杏吧原创

Andrew Greentree wants to play tricks with light. At the University of Melbourne in Australia he鈥檚 aiming to do something to photons that no one has done before: trap them in cages and make them talk to each other. Unlike particles of matter such as electrons, photons are notoriously unsociable. They pass like ships in the night, even going straight through one another without noticing. Yet if they could be made to interact they might be compelled to form a peculiar new kind of quantum material 鈥� one made of light.

Though this wouldn鈥檛 be a material in the everyday sense of the word, like a solid you can touch, it would behave like one in some important ways. Materials get their characteristic properties from the way the atoms within them interact, and the more we know about this, the more we can do with them. Likewise, the way photons interact when they form a quantum material could give us insights into how real materials with quantum properties work. This could, for example, help us explain exactly why high-temperature superconductors can make electricity flow with no resistance at all.

Why is all this important? First, because it may signal the start of a materials revolution that could lead to new kinds of computers and other devices. Materials made of light could be used to model 鈥� and then build 鈥� optical circuits that store and process information far faster than anything around at the moment. They may even offer a route to a practical quantum computer 鈥� something that as yet exists only in theory 鈥� that would be able to perform calculations millions of times faster than a supercomputer.

So far, photon materials also exist only in theory, but there is a buzz of anticipation that they may not be far away. In the past few months, three separate groups, Greentree鈥檚 among them, have developed models showing that it should be feasible to build and test such materials within a few years. 鈥淭he excitement really comes from our ability to make what are essentially new forms of matter,鈥� says Charles Tahan of the University of Cambridge, one of Greentree鈥檚 collaborators.

The idea that light can be manipulated to produce material-like behaviour is not entirely new. In recent years, researchers have experimented with so-called optical lattices, and exploited these devices to model the quantum behaviour of materials. Optical lattices use criss-crossing laser beams to trap and control atoms, which can be kept rigidly in position as if they were in a crystal, or allowed to move and interact freely as if in a gas. Technical constraints, however, have prevented researchers from studying what happens in these gizmos down to the level of individual particles. Greentree and the others hope to remedy that by building complementary gadgets in which photons act upon each other in more easily measurable ways 鈥� and that鈥檚 the key to the proposed photon materials.

Most of the time, it鈥檚 a good thing photons don鈥檛 interact the way atoms and electrons do. It鈥檚 the reason we can see: light reflecting off a mirror never backs up in a traffic jam, and two crossing flashlight beams never fight with one another. In the air and other ordinary materials, photons just don鈥檛 talk to each other. 鈥淭o make that happen you need to do something special,鈥� Greentree says.

Ideas for doing just that have circulated for a decade. In 1997, Ata莽 Imamoglu at the University of California, Santa Barbara, proposed that a special kind of box or 鈥渃avity鈥� could be made to absorb one, and only one, photon. That photon would change the resonant properties of the cavity, preventing it from absorbing further photons, a process called photon blockade. Any photon that enters the system would effectively repel all others. 鈥淧ut one in,鈥� says Dimitris Angelakis of the University of Cambridge, 鈥渁nd the second one will bounce back.鈥�

In 2005, Jeff Kimble and colleagues at the California Institute of Technology in Pasadena showed a practical way to do this (Nature, vol 436, p 87). They built a small cavity between two mirrors, 80 micrometres across, in which only photons of a particular wavelength could exist. To this cavity they added a supercooled caesium atom and showed in experiments that once the cavity-atom complex absorbed one photon, it could not absorb another, because it was no longer in a resonant state that allowed it to attract light.

This striking result spurred three research teams led by Angelakis, Greentree and Martin Plenio of Imperial College London to develop the idea further. Working independently, they began to design full-blown quantum materials out of photons. Their approach, says Angelakis, was to get 鈥渓ight to talk to light through matter鈥�, and to do so in such a way that many photons could be made to interact at the same time ().

What each group now envisions begins with a photonic crystal 鈥� a structure that can channel photons rather in the way that an irrigation system controls the flow of water across fields. Made of glass, plastic or diamond, the crystal allows only photons with certain wavelengths to flow inside it. A suitable pattern of holes drilled into the material using an ion beam can create cavities where photons can sit, surrounded by regions where they cannot.

Each cavity then needs something that will produce a photon blockade. Angelakis and Greentree each proposed adding an atom with exactly two energy levels, the difference tuned to match the energy of photons trapped by the cavity (Nature Physics, vol 2, p 856). This would produce an array of atom-containing cavities, each capable of taking up one and only one photon. Add photons to the cavities and they will interact with their nearest neighbours by effectively repelling one another. 鈥淵ou end up with something that looks like a classic material system,鈥� says Greentree, 鈥渂ut made out of light.鈥�

鈥淵ou end up with something that looks like a classic material, but made of light鈥�

How exactly? In models, the photon materials show behaviours every bit as rich as those of real materials. One of the hallmarks of any material is the ability to change its state 鈥� liquid water freezing into solid ice, for example 鈥� a process known as a phase transition. All three research groups have shown that their proposed photon systems would undergo the quantum equivalent of a phase transition, from a so-called 鈥淢ott insulating鈥� state, with a photon contained in each cavity unable to move, to a strikingly different 鈥渟uperfluid鈥� state, in which each photon can flow without resistance over the entire array, like an electron in a superconductor.

This effect really proves the principle. 鈥淚t鈥檚 been an open question for a long time as to whether you could ever see a quantum phase transition for light,鈥� says Angelakis. The answer, it seems, is a resounding yes. The transition could be triggered by changing the energy levels of the two-level atoms so that the atom-cavity systems no longer produce photon blockade. That could be induced in a fraction of a second by shining a laser onto all the cavities at the same time.

Plenio鈥檚 team go further. They reckon that each cavity could contain not one but an ensemble of atoms, and have shown that photon blockade can be realised by driving the atoms with laser light (Nature Physics, vol 2, p 849). Each group has confirmed in calculations that the behaviour of any photon in the array is intimately linked to the behaviour of others, as is the case with particles that make up liquids, solids and other states of matter. 鈥淭he interactions are so strong in such systems,鈥� says Gerard Milburn of the University of Queensland, Australia, 鈥渢hat thinking in terms of the bits, the particles, that make them up can be misleading. You need an effective theory for the system as a whole.鈥�

鈥淭his is really elegant work,鈥� says Pieter Kok at the University of Oxford. 鈥淚t will be fascinating if they can show it experimentally.鈥� If that can be done, physicists will have a new tool to help them understand a range of weird quantum effects. Richard Feynman first suggested in 1981 that simulations of quantum systems such as photons and electrons could best be carried out with other quantum systems: you should simulate like with like. With a 鈥渜uantum simulator鈥� made of interacting photons, researchers might gain clues into the baffling nature of high-temperature superconductors and other exotic materials in which strong quantum interactions play a dominant role.

It is thought, for instance, that the bizarre behaviour of such materials might have something to do with what goes on when a system is close to the boundary of a quantum phase transition, such as the transition from Mott insulator to superfluid. At such a boundary, fluctuations in the microscopic organisation of a material become unusually large and seem to dominate the system鈥檚 properties (New 杏吧原创, 28 January 2006, p 40). At the moment researchers have few techniques for probing the fine details of this state in real materials. 鈥淲e haven鈥檛 yet cracked the basic rules,鈥� says Jan Zaanen of Leiden University in the Netherlands. 鈥淭here鈥檚 a great demand for well-controlled experimental models of these systems. This light stuff might work.鈥�

A quantum simulator could also be an important step towards the fabled quantum computer (see 鈥淐omputing by light鈥�). That鈥檚 because these simulators can be thought of as rudimentary computers which carry out calculations in strictly quantum terms.

But how to actually build one? Greentree鈥檚 team proposes that an alternative and promising way to construct photon materials would be to use not two-level atoms in cavities, but defects in the crystal lattice of thin sheets of diamond. These defects 鈥� created where a nitrogen atom replaces a carbon atom next to a vacant spot in the lattice 鈥� can bind an electron and act like simple atoms. A system of this sort would naturally have many quantum bits, or qubits, that could be exploited to do computations.

Greentree and his colleagues have started working on this idea. They believe they can build the right kinds of devices within a few years, using existing fabrication techniques. 鈥淲e鈥檝e never made something this complicated,鈥� Greentree admits, 鈥渂ut it is like things being made in silicon already. We certainly think it should be feasible.鈥� The end result could be a quantum computer made of diamond.

Even if this proves impossible, Greentree and the others won鈥檛 be too upset. 鈥淭here is so much we still have to learn about quantum computing,鈥� he says, 鈥渢hat I hesitate to make any predictions. This is a great toy to let us understand new physics. We can combine things in new ways and see brand-new effects, hopefully ones never seen before.鈥�