British scientists have announced a breakthrough in x-ray microscopy which could be used to picture individual atoms in living cells without using a lens. The method, they claim, marks significant progress towards the 'ultimate microscope', with the potential to reach the greatest resolution allowed by an x-ray's wavelength: a span of around 0.1 nanometres, roughly the width of a single carbon atom.

John Rodenburg and colleagues from the University of Sheffield and the Paul Scherrer Institute in Switzerland hope their lensless x-ray microscope, which has just been unveiled,¹ could eventually take high-resolution 3D images of any molecular structure. And since their innovation relies on a special type of computation, rather than specific equipment, the technique could also be used to boost the power of optical microscopes² and maybe even electron microscopes. Since lenses aren't required to focus the beams, 'it is even possible to contemplate a solid-state microscope, built into a single chip with no optical elements at all,' Rodenburg told Chemistry World. His team have just received a £4.3 million Engineering and physical sciences research council (EPSRC) grant, starting in April 2007, for their work.

Traditional microscopes use glass lenses to focus light waves through an object. For shorter-wavelength beams of electrons or x-rays, which provide higher resolution, the lenses are based on magnetic fields or light-scattering gratings called 'zone plates'. But with that sophisticated technology comes a host of technical problems, so the images are about one hundred times more blurred than should theoretically be possible.

The alternative to such lenses is to shoot waves right through the object and collect their remains on the other side, working out what the atomic structure of the material must have been based on the patterns formed by the overlapping, diffracted waves. For regular crystalline structures, this is not too difficult: the structure of DNA, for example, was solved by x-ray diffraction in the 1950s. But rays diffracting through disorded atomic structures are far harder to process. One way to work around this problem is to focus on a tiny, isolated part of the sample, but this is no use for biologists trying to image a whole cell.

However, Rodenburg's method means that objects of any size or shape can be imaged, even if they have a disordered structure. 'The technique has revolutionary implications for x-ray imaging of all classes of specimen,' he said.

Lose the lens

The new technique relies on collecting diffraction patterns from several overlapping areas in space, which provides information about how the rays interfere with each other after they have been diffracted through the object. This interference, allied with a sophisticated computer program and a mathematical equation that Rodenburg and his postdoctoral research assistant, Helen Faulkner, took all of 2004 to figure out, can be related back to what the rays' previous phase changes must have been, giving a complete picture of the structure.

Rodenburg compares the idea of using overlapping diffraction patterns to the techniques used in radioastronomy, where signals from different radio telescopes are added together to simulate one large detector. 

'This is wonderful work, demonstrating what John Rodenburg and his group have shown many times in simulation,' commented John Spence, who works on imaging techniques at Arizona State University, Tempe, US. 

So far, the lensless x-ray microscope has achieved an impressive - though hardly 'ultimate' - 50nm resolution. That's almost as good as the very best conventional diffraction x-ray imaging, which uses low-energy x-rays. However, those soft x-rays cannot penetrate through an object anywhere near as far as Rodenburg's high-energy x-rays can, and they must also be used in vacuum conditions - not the biologist's best friend.  

Although Rodenburg has only used his microscope for surface imaging so far, he expects that the method should be extendable to three dimensions. And with some technical improvements, like brighter light sources, Rodenburg feels his microscope could image close to the 0.1nm theoretical limit for x-rays. 

'We need to be careful at this point in overstating the applications,' said J Murray Gibson, associate laboratory director at Argonne National Laboratory, Chicago, US. 'But this is a very clever approach, and I do believe it is going to be a practical method of lensless imaging.' 

Richard Van Noorden