortium of scientists who have devised a source of X-rays as powerful as those from a building-sized synchrotron – but small enough to sit on a lab bench.

The team, which included researchers from Imperial College London, the University of Michigan and the Instituto Superior Téchnico Lisbon, has generated a pencil-thick beam of high-energy X-rays from apparatus about a metre across. Benefitting from a €10m (£8.7m) grant from the European Commission’s LaserLab Europe project, the system was developed at the University of Michigan’s Centre for Ultrafast Optical Science, using its high-power petawatt laser beam, Hercules.

Supersize: the X-rays from the compact source would be useful for studying cracks in turbine blades

The X-rays were generated by passing the intense laser beam of Hercules through a plasma, which in turn was generated by the energy of the laser stripping away the electrons of helium in a 3-10mm jet of the inert gas. The laser generates a phenomenon known as a ’wakefield’ in the plasma – an electric field that can be more than a thousand times stronger than those generated by conventional synchrotron technologies. This field accelerates the electrons from the plasma to incredibly high velocities – imparting kinetic energy in the GeV range – over a distance of 10mm.

This is the first part of the process to generate X-rays. To actually produce the radiation, the electrons have to be induced to change direction; this forces them to slow down and hence shed some of their energy. In synchrotrons, this is done using magnetic fields; other wakefield sources have also used this technique, which has produced visible or near-visible light. The LaserLab group used a different method: a phenomenon called betatron oscillations, which arises because of the separation of charge created when the negatively charged electrons are stripped away from their positively charged nuclei.

“The beam is what is known as a collimated, coherent beam – a variety that is very useful for engineering”

The laser forms a bubble of plasma within the gas jet, and the beam of electrons is forced to ’wiggle’ by the charged particles surrounding it.

This disrupts the flow of electrons and generates X-rays with energies of 1-100keV, which is around a thousand times brighter than previous laser-driven X-ray sources. Moreover, the beam doesn’t spread out along its path; this is what is known as a collimated, coherent beam – a variety that is very useful for science and engineering, because of the closely controlled position and known energy of the X-rays. The beam is also very narrow – only a micron across – which would allow researchers to use it to probe very small features and samples.

’This is a very exciting development,’ said Stefan Kneip of Imperial’s physics department, the lead author on the research paper the group has published in the journal Nature Physics to describe their results. ’We have taken the first steps to making it easier and cheaper to produce high-energy, high-quality X-rays.’ The system produces X-rays that rival those from synchrotron sources hundreds of metres long, he added. ’Although our technique will not now directly compete with the few large X-ray sources around the world, for some applications it will enable important measurements that have not been possible until now.’

One apparent limitation of the system is that it is only capable of producing very short pulses of X-rays; around a femtosecond. This means that it isn’t capable of conducting studies that require a sustained source; those will still require large synchrotrons such as Diamond.

However, femtosecond pulses have a very valuable ability. Like the brief flashes of a strobe light, they can freeze the motion of events that happen extremely fast. At the femtosecond scale, this means they can isolate events happening at atomic sizes.

A system such as ours could eventually increase dramatically the resolution of medical imaging systems using high energy X-rays

Zulfikar Najmudin, Imperial College

The leader of the Imperial team, Zulfikar Najmudin, explained that the applications of this could stretch from life sciences to engineering. ’We think that a system such as ours could have many uses,’ he said. ’For example, it could eventually increase dramatically the resolution of medical imaging systems using high-energy X-rays, as well as enable microscopic cracks in aircraft engines to be observed more easily. It could also be developed for specific applications where the ultra-short pulses of these X-rays could be used by researchers to “freeze” motion on unprecedentedly short timescales.’

Of course, the research is still at a very early stage; although the equipment that generated the X-rays is only a metre on each side, the petawatt laser used to generate the plasma is hardly a commonplace tool. ’Our technique can now be used to produce detailed X-ray images,’ Najmudin said. However, he added: ’High-power lasers are currently quite difficult to use and are expensive, which means that we are not at a stage where we could make a cheap new X-ray system widely available. Laser technology is advancing rapidly, so we are optimistic that in a few years there will be a reliable and easy-to-use X-ray source available that will exploit our findings.’

In the meantime, Najmudin is concentrating on investigating the physics behind the new X-ray source. ’We are currently developing our equipment and our understanding of the generation mechanisms so that we can increase the repetition rate of this X-ray source.’

from the archive in synch

In 1956, The Engineer reported on the early days of what’s now perhaps the world’s best-known synchrotrons at CERN in Geneva

CERN is committed to the task of building and operating an international laboratory. The main installations will consist of two accelerators for research on high-energy particles. One of these accelerators is a 600MeV synchrocyclotron, the design of which was finalised in 1955 with a view to the completion of the equipment in 1957. The second is a 25GeV proton synchrotron, which is scheduled for completion in 1960.

In the circumferential path described by the particles there are 50 magnet periods and 100 magnet units. Each magnet unit consists of a focusing sector and a defocusing sector. Each magnet unit is 4.4m long and the free field gap between the magnet units is 1.6m long. Near the instant of injection of the particles the maximum beam section is calculated to be 6cm high by 10cm wide, assuming certain tolerances in the construction of the machine. All the magnet units have to be aligned accurately to a circle, the radial tolerance being 0.6mm at the end of the units.