Number 524, February 8, 2001 by Phil Schewe, James
Riordon, and Ben Stein
Data Storage densities of 100 Gbit/sq-in
Data storage densities
of 100 Gbit/sq-in or greater may be possible using a
special patterned magnetic medium demonstrated by physicists at IBM-Almaden.
By comparison, today's highest-density disk drives store up to 20 billion
bits (20 Gb) per square inch of disk surface. In recent years, increasing
data storage density has meant decreasing the number of magnetic grains
needed to store a bit of data from a thousand down to a few hundreds
and also
shrinking the size of the grains themselves.
A patterned medium
allows a different approach. By cutting into the magnetic medium with
a focused ion beam, arrays of isolated magnetic islands are created.
Charles Rettner, Bruce Terris (408-927-1517; terris@almaden.ibm.com)
and colleagues at IBM showed that when the islands are made small enough--below
130 nm--each will have only a single magnetic domain. Since each bit
is now only a single magnetic domain, the domains are large enough to
be thermally stable (not susceptible to thermally excited reversal at
room temperature) even at very high bit areal densities. The IBM team
demonstrated writing and reading on arrays patterned at densities up
to 100 Gbit/in2.
The lab samples
are still quite small, and the patterning method must be extended economically
over many square inches before actual products could be produced. Related
support technology must also follow along, such as advanced read-write
heads, air bearings that can fly reliably over a patterned-media surface,
and mechanical actuators capable of positioning the head accurately
over data bits only a few tens of nanometers across, which is more than
an order of magnitude narrower than today's bits. (Lohau et al., Applied
Physics Letters, 12 February 2001.)
Negative Heat Capacity
Physicists at
the University of Freiburg in Germany have performed an experiment in
which clusters of sodium atoms respond to added energy by cooling down.
The clusters, typically consisting of 147 atoms, are made by blowing
cold helium gas over a surface of boiling sodium. This leads to formation
of clusters in a process which is similar to cloud formation in nature.
The clusters are swept by the helium gas into a cell, where they are
cooled or heated to some temperature. Afterwards the clusters are sorted
by size and irradiated by a laser.
The laser light
can fragment the clusters and the Freiburg group has developed a method
on how to read the energy (i.e. the energy before the laser light was
absorbed) from the fragmentation pattern. Near the melting point of
the cluster, the measured internal energy can actually decrease even
as the temperature rises. This may sound counter-intuitive, but is in
agreement with theory, and no law of thermodynamics is violated.
Negative heat capacity
has been predicted to occur in such systems as stars and atomic nuclei
in the act of fragmentation, but this is the first time the phenomenon
has been observed experimentally in atom clusters. (Schmidt et al.,
Physical Review Letters, 12 February; contact Hellmut Haberland,
49-761-203-5726, haberland@physik.uni-freiburg.de)
Near-field Scanner for Moving Molecules
A near-field scanner
for moving molecules has been built and demonstrated
by a multinational research team (Robert Austin, Princeton, 609-258-4353,
rha@suiling.princeton.edu),
offering a potentially fast way to make high-resolution images of molecules
such as DNA. Traditional scanning-probe
microscopes offer molecular-level images, but at the cost of slow scanning
speeds for large molecules.
In the new device,
molecules travel in a microscopic fluid channel (5 microns wide by 1
micron deep) and pass directly under a trio of 100-nm-wide slits that
are just a few hundred nanometers above the molecules. The fluid channel
contains an array of posts to stretch out the DNA molecules. A laser
causes the molecules to fluoresce, providing light that yields an image.
The slits' narrow width, along with their proximity to the molecules,
enables high-resolution images, 200-nm resolution in this initial experiment.
To ensure high-quality
images, the microscope accepts data only from those molecules that pass
through the three slits at roughly equal time intervals.
For a DNA molecule
with 200,000 base pairs (corresponding to about 74 microns in stretched
form), the researchers obtained imaging data in just 100 milliseconds,
considerably faster than AFM or traditional near-field optical microscopes.
Resolution improvements are possible by narrowing the slits or making
them thinner; future versions of the device will employ shallower fluid
channels for confining DNA molecules to a greater degree.
Ultimately, the
researchers envision massively parallel data acquisition by creating
multiple slits that simultaneously scan many molecules. This microscope
design could potentially obtain high-resolution maps of the binding
sites of repressor/promoter proteins critical for the expression of
genes, part of an emerging field called epigenetics. (Tegenfeldt
et al., Physical Review
Letters, 12 February 2001.)
The Muon's Magnetic Moment is Misbehaving
In developing
a better theory for describing how electrons interact with light, Richard
Feynman and others showed that certain mathematical problems with quantum
theory, such as calculations becoming infinitely large, could be avoided
by reinventing the electron, as it were. This could be done by taking
into account all the possible interactions between the electron and
different combinations of virtual particles hiding out in the universal
vacuum. These interactions, portrayed graphically in Feynman diagrams
(which he invented for the purpose), serve to "renormalize"
the electron and, in the process, tame all the mathematical catastrophes
of the earlier theory.
Eventually Feynman's
theory, quantum electrodynamics (QED), was expanded to take into account
the weak and strong nuclear forces. One of the predictions made by QED
(since subsumed into the Standard Model) is that the strength of an
electron's inherent magnetism, its magnetic moment, should depart slightly
from its value in the absence of the interactions with virtual particles.
Physicists readily test this proposition since it is an area where both
theory and experiment can attain very high precision. In practice, though,
one often uses muons instead of electrons since the expected modification
to the magnetic moment gets larger with mass, and the muon is some 200
times heavier than its cousin the electron, which outweighs the difficulty
of making the muons in the first place (they don't ordinarily exist)
and having them decay quickly (but not before they can be studied).
Scientists at Brookhaven,
looking at the decay of a billion muons, have now detected a further
anomalous magnetic moment beyond what the Standard Model predicts. The
new results, reported at a Brookhaven seminar today (co-spokesmen are
Vernon Hughes of Yale and Lee Roberts of Boston Univ) correspond to
a divergence of 2.6 "standard deviations" from the Standard
Model, not yet a definitive statement, but enough to spur discussion
of explanations outside the customary model. These include the possible
effect of hypothetical "supersymmetric" particles. Meanwhile,
the experimental team continues to analyze a data sample some four times
the present sample, so an even higher precision test will be forthcoming.
(http://www.phy.bnl.gov/g2muon/home.html)
Enlarge text
Shrink text
Print
E-mail
Digg this
Del.icio.us
Furl