[ Pobierz całość w formacie PDF ]
confined electron into a complex tangle which resembles, in some ways, the
quantum waveform of an orbital. (Image courtesy of Charles Marcus)
"We might still dispense with temperature," Marcus tells me, "Except that
decoherence is strongly correlated with it. At
1 Kelvin, our decoherence time is extremely short -- 100
picoseconds. Colder, at 100 millikelvin, it jumps to 1
nanosecond, which is comparable to the clock cycle of a 1-
gigahertz computer."
Long enough, in other words, to be caught in the act, before the wave
functions collapse into classical particles.
Long enough to be computationally useful. And at lower temperatures,
decoherence can be staved off for even longer.
"For electrons," Marcus notes, "the Pauli exclusion principle tells you that
Page 58
ABC Amber Palm Converter, http://www.processtext.com/abcpalm.html
if low-energy states are filled up, they're off-
limits to another electron. That's the origin of the effect.
The temperature dependence of dephasing is from the fact that there's nowhere
for the electrons to scatter to when all the states are full."
As another physicist, Ken Wharton, explains, the temperature correlation may
also depend on "the sheer number of interactions between the quantum state and
the rest of the universe. In general, higher temperatures tend to equal more
interactions, because there are a lot more blackbody photons emitted from hot
surfaces, which can then be absorbed and destroy atomic superpositions. But
photon - photon interactions have such a low cross section you don't have to
worry about it for optical quantum states. A photon that's in a quantum
Hacking Matter Thermodynamics and the Limits of the Possible
81
superposition is therefore going to be a lot more stable at room temperature."
This may be one more reason to downplay the application of chilled quantum
dots to the field of computing; optical quantum computers appear to sidestep
all the messy thermodynamics and simply function at room temperature. Marcus
cautions, though, that these effects can't be produced by ordinary linear
optics, but rely instead on the nonlinear responses of certain exotic
materials -- including quantum dots. So the primary advantage of quantum dot
computers may not be in qbits or single-electron transistors, but in the
near-instantaneous reconfiguring of optical and electrical properties. We'll
return to this concept later; the point is simply that decoherence, while
important, affects only a small subset of quantum dot capabilities.
As Bawendi's work clearly shows, the world has many potential uses for quantum
confinement and atomlike behaviors at higher temperatures -- a point Marcus
happily concedes.
"There's a lot of interest in buckyballs and nanotubes for this reason: they
work in the quantum domain at room temperature.
They have a low buy-in cost, and they're easy to fool around with. They're
interesting because they're very small, yet very easy to make."
The term "buckyball" is short for "buckminsterfullerene," a form of carbon
first discovered by Richard Smalley and Robert
Curl at Rice University in 1985. The molecule, about 1.4 nm across, contains
60 carbon atoms arranged in a spherical pattern of pentagons and hexagons that
is, by sheer coincidence, identical to a soccer ball. In mathematical terms,
this shape is a truncated icosahedron, but its more popular name is an homage
to architect Buckminster Fuller, who used it (and related shapes) for the
construction of domes from flat polygons. But
C turns out to be only the material's stablest form; the same
60
chemical and physical processes can produce larger or smaller molecules in a
variety of shapes, collectively known as
"fullerenes." Nanotubes are a specific form of fullerene, essentially a
buckyball which has been cut in half and used to cap a graphite cylinder of
arbitrary length. By the late 1990s, a powder of purified, single-walled
nanotubes, vaguely resembling black felt, was available on the Internet for as
little as $200 a gram.
Page 59
ABC Amber Palm Converter, http://www.processtext.com/abcpalm.html
Hacking Matter Thermodynamics and the Limits of the Possible
82
Figure 4-2: Buckyball and Nanotube
These intricate molecules, made of pure carbon, display interesting quantum
peroperties even at room temperature. (Image courtesy of Richard Smalley and
Rice
University.)
There is considerable evidence that fullerenes can act as quantum wires and
quantum dots. This is not surprising considering their small size, and
considering that carbon can, under some circumstances, act as a semiconductor
or metal. (In fact, as a conductor it supports much higher current densities
than normal metals do, and even sometimes displays superconductivity at liquid
nitrogen temperatures, or perhaps even higher.) A buckyball is different than
a Bawendi-type CdSe dot of equivalent size, though, because it's hollow.
Every atom in the ball is a surface atom, and every electron trapped on it
will be trapped on its surface rather than bouncing around in its interior.
This makes their actual behavior hard to predict or model with current
techniques, while their small size makes it extremely difficult to attach
wires to them for any sort of direct measurement. What we know about their
properties generally comes from monkeying with bulk materials rather than
individual molecules.
There are also "dopeyballs" and "dopeytubes," which have other atoms trapped
inside the fullerene cage, and these materials have properties that are even
more different and even more complex. But in the end, it may not matter;
[ Pobierz całość w formacie PDF ]