IT IS a realm more alien to our experience than the heart of the sun or the other side of the universe, separated from us by a great gulf not of distance, but of brevity. To the inhabitants of this superfast world, the beat of a human heart is as imperceptibly slow as the drift of the continents is to us.
The world is the atom, and its denizens electrons. Their mercurial movements can be over in just attoseconds – billionths of a billionth of a second – yet they drive our electronic devices, every chemical reaction in nature and every thought in our heads.
What wouldn’t we give to capture them in action. Understand how electrons move, and we might control reactions better, design more effective drugs and better materials for generating energy, and boost today’s sluggish computers to warp speed.
Advertisement
The premiere of such high-speed action movies is fast approaching. As yet the camera work is a little shaky and the picture somewhat grainy, but the technical finesse of the electron film-franchise can only improve with time. Soon we could be watching movies of the microworld in ultimate slow-mo.
It was the late 19th century when photographic pioneer Eadweard Muybridge felt the urge to witness what is merely a blur before our unaided eyes. Famously he captured the motion of a horse galloping, proving the disputed notion that there is a moment in each stride when all four hooves are off the ground. Muybridge used a tripwire as a trigger for his cameras, and then put a series of stills together onto a glass disc. Spinning the disc within his primitive film projector, the , he produced an illusion of motion.
The same principle has kept movie-goers happily fooled ever since. It illustrates the central paradox of film-making: to make motion flow visibly, you first have to freeze it. For a speeding horse, that means a fast shutter on your camera. Eventually, though, mechanical shutters reach their limits, and film or detectors get too little light in the brief exposure to reveal anything. Then what you need is a strobe light that provides intense illumination for just a split second. In the 19th and early 20th century, electric sparks were used to provide such flashes, capturing the motion of sound waves, speeding bullets and other fleeting phenomena.
“It is the central paradox of film-making: to make motion flow visibly, you first have to freeze it”
Sparks can last just a few nanoseconds, but even that is far too ponderous to illuminate the attosecond world of electron movements. “These are the fastest motions outside of the atomic nucleus,” says Ferenc Krausz of the Laboratory for Attosecond Physics in Garching, Germany. Capturing them needs the right kind of flashlight to generate ultrabrief, ultrabright pulses of light that can freeze electrons in action.
That means a laser – in particular, one that produces many different frequencies of light. Carefully manipulated, these will periodically all come into step, reinforcing one another in a sudden burst of light that is over in a trice.
In the late 1980s, of the California Institute of Technology in Pasadena developed such a flashlight to watch chemical reactions unfold on timescales of femtoseconds – millionths of a billionth of a second. He observed the making and breaking of bonds and the formation of ephemeral intermediate molecules, and received the for his efforts.
Getting any faster posed a fundamental problem, however. Light is made of oscillating electric and magnetic fields, and for a light pulse to exist at all its field must rise from zero to a peak and fall back again. The shortest-wavelength lasers produce light in the visible and infrared regions of the electromagnetic spectrum, where one wave cycle lasts just a few femtoseconds. There seemed no chance of making a laser pulse any shorter.
Not so. Around the time that Zewail was doing his seminal work, some laser physicists began to see strange, short bursts of ultraviolet light emerging when they hit certain gases with a powerful blast of infrared laser light. Ken Kulander and his team at the Lawrence Livermore National Laboratory in California . Bright light means strong fields, and if a laser pulse is intense enough, its electric field can rip an electron from an atom. Then, as the field reverses, the kidnapped electron is hurled back towards its atom, emitting its excess energy as a high-frequency ultraviolet or X-ray photon as it slams back in and is recaptured.
This process, called high-harmonic generation, is a little like twanging an elastic strap to create a burst of high-frequency sound waves. Crucially, the rip and the re-collision both happen on very short timescales, lasting much less than a single laser oscillation. Master the fiddly details and you might use an infrared laser to produce X-ray flashes just attoseconds long.
It worked. In 2001 Krausz’s group, together with at the University of Ottawa, Canada, demonstrated that high-harmonic generation could produce flashes of light lasting about 650 attoseconds, showing for the first time that the femtosecond barrier had been broken (). Since then, they and others have trimmed the fastest flashes down to about 80 attoseconds.
This, finally, was a camera-flash fast enough to snap an electron in action. And last year, Krausz’s team did just that. They tuned an infrared laser to create a pulse lasting less than 3 femtoseconds – just a couple of wave cycles – and fired it through neon gas to create high harmonics. Neon atoms grip their electrons tightly, resisting until close to the peak of the intense laser pulse, and so generate the briefest and highest-frequency light.
The twanging of this electron created an ultraviolet pulse lasting less than 150 attoseconds. It is combined with the infrared pulse to hit a nearby sample of krypton gas. This double whammy is vital. Most of the time, an electron bound to an atom sits in an unchanging energy state. To catch one being interesting, you must first throw something at it to provoke it to move, and then snap its picture a moment later.
That is, in a rudimentary fashion, what the two-pulse method does – with an additional trick to take account of the blisteringly quick timescale. To delay the ultraviolet flash very precisely with respect to the infrared pulse, Krausz’s team has developed a special two-part mirror. The outer part focuses the broad infrared laser beam onto the krypton to excite its electrons, while the central part is movable, adding a slight, variable delay to the narrow ultraviolet flash used to take the picture. With a series of different delays in repeated experiments, you get a series of snapshots that can be run together to reveal motion.
A deal of post-production wizardry still goes into making a sensible picture. An electron is not a sharply defined object, but a blurred-out beast, its hazy presence described by a quantum wave function. This wave function adopts various states, or orbitals, each with a different fuzzy outline and energy. Recording the frequencies of ultraviolet light absorbed at a given instant reveals the electron’s energy and quantum state, which tells you the orbital’s shape. Krausz and his team used some CGI trickery to convert this quantum information into a movie of a krypton ion’s outermost electron wobbling its way from dumb-bell to doughnut shape and back again (). The shot of around 40 “frames” lasts about 6 femtoseconds.
“A deal of quantum post-production wizardry goes into making a sensible electron picture”
Similar quantum cunning lies behind a film reel produced in March this year by Corkum and his group. In the instant between having an electron ripped off and slammed back on again, an ionised atom or molecule can be in more than one quantum state. These states interfere with one another, resulting in a series of short flashes that automatically provide snapshots of electron action in the molecule ().
Just as Muybridge’s original film reel of a galloping horse is a little primitive to modern eyes, these short movies are hardly 3D IMAX immersive experiences. But at a million thrills per nanosecond they are action-packed – and promise some profound insights into the workings of nature.
Take the frailties of our existence, for example. “The energy and information transport in biomolecules is based on charge transfer,” says , another Garching researcher. “But we don’t know exactly how that happens – we only know that if it doesn’t work right then people get sick.” The same goes for the signalling process in our nerves, a particular target for Krausz. He wonders whether it’s a purely electronic process involving just charge transfer from one molecule to another, or whether it is also chemical, involving the movements of whole atoms. “No one knows, but we now have the tools to study such questions,” he says.
There is still a hurdle to be jumped before such experiments become reality. The immensely strong electric fields of attosecond light pulses tend to destroy delicate biological samples. The Garching team is now trying to find a way of sticking molecules onto a surface in a way that holds them tight and protects them from electrical disintegration.
More than just voyeurs
At Imperial College London, and his research group are training their own flashlight, which currently delivers pulses as short as 250-attoseconds, on different targets. Among other things, they plan to look at the mechanism behind radiation damage in biomolecules. Radiation damage is thought to begin when a high-energy photon knocks an electron out of a molecule, leading to a chain of events that breaks bonds, damages molecules and may lead to cancerous tissue. No one is too sure about the details of this process on the shortest time scales. “That could have important implications for how radiation damage proceeds,” says Marangos.
Eventually he also hopes to watch photosynthesis in action. This process begins when a photon of sunlight excites an electron in chlorophyll, but exactly how does that electron’s motion lead to the chemical changes that make sugar out of water and carbon dioxide? “It is possible that we are missing key steps,” says Marangos – for instance the mooted role of collective quantum excitations in making the whole thing work. If we can find out, it could enable us to develop artificial photosynthesis to generate power or directly create fuels. It could even help us to bioengineer plants to increase the efficiency of natural photosynthesis.
Ultimately we would like to become more than just voyeurs of the attoworld. We would like to manipulate it, too. That might involve, for example, steering an excited electron to a particular site in a molecule, perhaps to break a bond at a certain place and trigger a chemical change in a specific way. “Perhaps we can make new molecules that would not be possible otherwise,” says Marangos. Krausz’s group is investigating using attosecond laser pulses to switch microchip currents 100,000 times faster than they can be switched today. First indications are encouraging. “We are very excited about this,” he says.
Time will tell. As others look to even faster times (see “Beyond the atto”), Krausz now has his eye on the ultimate subatomic film-maker’s goal: real documentary movies, rather than footage created by computer wizardry from an absorption spectrum and some quantum mechanics. To see the fine spatial detail of an atomic process, you need light of a wavelength shorter than the size of individual electron orbitals, much smaller than a nanometre. That means producing much higher-energy photons.
Krausz says he has a few technical barriers to surmount before he can produce and manipulate that ultimate light, but he is confident they will be overcome. “Then the long-held dream will come true,” he says. Coming soon to theatres near you: the shortest story ever told.
Beyond the atto
Can we make light pulses shorter even than 1 attosecond? Perhaps, if research from the group of at the JILA research institute in Boulder, Colorado, is anything to go by. They are investigating how to make high harmonic light (see main story) from a long-wavelength infrared laser. The longer wave cycle means that an electron ripped from the atom by the light is accelerated for longer before the field reverses, gaining more energy before it is pinged back. The result is extremely high-energy X-rays that could be packed into a pulse of just a few attoseconds or fewer.
at the Laboratory for Attosecond Physics in Garching, Germany, is following the same lead. Achieving the very highest energies turns out to be particularly tricky, though. The infrared pulse rips out some electrons even before it reaches its central peak, resulting in messy interference. Krausz won’t describe his workaround in any detail as it is currently under peer review, but it involves combining three separate infrared laser pulses so that interference cancels most of them out, leaving almost all of the power in a single wiggle of the wave.
Is there much point in these blisteringly fast pulses? After all, nothing much of interest in nature is faster than the attosecond movements of electrons. “A sub-attosecond pulse might allow for some control of atomic dynamics not previously possible,” says Kapteyn. “But in most cases this is an inconsequentially short period of time.” He is instead looking to the potential of longer pulses from what is effectively a desktop X-ray laser. “Its spatial resolution will be a thousand times better than an ordinary medical X-ray,” says Kapteyn. His group is now collaborating with electronics companies to picture the digital bits written on a disc drive, and watch the effects of heat flow in nanoscale components.
