EVER watched a protein fold in real time? Or seen what happens when photons strike a molecule of chlorophyll? Nobody has yet: these events happen too fast and involve structures that are too small to image with today’s technology. The detailed workings of some of the most basic processes in biology, from cell death to photosynthesis, remain unseen by human eyes.
That could all change soon, thanks to a new generation of X-ray lasers that will concentrate their energy 10 billion times more effectively than current instruments, allowing scientists to pry open the physics of the smallest and fastest things. Free-electron lasers, as they are called, have been around for a while, but so far they haven’t been able to produce the kind of radiation needed to resolve the tiniest, most delicate structures. Within the next few years, massive machines capable of generating such laser beams will come online.
Some two-dozen free-electron lasers, spanning the light spectrum from infrared to ultraviolet, are already in operation or under construction. Now efforts are under way to extend the range to X-rays. In the lead and with a budget of $379 million is the Linac Coherent Light Source, scheduled to open in 2009 at the Stanford Linear Accelerator Center in Menlo Park, California. Not to be outdone, the German Electron Synchrotron (DESY) research centre in Hamburg – which inaugurated a state-of-the-art ultraviolet free-electron laser last August – plans to revamp its facility to house the European X-Ray Free-Electron Laser by 2012, at an estimated cost of €908 million. Other projects are further behind, such as at the RIKEN Harima Institute near Kobe, Japan.
Advertisement
In return for this staggering investment, the devices will open the door to experiments that have never been possible before. The lasers could act as strobe lights, capturing images of electrons as they hop between atoms, or of biomolecules as they interact. They could decipher the structure of complex proteins and whole viruses. They could even wring some surprises from DNA, which was first imaged using X-rays in the 1950s, but only in an artificially stretched-out form. “Nobody has ever seen a genome in 3D,” says biophysicist Janos Hajdu of the biomedical centre at Uppsala University in Sweden. Hajdu advises the Stanford and DESY facilities about how to exploit the technology, which could also be used to manipulate matter at a subatomic level. “The race is really heating up,” he says.
So how does it work? Strictly speaking, free-electron lasers are not really lasers at all. In a traditional laser, light or electricity excites matter of some sort – usually a gas or a semiconductor – and that prods the atoms to emit radiation at a specific wavelength. The atoms’ emissions resonate with one another, and what comes out is a tightly focused beam whose photons all swing in time. This “coherent” radiation is to ordinary light what a marching band is to a swarming mob.
Instead of exciting atoms, a free-electron laser uses a particle accelerator to shoot microscopic clouds of electrons through an undulator or “wiggler”, a sandwich of magnets arranged as two metal slabs (see Diagram). The orientation of the magnets alternates along the length of the slabs, so that as electrons fly between them at close to light speed, they traverse regions of magnetic field that switch up and down. The magnetic forces shake the electrons from side to side. This wiggling makes the electrons slow down, converting some of their energy into bursts of photon radiation, which travel in a thin beam along the electrons’ path. The photons interact with the electrons, squeezing them together, which slows the electrons further and sucks more energy out in the form of more photons. By tuning the speed of the incoming electrons, it is possible to get the electron clouds in sync with the photon bursts, a bit like surfers on sea waves – but in this case, the surfers push on the waves from behind, rather than riding along in front. The result is a train of highly concentrated pulses of coherent radiation.
Existing free-electron lasers use their wiggling electrons to produce light beams with wavelengths that range from infrared (1 to 100 micrometres) through visible light to ultraviolet (10 to 100 nanometres). But real-time imaging of two atoms as they combine into a molecule, or of a protein as it folds, requires much shorter X-ray wavelengths (1 picometre to 10 nanometres) and femtosecond (10-15 second) pulses. Shorter wavelengths resolve smaller details, and shorter pulse durations make it possible to record faster events – just as a sports photographer needs a fast shutter speed to freeze a ball in mid-air.
In conventional lasers, the wavelength of radiation is set by the nature of the emitting materials, so they can only produce small amounts of energy at X-ray wavelengths and femtosecond timescales. Free-electron lasers don’t have this limitation. In principle they can be tuned to any wavelength, by varying the energy of the electrons and the geometry of the wiggler. But it is still difficult to reduce the wavelength and duration of their pulses since it means squeezing electrons closer together, against the electrical forces that push apart particles that have the same charge.
Pulse power
The solution: bigger and more powerful devices. The first free-electron machines, built in the 1950s at Stanford University, produced microwaves, not light; the shorter wavelengths of visible light were not achieved until the early 1980s. Meanwhile, wigglers a few metres long have been used for decades to produce non-laser X-ray beams in synchrotrons – the doughnut-shaped particle accelerators that have become the workhorses of structural biology experiments. But the new free-electron lasers will have wigglers up to 150 metres long and will use very concentrated electron clouds. Together, these features will produce highly focused and coherent beams with X-ray pulses down to a femtosecond – 1000 times shorter than today’s best synchrotron sources (Science, vol 308, p 392).
The Stanford and Hamburg machines will use slightly different technologies for their kilometre-long electron accelerators. DESY will produce more pulses per second and shorter wavelengths, but each pulse will be longer. The technology is so new that researchers don’t yet know exactly what they will use it for, or even whether it will work as they expect. “Any machine that brings 10-billion-fold improvements [in concentrating energy] is bound to bring new science, some of which can be predicted, some of which can’t,” Hajdu says.
Perhaps the most coveted application will be to study the structure and behaviour of proteins and other biomolecules. That is already the most common type of experiment with synchrotrons: proteins are arranged in regular crystal structures, typically 2 to 50 micrometres across, and ordinary X-rays are scattered through them. From these images the proteins’ 3D shape can be reconstructed. But obtaining good crystals is tedious and time-consuming. Moreover, many biologically important proteins – especially those that sit on cellular membranes and ferry materials such as drug compounds in and out – tend to resist crystallisation.
A big promise of free-electron lasers is that they will eliminate the need for crystals. That’s because the laser energy will be concentrated enough to image a single molecule, says Jerry Hastings, who heads the science programme for the Stanford device. “You should be able to take a bucket of proteins and drop them one by one into the beam,” he says. Researchers could then resolve the 3D structure of each protein from images of the scattered X-rays.
Hajdu says it is possible that whole viruses could also be studied in this way. Virus researchers typically take fuzzy electron-microscope images, for which they need to freeze the viruses and coat them with gold. But the technique has provided little information about some viruses, including HIV and genital herpes. Also unknown is what’s inside. “Nobody really knows how DNA is so tightly packed into a virus,” Hajdu says. It is hard to predict the results in advance, but researchers hope that free-electron lasers will enable them to take detailed snapshots of live viruses, and possibly help them find lines of attack for new drugs.
Some obstacles remain to be overcome. For one thing, the new lasers might be too strong for their own good. The X-rays will be so concentrated that any molecule they are trying to image will almost certainly be smashed to pieces. The hope is that the pulses will be short enough that by the time the molecule starts to fly apart, the scattered radiation will already have left and will be carrying information to the imaging detectors. Whether it will work out like that, no one knows for sure. “That’s the big question,” says physicist Abraham Szöke, an expert in X-ray imaging at Lawrence Livermore National Laboratory in California. Sceptics say that even if the samples don’t have time to vaporise, they might be distorted.
Another challenge is synchronising the pulses with what they will measure – no mean feat when you are dealing with kilometre-long accelerators and quadrillionth-of-a-second processes. But so far, so good: in a first step toward demonstrating the necessary precision, researchers from DESY, Stanford and elsewhere managed to get the timing right for non-laser X-ray pulses from an accelerator source (Physical Review Letters, vol 94, 114801).
There is still a long way to go. But ultimately if the X-ray laser technology is successful, some long-standing scientific questions could be banished, quite literally in a flash.
Big science
What are free-electron lasers good for? Within a decade, researchers from around the world will converge on the Linac Coherent Light Source (LCLS) in Menlo Park, California, and the European X-Ray Free-Electron Laser in Hamburg, Germany, to find out. Three leading ideas for experiments:
Stripping down
Philip Bucksbaum from the University of Michigan in Ann Arbor plans to use the LCLS to probe new subatomic phenomena, such as what happens when an atom is stripped of its inner shell of electrons while retaining its outer electron shell – a feat that is only possible with highly concentrated X-rays.
Changing state
Jo Stohr at Stanford will use the extremely short pulses of the LCLS like a strobe light to study the critical temperature above which metals such as iron lose the ability to stay permanently magnetised. Another goal is to image chemical reactions – to watch molecules combine, chemical bonds break and proteins fold, all in a quadrillionth of a second.
Simulating stars
Thomas Tschentscher from the DESY centre in Hamburg wants to create entirely new states of matter. In one experiment, the atoms of a small solid sample will have all their electrons removed by an X-ray pulse that is so short it will not give their nuclei time to heat up and expand. This “warm dense” state is thought to be commonplace inside small stars and gas-giant planets such as Jupiter, but nobody has been able to study it in a lab.
