麻豆传媒

GOOD old Dmitri Mendeleev. He did a remarkable job with his periodic table,
even leaving spaces in the right places for elements that had yet to be
discovered. But ninety years after his death, physicists are on the verge of
building a whole new table鈥攖his time with a distinctly futuristic
design鈥攆rom artificial atoms.

These brand new atoms can be combined into strange molecules with unearthly
properties. And it doesn鈥檛 take huge particle accelerators or a star鈥檚-worth of
energy to achieve. All our ambitious physicists need are microscopic
semiconductor dots and a handful of electrons. The tough part will be getting
from theory to the real world鈥攖ry bonding single artificial atoms together
to function as a single molecule or keeping those tricky new creations at well
below room temperature so they can show off their strange powers.

Atoms are little more than boxes that hold electrons. It is the way these
electrons are arranged that gives each element its unique character, defining
its chemical or magnetic properties and determining how it will bond to other
atoms. With this simple picture in mind, physicists have started to build their
own tiny 鈥渂oxes鈥� from semiconducting materials.

Just shove in as many electrons as you like and suddenly you have an
artificial element which can be used to test the rules of quantum physics or to
create matter with properties never before seen on Earth. Change the arrangement
and interactions of the electrons inside your atoms and you could build
fantastic new materials, or change the very fabric of the world. Imagine
switching the properties of materials at will, turning magnets on and off, or
converting conductors into insulators. 鈥淲e have already seen a lot of physics
that simply hasn鈥檛 been, or can鈥檛 be seen in real atoms,鈥� says Ray Ashoori of
the Massachusetts Institute of Technology, one of the pioneers of the field.

To build an artificial atom, physicists use lithography to etch out their
鈥渂ox鈥� from a solid lump of semiconductor. The result is a tiny pillar just a few
micrometres tall that is made from several layers of material, one stacked upon
another. About a third of the way up the pillar is an indium gallium arsenide
layer just a few nanometres thick, sandwiched between two thin insulating
layers. In this layer鈥攃alled a quantum dot鈥攖he artificial atom is
formed.

Caught in a trap

In real atoms, electrons are trapped by the attraction of a positively
charged nucleus that lies at their centre. Artificial atoms, on the other hand,
use an external electric field to ensnare their electrons. The top of the
semiconductor pillar is coated with a thin layer of chromium to act as an
electrode. Put a small positive charge on it and electrons in the 鈥渟ource鈥�, an
electron-rich gallium-arsenide layer at the base of the pillar, are attracted
upwards. As they move up the pillar, they bump into a barrier鈥攖he
insulating layer beneath the dot.

However, under the influence of the electric field and thanks to the weird
world of quantum mechanics, they can 鈥渢unnel鈥� their way through to the dot beyond
(see Diagram). By carefully adjusting the strength of the
electrical field, physicists can attract anything from one to one thousand
electrons into the quantum dot. Switch the electric field off, and they are
trapped.

Held inside the dot, these electrons behave as if they are all alone. This is
because while the quantum dot contains thousands of semiconductor atoms, each of
which has its own electrons, they are all tightly bound to their own nuclei and
invisible to the trapped electrons. So the micrometre-sized quantum dot behaves
just like its smaller cousin with a diameter one thousand times smaller.

Electrons in a real atom are restricted to certain orbitals or 鈥渟hells鈥�
around the nucleus. The energy of each electron depends on which shell it
occupies. A similar situation exists in the artificial atom. The dimensions of
the quantum dot are so small that the electrons are confined to discrete energy
levels鈥攋ust as you鈥檇 expect from the laws of quantum mechanics. They
display all the properties shown by electrons in a real atom. Push them up to a
higher energy level by changing the electrical field around the quantum dot, for
example, and the excited electrons will lose their extra energy by emitting
light of exactly the wavelength predicted by the Schr枚dinger
equation鈥攖he mathematics that describes the fundamentals of quantum
mechanics.

In real atoms, the electrons are thought to orbit around the nucleus in a
series of concentric shells. It is the way that electrons fill these
3-dimensional shells that gives Mendeleev鈥檚 periodic table its shape. But in the
quantum dots built by Leo Kouwenhoven of Delft University of Technology in the
Netherlands, the orbits are very different. They form a series of circles,
spread out on a flat disc鈥擪ouwenhoven nicknames them 鈥減ancake鈥� atoms.

This two-dimensional symmetry means that the number of electrons needed to
fill each shell is different to a real atom. Kouwenhoven has even written out a
new periodic table for his pancake atoms
(see Diagram).
Its rows are shorter than those of Mendeleev鈥檚 original鈥攅ach shell accepts
fewer electrons鈥攂ut Kouwenhoven鈥檚 table also shows certain
similarities.

Just as in Mendeleev鈥檚 original, when two elements have a similar number of
electrons in their outer shell they display similar characteristics and are
placed in the same column. In Kouwenhoven鈥檚 table, there is a column of
chemically inactive substances with outermost electrons that can only be removed
with a large blast of energy. The inert gases in Mendeleev鈥檚 table require a
similar burst of energy鈥攌nown as the ionisation energy鈥攖o lose an
outer electron.

Taking Kouwenhoven鈥檚 pancake atoms apart also reveals that their electrons
show the same quantum-mechanical interactions as those in real atoms. In pancake
helium, for example, Kouwenhoven and his team noticed the effects of the
鈥渆xchange force鈥�, a quantum-mechanical effect seen in real multi-electron
atoms.

But they have found completely novel effects too. Applying a magnetic field
of a few tesla鈥攕trong, but fairly easy to produce in a lab鈥攖o
pancake helium atoms has an astounding effect. The magnetic field created
completely new energy levels. By forcing electrons to move between these states,
they were able to see electron transitions that have never before been seen in
real atoms. Similar transitions are predicted for helium atoms near the
ultra-high magnetic fields鈥攏early a million tesla鈥攁ssociated with
pulsars and white dwarf stars. But since such fields are entirely inaccessible
to researchers here on Earth, no one had ever seen the proof.

Designer solids

The reason for this extraordinary behaviour is that real atoms are so small
that an immense magnetic field is needed to get a large magnetic flux into the
atom. Since artificial atoms are scaled-up versions of their real cousins, the
same magnetic flux can be generated inside using a scaled-down magnetic
field.

Such potential will keep artificial atoms at the forefront in investigating
quantum mechanics. Researchers are already designing chip-scale labs containing
all the equipment necessary to verify textbook quantum mechanics. As electronic
devices are made smaller and smaller, a good understanding of their behaviour
will only come through quantum-dot research, Kouwenhoven believes: 鈥淚n ten years
time [artificial atoms] will describe the fundamental physics of commercial
诲别惫颈肠别蝉.鈥�

But a single artificial atom is of limited use: the ability to link them
together while still obeying the laws of quantum mechanics is what would really
make this technology the tool of the future. And things are beginning to look
very promising indeed.

In December last year, researchers from the Walter Schottky Institute (WSI)
in Munich announced that they had achieved quantum-mechanical bonding between
artificial atoms. With this breakthrough, the road to artificial solids is
officially open. If two artificial atoms can bond using the same rules as real
atoms, then there may be no limit to the number of atoms that can be strung
together to form unnatural materials. Physicists may be about to redefine
鈥渕an-made鈥� material.

Getting a pair of artificial atoms close enough so they start to behave as a
single molecule isn鈥檛 easy. The German researchers found they had to construct
tiny pillars that were less than 60 nanometres apart before the electron in one
of the dots could be shared with a neighbouring dot to form a bond.

Werner Wegscheider, one of the WSI team, believes that, having formed an
artificial molecule, it should be possible to build 鈥渄esigner solids鈥� from a
host of artificial atoms joined together. Changing the bonding could give
researchers the chance to manipulate the properties of the material: eventually,
Wegscheider hopes, this technology could be used to produce new optical
materials, magnets and superconductors.

Artificial atoms may promise much stronger magnetic properties. Under the
influence of magnetic fields, for instance, the spins on all of the electrons in
an artificial atom align themselves with the field. In conventional magnetic
materials such as iron, only two of each atom鈥檚 electrons do this: 鈥淭he
artificial atom acts as a perfect ferromagnet,鈥� says Ashoori.

The pioneers of artificial atoms believe that researchers may learn to
engineer desirable properties into artificial solids by carefully tuning the
energy levels of the electrons that control the coupling between atoms. 鈥淭here鈥檚
an infinity of things you can create by changing the electron states,鈥� says
physicist Mark Reed of Yale University. 鈥淣ature gives you a certain
set鈥攖hat鈥檚 the periodic table. Now we鈥檙e moving into a different realm
where, hopefully, we can create electronic states at will.鈥�

But researchers are still cautious about real applications. Operating
temperature is one problem. At present, artificial atoms can only form below 1
kelvin鈥攆ar too cold to be of any real use outside the lab.

Size matters

Quantum dots only show their behaviour at extremely low temperatures because,
in the quantum world, size matters. The difference between the energy levels
that an electron can occupy depends on the diameter of the atom it is trapped
in, whether it is a real atom or not.

In real atoms, the electron鈥檚 energy levels are well-enough spaced to keep
electrons confined in their appropriate place at room temperature. But quantum
dots are much larger, so their energy levels have a much smaller separation.
Warm a dot up to room temperature, for instance, and thermal energy will give
its electrons enough energy to make random jumps between energy levels, smearing
out their quantum characteristics.

Overcoming this challenge is a major concern. 鈥淚 don鈥檛 think there鈥檚 any use
for the materials we have produced because they only work at these low
temperatures,鈥� says Wegscheider. 鈥淲e need better structures to make the effect
more stable at higher temperatures.鈥�

Moving artificial atoms toward room temperature operation will be a
significant challenge. It may only be overcome by making quantum dots using
combinations of semiconductors with extremely wide separations between the
energy levels that electrons can occupy.

But there are also other problems. To form an artificial solid with
controllable properties, physicists must create dots that are identical鈥攁n
extremely difficult task. 鈥淲hen you get down to these scales, the fluctuation of
a single atomic layer on a device has a dramatic effect,鈥� says Reed. 鈥淚t doesn鈥檛
give much room for error.鈥�

Reed believes that the top-down approach 鈥攅tching out pillars of
semiconductor鈥攎ay eventually prove useless for building artificial solids.
Luckily, researchers have other cards to play. 鈥淪elf-organised鈥� structures,
created by chemical synthesis of a vast number of identical molecules, or
growing semiconductor clumps, might turn out to be better ways of building
quantum dot arrays.

Whichever way the researchers turn, there is a great deal still to be done.
鈥淭he road to applications is a very long one,鈥� says Kouwenhoven. But while
technological difficulties may make the future look uncertain, the rate of
progress has been astonishing considering the field is less than ten years old.
If this technology develops at the same rate as silicon-based technology,
quantum dots could be used in real-world applications within 20 years.

There may be other applications for quantum dots, too. They could turn up in
new computing and memory devices, or be used as sensors to make highly sensitive
measurements of charge.

But nothing seems so enticing as the potential for artificial solids. Reed
may not be overstating the case when he talks about the creation of new
materials as 鈥渁 triumph of human ingenuity and imagination over the natural
rules鈥�. After all, improving existing technology is one thing, but using
fundamental particles to create new forms of matter鈥攖hat is the stuff of
scientists鈥� dreams.