ScienceWeek Free Shareware Edition July 28, 2000

Claire Haller prismx at
Mon Jul 31 20:01:35 EST 2000

This is the Free Shareware Edition of SCIENCEWEEK. To upgrade to
the regular edition:

If you are not already subscribed to the Free Shareware Edition
of SCIENCEWEEK and you wish to receive it by Email each week,
transmit "SUBSCRIBE SWX" to request at
To unsubscribe, transmit "REMOVE SWX" to request at
Note: You may freely transmit or post this issue of SW anywhere,
provided you transmit or post it complete and without editing.

SCIENCE-WEEK (Free Shareware Abridged Edition)

A Weekly Email Digest of the News of Science

A journal devoted to the improvement of communication
between the scientific disciplines, and between scientists,
science educators, and science policy-makers.

July 28, 2000 -- Vol. 4 Number 30


We are the strangest species. We question everything,
measure the stars, sift the sand through our fingers,
gauge the bowels of the Earth. It is our destiny and
it will not stop.
-- Anonymous


1. Computer Science:
Prospects in Molecular Electronics
Within the next few decades a new technology will supplant the
transistor paradigm that currently forms the basis of computing
machines, and many believe this new technology will involve
molecular scale devices and the utilization of quantum effects
manifested at such scales.
(Includes related background material.)

2. Condensed-Matter Physics:
On the Optical Activity of Quantum Dots

3. Earth Sciences:
On Ocean Tides and Climate Change

4. Evolutionary Biology:
On the Origin of the Cell Nucleus and the Woese-Mayr Controversy

5. Plant Biology:
On Photosynthesis

6. Medical Biology:
Environment vs. Heredity in the Causation of Cancer

7. Focus Report:
On Cosmic Heat Death

8. From the SW Archive:
Marietta Blau: The Destruction of a Career in Physics

[To receive a copy of the complete issue of this week's
ScienceWeek via Email, send one dollar US to ScienceWeek, 3023 N.
Clark Street #109, Chicago, IL 60657 USA. Be sure to include your
Email address and issue date and number.]


It is apparent that within the next few decades there will be
developed a new technology that will supplant the transistor
paradigm that currently forms the basis of computing machines. It
is also apparent that this new technology will involve molecular
scale devices and the utilization of quantum effects manifested
at such scales. The basic science and engineering challenges are
enormous, but there is now such intense global hunger for
increased computing power (and excitement over the prospects)
that success in overcoming the obstacles seems highly probable.
... ... Craig S. Lent (University of Notre Dame, US) presents a
commentary on current efforts in this field, the author making
the following points:
     1) Modern computers are the result of two ideas: a) the use
of binary numbers to represent information in a machine, and b)
encoding the binary "1" and "0" as the "on" and "off" states of a
current switch. Prior to approximately 1940, electromechanical
relays were used as the current switch devices. Later, these
switches were replaced by vacuum tube triodes ("valves"), which
in turn were eventually replaced by the solid-state version of
the vacuum tube, the modern transistor. In general, representing
binary information by turning a current stream on or off has been
one of the most fruitful ideas in the history of technology.
     2) This idea, however, has serious drawbacks as device sizes
are reduced. The smaller the switch, the lower its ability to
cleanly turn the current off and on. Also, the current through
each single switch is reduced, making it more difficult to charge
up the interconnect lines between devices. Finally, switching
current requires electrons to move from the power supply through
a resistance to ground, resulting in considerable energy
dissipation. If present transistors could be magically shrunk to
the size of single molecules, a chip with phenomenal device
density could be fabricated, but the heat generated would cause
such a chip to melt as soon as it was turned on.
     3) It is thus clear that if electronic devices are to be
shrunk to the ultimate limit of molecular size, a new paradigm
must be developed. The Quantum-dot Cellular Automata (QCA)
paradigm retains the notion of a binary representation of
information, but this binary information is stored in the charge
configuration within a quantum-mechanical microscale system
("cell") rather than in the on/off state of a current switch. In
general, each cell consists of four "dots" and contains two
mobile electrons, the electrons occupying antipodal sites. In a
molecular implementation, each cell is a single molecule. No
current flows between cells. Instead, the Coulomb interaction
between cells tends to induce the same state in neighboring
cells, and this interconnection is sufficient to support general-
purpose computing. [Editor's note: For more on "quantum dots",
see related background material below and report #2 this issue.]
     4) At the present time, researchers are exploring a class of
molecules known as metal cluster carboxylates for use as QCA
molecules. Each of the four dots is composed of a cluster of
three metal atoms, with additional metal atoms forming a bridge.
Electrons hop from one outer dot to another through the central
cluster, which also acts structurally to place the four dots on a
plane. The molecules have been synthesized with several different
clusters and centers in an effort to understand the role of the
linkers in electron hopping and to design molecules with
appropriate switching behavior. Another experimental design
involves the use of single ruthenium atoms as the dots of a QCA
     5) The author concludes: "The QCA approach to molecular
electronics is very promising, but creating functional QCA
molecules is just the first step. In a device, the molecules need
to be attached to a surface in a predetermined geometry, inputs
and clocking signals need to be applied, and the state of the
output cells must be read. Each of these steps presents
substantial challenges. In addition, the whole approach to
circuit architecture must be rethought if circuits are going to
be based on QCA cells rather than transistors."
Craig S. Lent: Bypassing the transistor paradigm.
(Science 2 Jun 00 288:1597)
QY: Craig S. Lent [lent at]
Summary by SCIENCE-WEEK 28Jul00
For more information:
Related Background:
     The consensus view among computer scientists and
microelectronics engineers is that in the near future the rapid
growth of information processing capability will require radical
innovations in technology because of inherent limitations in
current microchip (integrated circuit) architecture and dynamics.
This has caused attention to focus on various possibilities for
the use of quantum mechanical effects in computing. One such
possibility involves the "single-electron transistor". A
single-electron transistor is a transistor of extremely small
dimensions isolated from its leads by potential barriers narrow
enough to permit electron tunneling, with a minute electron
source that is essentially a droplet of electrons. A
single-electron transistor switches on and off with the addition
of each electron, in contrast with the ordinary transistor which
sustains a switched-on state given a flow of added electrons.
This quantized behavior of the single-electron transistor is due
to its dimensions, the electron droplet essentially behaving as
an artificial atom or "quantum dot". As realized in the
laboratory, quantum dots are small electrically conducting
regions, typically less than 1 micron in diameter, that contain
from one to a few thousand electrons. Because of the small
volume, the electron energies within the dot are quantized, and
the behavior of the quantum dot is intermediate between that of
an atom and that of a classical macroscopic object.
     In general, in this context, the term "cellular automata"
refers to theoretical computer processing units consisting of a
large number of cells, each cell containing a relatively small
number of responding entities (devices), with each cell
communicating only with neighboring cells. The main advantage of
cellular automata systems is that they eliminate the requirement
for long interconnections between devices, a requirement that is
one of the ultimate limitations of the speed of the conventional
computer microchip. What is of interest here is that quantum dots
can be configured as quantum cellular automata, and such
configurations have been shown to have the ability to perform
logic operations. Until now, however, these devices have worked
only at very low temperatures (millikelvins) -- unless the dots
are made extremely small (less than 2 nanometers in diameter).
... ... R.P. Cowburn and M.E. Welland (University of Cambridge,
UK) now report experimental development of quantum cellular
automata using relatively large quantum dots (approximately 100
nanometers) that work at room temperature, provided one uses
magnetic metals in the construction of the dots. The authors make
the following points:
     1) In the system developed by the authors, each quantum
cellular automata network consisted of a single elongated input
quantum dot followed by a chain of 69 circular dots. Each dot was
110 nanometers in diameter and placed so that the edge of each
dot was 25 nanometers from the edge of each neighboring dot
(i.e., pitch = 135 nanometers). The dots were 10 nanometers thick
and made from a common magnetic alloy (Supermalloy), which is
Ni(sub80)Fe(sub14)Mo(sub5)X, where X is other metals, all of the
dots on a single-crystal silicon substrate. The dots were
fabricated by high-resolution *electron-beam lithography.
     2) The authors point out that in electronic quantum cellular
automata, the term "quantum" is used because the system involves
*quantum mechanical tunneling of charge between dots to change a
logic state; classical electrostatics are involved thereafter in
the propagation of the change in logic state. The quantum
mechanical interactions in magnetic quantum cellular automata
networks are *exchange interactions between *spins within a
single dot in order to form a single giant classical spin. A
logic 1 is signaled when the magnetization vector of the dot
points to the right, for example, and a logic 0 when it points to
the left. The magnetic field emanating from such a magnetic
particle can be extremely large, with the result that one
magnetic dot is strongly influenced by the magnetic field
emanating from its nearest neighbor. These classical
magnetostatic interactions are then involved in the propagation
of information along the chain of dots. A further feature of
magnetostatic interactions is that they force the magnetization
to point along the length of the chain. The system is thus
intrinsically binary, with only right- and left-pointing
magnetization states being stable. An applied oscillating
magnetic field feeds energy into the system and serves as a
     3) The authors state: "These networks offer a several
thousandfold increase in integration density and a hundredfold
reduction in power dissipation over current microelectronic
R.P. Cowburn and M.E. Welland: Room temperature magnetic quantum
cellular automata.
(Science 25 Feb 00 287:1466)
Text Notes:
... ... *electron-beam lithography: In this context, lithography
is a technique used for integrated circuit fabrication, the
technique in general involving a silicon chip coated uniformly
with a radiation-sensitive film ("resist"), and an exposing
radiation source (e.g., ultraviolet light or an electron beam)
illuminating selected areas of the surface through an intervening
master template (mask) to obtain a particular pattern of resist-
coated surface after unexposed resist is washed away. Non-resist
coated portions of the silicon surface are then etched away by
acid. In electron-beam lithography, the radiation-sensitive film
used in microchip fabrication is placed in the vacuum chamber of
a scanning-beam electron microscope and exposed by an electron
beam under digital computer control. In the present report, the
quantum dots were fabricated by high-resolution electron-beam
lithography in a polymethylmethacrylate resist followed by
metalization and ultrasonically-assisted lift-off in acetone. The
report contains photographs of linear arrays of identical quantum
dots 110 nanometers in diameter fabricated by this method.
... ... *quantum mechanical tunneling: "Tunneling" is a quantum
mechanical phenomenon involving an effective penetration of an
energy barrier resulting from the width of the barrier being less
than the wavelength of the particle.
... ... *exchange interactions: In quantum mechanics, an
"exchange interaction" is an interaction represented by an
exchange of space or spin coordinates or both --  an interaction
that can be viewed as an effective exchange of particles.
... ... *spins: In quantum mechanics, electrons, protons, and
neutrons have an intrinsic angular momentum known as "spin", and
a *magnetic moment parallel or antiparallel to that angular
momentum. When electrons are combined together to form an atom or
ion, there is a resultant angular momentum which is a combination
of the intrinsic spin of the electrons and the angular momentum
due to their motion about the nucleus, and this is the "spin" of
the atom or ion. Atoms or ions with non-zero spin are magnetic
atoms or ions.
... ... *magnetic moment: (magnetic dipole moment) The intrinsic
spins of the electrons in an atom, together with the motion of
the electrons around the nucleus, give rise to a magnetic field
around the atom, and the magnitude of this field is related to
the magnetic dipole moment of the atom or ion.
Summary & Notes by SCIENCE-WEEK 12May00
For more information:
Related Background:
It is now possible to create extremely small crystals which
contain less than 1000 atoms, each crystal measuring a few
millionths of a millimeter across and thus in the nanoscale
domain. Certain of these nano-crystals, those of cadmium
selenide, for example, have peculiar attributes: crystals of
exactly the same composition but of different size exhibit quite
different properties, with the large nano-crystals of cadmium
selenide red in color, smaller crystals orange, and the smallest
(containing barely 100 atoms) yellow in color. The differences in
properties are due to quantum mechanical effects. These extremely
small atomic arrays are called "quantum dots", and there is a
current consensus that if quantum dots could be integrated onto a
chip, their unique electrical properties could be harnessed to
perform a function similar to a conventional transistor, while
requiring only a small fraction of the space. In consequence, the
creation of an appropriate regular array of quantum dots would
allow a computer processor many times more powerful than any
current supercomputer to be constructed on single chip.
... ... F. Remacle and R.D. Levine (2 installations, BE IL)
present a theoretical discussion of assemblies of metallic
quantum dots with each dot considered as an "atom". The dots are
taken as being packed close enough to be interacting. The authors
suggest that the key point is that such dots are essentially
"designer" atoms, since their electronic properties can be
controlled via the synthetic method used to prepare the dots. Of
direct significance are the size of the dot and the nature of the
ligands used to prevent coalescence of the dots. The energy
required to remove or add an electron to the dot is determined by
the size of the dot. The ligands control how closely the dots can
be packed and hence the strength of the coupling between adjacent
dots. An important parameter is the energy cost of adding an
electron to a dot: because of the large size of the dots, the
Coulomb repulsion of the added electron is low. Unlike most
ordinary atoms, quantum dots have a high capacity for
accommodating an additional electron.
F. Remacle and R.D. Levine: Architecture with designer atoms:
Simple theoretical considerations.
(Proc. Natl. Acad. Sci. US 18 Jan 00 97:553)
QY: R.D. Levine [rafi at]
Summary by SCIENCE-WEEK 21Apr00
For more information:
Related Background:
In general, a "transistor" is a semiconductor device in which it
is possible to control voltage or current in such a way as to
achieve gain or switching action. An "integrated circuit" is a
miniature electronic circuit produced within a single crystal of
a semiconductor such as silicon. Such devices, usually called
"chips", range from simple logic circuits to large-scale circuits
containing approximately 10^(6) components (transistors,
resistors, capacitors), and these devices are widely used in
memory circuits, microcomputers, pocket calculators, etc.,
because of their low cost and high speed.
... ... Robert D. Miller (IBM Almaden Research Center, US)
presents a review of current research in chip design, the author
making the following points:
     1) Within the next few years, high-performance chips
containing as many as 5 x 10^(8) transistors on a single chip
will be produced. These advanced chips may contain up to 10^(4)
meters of on-chip wiring. Such increased component and wiring
densities, however, cannot be achieved with currently used
materials. Although an intensive search is now underway for
materials that can replace silicon dioxide [SiO(sub2)] as the
insulator in these future devices, a clear candidate material has
yet to be identified.
     2) In a typical microchip, layers of copper-interconnect
wiring are separated by a dielectric insulator, traditionally
silicon dioxide. Both the resistance of the metal and the
capacitance of the insulator increase markedly as the wiring
dimensions and distances between chip components (pitch)
decrease, with resulting crosstalk, capacitative coupling between
metal-interconnect lines, and consequent increased signal delays.
Traditional aluminum-copper wiring can be replaced by pure
copper, which has a lower resistance, and then the performance
gain is limited primarily by interlayer and intralayer
capacitance, which in turn is dictated primarily by the
*dielectric constant of the insulator. Thus, there is now a
intensive search underway for new dielectric insulators with
lower dielectric constants than silicon dioxide. (The dielectric
constant (k) of silicon dioxide is in the range 3.9 to 4.2.)
     3) Any replacement low-dielectric constant material must
meet current requirements for integrated circuits: thermal
stability in excess of 400 degrees centigrade, good mechanical
properties, low ion content, breakdown fields in excess of 2
million volts per centimeter, low water uptake, lithographic
processability, low thermal expansion coefficients and film
stresses, good adhesion to a variety of substrates, and low
reactivity with conductor metals at elevated temperatures.
     4) Although the drive toward increased device densities and
improved performance in semiconductor devices makes the switch
from silicon dioxide on-chip insulators to low-dielectric
constant materials inevitable, no clear winner has yet emerged
among materials with dielectric constants less than 3.0. The
author concludes: "The switch to low-k on-chip insulators
continues to be a formidable challenge to chemists, physicists,
materials scientists, and integration engineers."
Robert D. Miller: In search of low-k dielectrics.
(Science 15 Oct 99 286:421)
QY: Robert D. Miller [rdmiller at]
Text Notes:
... ... *dielectric constant: In general, a "dielectric" is a
substance that can sustain an electric field and act as an
insulator, e.g., a nonconductor of electric charge in which an
applied electric field causes a displacement of charge but not a
flow of charge. In physics, "permittivity" is the ratio of the
electric displacement in a dielectric medium to the applied
electric field strength, and "relative permittivity" refers to
the ratio of the permittivity of a medium to the permittivity of
*free space. Relative permittivities vary from 1 (for free space)
to over 4000 for certain ferromagnetic materials, but for most
materials, relative permittivities are less than 10. In general,
in physics, the term "dielectric constant" has been replaced by
the term "relative permittivity", but in the context of electric
circuits, the term dielectric constant is still used, and in this
context the term "dielectric constant" of a material is perhaps
best defined as the ratio of the capacitance of a capacitor
constructed of the material to the capacitance the capacitor
would possess if the material were replaced by free space.
... ... *free space: In physics, the term "free space" refers to
a region in which there is no matter and no electromagnetic or
gravitational fields, the region having a temperature of absolute
zero, unit refractive index, and the speed of light at its
maximum value.
Summary & Notes by SCIENCE-WEEK [] 31Dec99
[For more information:]


This free shareware abridged edition of ScienceWeek contains one
or more complete reports, but it does not contain all the reports
present in the regular edition (see Contents at beginning).
Regular edition subscriptions are available at:

Queries concerning subscriptions to: info at

There is no print edition of ScienceWeek. The first issue of
SCIENCE-WEEK appeared online May 1, 1997, and it has been
published regularly each week since that date.

Editor/Publisher: Dan Agin dpa at
Managing Editor: Claire Haller haller at
Associate Editor: Joan Oliner oliner at

Copyright (c) 1997-2000 SCIENCE-WEEK/Spectrum Press Inc.
All Rights Reserved
US Library of Congress ISSN 1529-1472
This publication is protected by U.S. and International Copyright
Laws. The free shareware edition may be copied, retransmitted,
and distributed only as a complete intact text containing both
front and back material. Changes in the text are prohibited
without explicit permission from the publisher. SCIENCE-WEEK is
published by Spectrum Press Inc., 3023 N. Clark Street #109,
Chicago, 60657-5205 IL, USA.
-----end file

More information about the Bioforum mailing list