TO GET a taste of the latest thing to hit the world of quantum physics, spend
a few minutes in St Paul鈥檚 Cathedral in London. Squeeze through the crowds in
the entrance, climb the steps to the whispering gallery, lean against the cold
stone wall and speak softly. As the sound waves bounce along the curved walls,
your words are whisked around the vast dome. Someone 40 metres away on the other
side will hear your every word.
It鈥檚 an age-old trick, yet the principle that architect Christopher Wren used
to create the whispering gallery is helping researchers build some of the
smallest lasers ever made鈥攋ust 50 micrometres across, or about the
diameter of a human hair. But their tiny size isn鈥檛 the only extraordinary
property of these lasers. They are powerful too, packing the same punch as
lasers sixty times as large, yet they consume only a third as much energy. And
unlike with normal lasers, physicists can alter the colour and direction of
their light beams simply by tweaking them with electricity.
The development of lasers with these characteristics is just what is needed
for the next great technological leap forward鈥攐ptical computing. Such
lasers are perfect for use as high-speed optical switches, and for replacing
wires with sparkling beams of light. They could enable oceans of data to be
swiftly sent between processors, for instance.
Advertisement
And it doesn鈥檛 stop there. The chaos at the heart of these lasers could allow
researchers to study the poorly understood relationship between quantum
mechanics and chaos theory, in which small initial perturbations鈥攖he
flapping of a butterfly鈥檚 wings鈥攃an lead to large changes. By adjusting
the behaviour of the photons inside the laser, physicists may at last have a
powerful tool with which to unravel the long-standing mysteries of quantum
chaos.
But how could 17th-century architecture lead to a new generation of lasers?
All lasers need two things: a source of photons and a space鈥攃alled the
cavity鈥攊n which to amplify them. In conventional lasers, the cavity is
formed by two mirrors. Blast the atoms inside the cavity with energy and they
emit light. This, in turn, triggers other atoms to emit light鈥攁 process
known as light amplification by stimulated emission of radiation鈥攁nd the
cavity fills with photons. Let just a few per cent of the photons escape through
one of the mirrors, and you have a bright beam of laser light.
Most lasers are designed along similar lines: they are long and thin, and
inside, light bounces from one end to the other. But there鈥檚 a problem with this
geometry. Make a rectangular laser too small, and it produces only a feeble beam
of light.
The difficulty lies with balancing the amount of light inside the cavity with
the amount escaping as the laser beam. If the cavity is too short鈥攎uch
less than a millimetre in length鈥攊t is almost impossible to trap the light
for long enough to get sufficient amplification. You can pump huge bursts of
energy into the laser, but you get no more than a trickle of light out. This is
the conundrum facing engineers trying to miniaturise powerful lasers.
It was in an effort to break this impasse that physicists borrowed a trick
from Wren. About ten years ago, they began to build circular
cavities鈥攕olid discs made of a material with a high index of
refraction鈥攖hat behave like microscopic whispering galleries. Create
photons inside and if they hit the walls at a shallow angle, they are reflected
instead of passing through. Rather than bouncing back and forth between two
mirrors, the laser light is trapped, skimming round the curved surfaces of the
disc, just like sound inside a whispering gallery.
This design was only a partial success, however. The lasers can be made so
small that hundreds could fit onto the head of a pin. And they do require little
power to operate. But it鈥檚 almost impossible to get a useful laser beam out. As
the photons race round inside, the laws of quantum mechanics allow only a few to
鈥渢unnel鈥� out through the walls. So most of the laser light is trapped, and a
fraction dribbles out in all directions, with a power of just a few
microwatts鈥攏othing like the intense beams engineers need.
For years, researchers have struggled to harvest the rich source of photons
inside a microdisc. But two years ago, Federico Capasso, the head of the
Semiconductor Physics Research Department at Bell Laboratories in New Jersey,
happened to read a popular book on chaos theory. He was interested to learn
that, within a container shaped like a flattened oval, a moving particle will
behave chaotically.
Capasso is the co-inventor of the quantum cascade laser, a device made from
stacks of thin layers of semiconductors. Apply a small voltage across these
layers, and the electrons in the semiconductor give off their energy as photons.
In this type of laser, the light is confined in a single plane: make the cavity
a flattened oval, thought Capasso, and there might be some interesting chaotic
effects. He passed the project to a colleague, Claire Gmachl.
Meanwhile, in the department of applied physics at Yale University, theorist
Douglas Stone was studying the strange behaviour of misshapen laser cavities. He
had discovered that inside a deformed laser cavity, light will often follow
chaotic pathways. Eventually, he believed, a degree of order should emerge
within the chaos, and the light should escape in specific, predictable
directions. Deforming a circular laser cavity could turn the dispersed emission
of whispering gallery lasers into narrow beams. In 1997 he and colleague Jens
Noeckel published a paper in Nature (vol 385, p 45) describing the
possibilities that chaos could open up.
Then Capasso and Stone met, and their groups began to collaborate. Capasso
and his team experimented with deformed laser cavities made from tiny discs of
gallium indium arsenide and aluminium indium arsenide, while Stone鈥檚 group at
Yale interpreted the results and pointed out new paths to follow.
The balance between chaotic and ordered behaviour in a distorted laser cavity
is extremely delicate. Light can bounce around inside a laser cavity in a number
of different ways. Only when it is reflected between the mirrors in a particular
pattern or set of pathways called modes can the light become intense enough to
form a laser beam. In a circular cavity, the photons can arrange themselves in
one kind of mode鈥攖he whispering gallery mode鈥攖hat approximately
follows the walls. But deform the circular cavity into an oval, and this mode
breaks up. The calm procession of photons breaks down into chaotic anarchy.
Whenever the photons strike the walls too steeply, they escape from the cavity
altogether. With too much light leaving, laser action stops.
But Gmachl continued to build new devices, each with a cavity a tad more
deformed than the previous one. Eventually, when the diameter of the cavity had
been reduced by about 15 per cent, something entirely unexpected occurred. The
disc leapt into life, abruptly transformed into a powerful 鈥渕icrodisc鈥� laser
blazing with bright beams of light that shone outwards like the spokes of a
microscopic wheel. Gmachl had created a tiny laser with four 10-milliwatt laser
beams, each a thousand times as powerful as that of a circular microdisc. The
researchers were stunned. 鈥淚t was something that neither Doug nor I realised
could happen,鈥� Capasso says.
It took four months to come up with an explanation. 鈥淲e were trying to use
our old theory to explain the results,鈥� he says. 鈥淏ut these modes are
conceptually different鈥攖hey don鈥檛 exist at all in the circular
肠补惫颈迟测.鈥�
Finally Evgenii Narimanov, working in Stone鈥檚 lab, realised that the only
possibility was something called a 鈥渂ow tie鈥� mode, in which the photons follow a
distinctive pattern inside the cavity (see Diagram).
鈥淗e presented it to me and there was a Eureka moment,鈥� says Stone.
The research was published last summer in Science (vol 280, p 1556).
Looking back, the bow tie mode is an obvious way to achieve directional
emission, Capasso admits. At the corners of the bow tie, where the light hits
the walls of the cavity at a steep angle, light can easily escape. The result is
four high-power beams shining out from the disc. The regular pattern of the bow
tie mode emerges from the chaos of the cavity, as stable as a rocky outcrop in a
raging sea.
Noeckel estimates that only about 60 per cent of the photons are reflected
back into the cavity when they strike the walls, far less than the 90 per cent
typical of other lasers. Yet this is enough to stimulate the emission of more
photons and produce powerful laser beams. And it needs just microwatts of power
to work.
Out of the chaos
Part of the key to this behaviour is the way the beams focus at the centre of
the cavity, making the light amplification process more effective. But the most
amazing thing is that the bow tie mode owes much of its power to the chaos that
surrounds it. 鈥淭he chaos is a very subtle point, but very important,鈥� Capasso
says.
When the cavity is distorted to just the right degree, the photons begin to
follow the regular pattern of the bow tie mode, and the energy that was
dissipated by the photons鈥� chaotic motion begins to feed into the bow tie mode.
Individual photons bounce around chaotically, then join the bow tie mode, then
either escape within one of lasers beams or return to their chaotic wanderings
again. Nobody understands exactly what鈥檚 happening yet, but the effect is that
the power in the laser goes up. 鈥淭he bow tie mode seems to concentrate the
energy,鈥� says Stone.
That concentration of energy could be crucial. 鈥淥ne of the most striking
things about lasers is that it鈥檚 very hard to make them small,鈥� says Stone.
鈥淭his device produces useful power at the 50-micrometre scale, and could well be
shrunk down to 5 micrometres.鈥� Impressive as this is, the researchers don鈥檛
think microdisc lasers will replace existing laser technologies. No one is
planning to take over the hi-fi market with microdisc laser-powered CD players.
鈥淚t is simply very difficult to displace an entrenched technology,鈥� says
Capasso.
But tiny, efficient lasers are ideal for sending information along networks
of optical fibres, and optical connections have several advantages over sending signals down a piece of wire
(see 鈥淭he future鈥檚 bright鈥�, New Scientist, 25 October 1997, p 40).
You can send as much as a billion times more information
down an optical fibre than along conventional metal connections, for instance.
And as electrical circuits shrink, the resistances of narrow wires lead to large
power dissipation and damage to the wires鈥攏ot a problem faced by photons
in optical fibres.
Microdisc lasers could even move data between chips inside a computer. Stone
calls it 鈥渋ntegrated optics鈥�. 鈥淚t鈥檚 about doing with light many of the things
that you can do with silicon integrated circuits,鈥� he says. 鈥淔or moving
information quickly over more than a centimetre, it鈥檚 much better.鈥�
And it鈥檚 in this field that microdisc lasers offer huge advantages over
traditional lasers. For starters, they can send light down four or more optical
fibres simultaneously. There are also many variations of the bow tie mode, each
slightly different and each with its own beam geometry and colour.
By 鈥渢uning鈥� the device with a pair of tiny electrodes, engineers should be
able to switch between these modes, changing the direction of the laser鈥檚 beams
at will. The team at Bell Labs is working on ways of achieving this, but Capasso
won鈥檛 yet give anything away. He believes that computer networks will soon be
linked by 鈥渟mart鈥� optical switches that can route signals down one fibre or
another, making optical networks even more efficient. Microdisc lasers might be
one way to make such switches: surround a laser with an array of optical fibres
and in theory you could send light down almost any combination of fibres by
changing the mode of the laser. The commercial potential is huge.
But there is a snag. The existing lasers produce infrared light. If they are
to be useful in optical computers and switches, they must emit light of much
shorter wavelengths, such as blue or green. Developing the technology is not
impossible, but it will take a few years.
While Capasso鈥檚 team works on the next generation of optical technologies,
theorists like Stone are using the laser to explore the mysteries of quantum
mechanics. The microdisc laser could prove to be a powerful new tool for
uncovering exactly how quantum physics relates to chaos theory, a poorly
understood area of physics
(see 鈥淲here two worlds meet鈥�, New Scientist, 18 May 1996, p 26).
The chaotic photons in the laser could enable hitherto
impossible experiments to be performed. 鈥淢ost quantum systems are just too
complicated,鈥� says Stone.
Noeckel believes that the bow tie mode may turn out to be an example of
chaotic systems that defy the randomness expected of them, and instead carve out
a regular pattern, a phenomenon known as scarring. 鈥淪carring is a hot topic in
quantum chaos,鈥� he says. But whatever turns out to be going on in the heart of
the microdisc laser, it is good news for chaos theory, Noeckel believes.
Generally, chaos has been perceived as a negative force鈥攕omething to be
avoided. Chaos can cause random, useless motion, it鈥檚 true, but new and useful
types of motion such as the bow tie mode are sometimes created in the transition
to chaos. Discovering these strange phenomena, says Noeckel, is like finding a
whole new set of tools. 鈥淐haos is the enabler,鈥� he says.
As the microdisc laser shows, the very randomness of chaos can produce weird
and wonderful behaviour, providing researchers with all sorts of new
opportunities. 鈥淭here is a lot of good quantum physics in these systems,鈥� he
says. 鈥淧eople are going to pursue it for years.鈥�