|
|
|
 |
|
Molecular Information Processing And Molecular Electronic Devices |
|
Fifth International Conference On Langmuir-Blodgett Films |
|
August 26-30, 1991 |
|
|
|
|
|
Joël de Rosnay
Director of Strategy
Cité des Sciences et de l'Insdustrie – La Villette – Paris – France
|
|
|
Computer chips with molecular sized circuits made from proteins still
belong to scientific fantasy. Yet reality has matched part of that dream
as microprocessors have shrunk to the size of a nail and transistor have
been made from semiconducting polymers.
The fields of molecular electronics and nanoelectronics have become
hot issues in the scientific and technological world in less then 10 years.
Thanks to the pioneering work of H. Kuhn in the 60's and André Barraud
from CEA, new materials made from LB films helped dozens of laboratories
around the world to study the properties of molecular electronic devices.
The first interdisciplinary research groups interested in molecular
electronics met in Washington at the Naval Research Lab in 1978 during
a workshop organized by the late Forest L. Carter. Since that time many
other conferences, seminars and workshops in the United States, Europe,
USSR and Japan have contributed to the fast progress of this field.
One of the most promising areas of molecular information processing
results from the convergence of microelectronics, solid state physics,
and molecular biology. This hybrid technology is leading to the development
of molecular electronic devices (MED). Technological breakthroughs in this
area are opening the way to the development of "chemical computers."
Molecular electronic devices could represent a significant breakthrough
for computers in the next 30 years. From the tube era (40s to 60s), to
the transistor era (60s to 80s), we are now entering into the molecular
electron devices era (90s to 2000 and beyond). The challenge is enormous.
We need to start from scratch and reinvent all the components of present
day micro-electronics.
Scientists will have to synthezise molecules sensitive to outside stimuli
(like photons, magnetic or electric fields) and which present new properties
when connected together For example as conductors or semiconductors, optoelectronic,
piezo-electric or iono-electronic devices. Then the researchers will need
to control the cooperative properties of supra-molecular assemblies made
of such molecules through LB films, orientation of liquid cristals or growth
of ordered films of molecular material. Eventually such supra-molecular
assemblies will have to be integrated in electronic, opto-electronic, magnetic
or catalytic components.
These components will mimic the fundamental elements of present day
micro-electronics.
1- First, engineers have to work on a switch, able to shift information
in one state or another. By interrogating this component. one should be
able to find out in which state the switch is.
2- The second major achievement will be the building of a true
memory with different switches, that is, an array of molecules which could
undergo reversible alterations, and can therefore be reused.
3- The third major step is to construct molecular wires, able
to transport information through distances, like conjugated chains of carbon
atoms on which solitons move rapidly in order to connect switches and memories.
4- The fourth step will be represented by the assembly of switches,
memories and wires in three-dimensional structures or arrays, organized
in assemblies of modules at several levels of communication and interconnections,
able to perform coordinated functions.
This is where one of the major breakthroughs of the new molecular engineering
techniques will come into play : from the properties of atoms and molecules,
the system will self-assemble in highly organized structures, putting atoms,
molecules, and macromolecules into place, in order to perform dedicated
functions. At this step, immunology techniques could be used to fit molecules
at predetermined spots, or to deposit heavy metals which could be used
as conductors at other levels of communications.
5- Finally, the system should be reparable. Modules which do not
perform correctly will have to be detected, corrections made, and sometimes
a device replaced. These types of self-repairing automata already exist
in biological systems. We begin to understand how they function at the
molecular level, and how they self-assemble. Other supramolecular assemblies,
like the quantosome, the photosome, the oxysome are small self-contained
factories which perform important functions for the cell. These minute
devices are packed with an enormous amount of molecules and information.
We are slowly starting to understand their structures, and are able for
the first time to copy them. But we still have to learn more from nature's
design, understand how nature operates, and then translate what we have
learnt into the new micromachines.
What has been achieved by research labs around the world to meet such
exciting challenges ?
In Strasbourg, France, the group headed by Jean-Marie Lehn, Nobel Price
in Chemistry has been able to build molecular wires and DNA-like double
helical molecules.
Last year Francis Garnier and his group of the Laboratoire des Matériaux
Moléculaires du CNRS à Thiais produced what is recognized world-wide as
the first "plastic transistor" made of polythiophène and with
properties comparable to those of devices made from amorphous silicon.
A new technique developped by Jean-Claude Wittmann and Paul Smith of
the University of Santa Barbara in California provides a versatile method
to produce ordered molecular films from a wide range of materials. They
use a thin single cristal-like film of PTFE ( Polytetrafluoroethylene)
deposited mechanically on a smooth surface like glass. This film acts as
a matrix to align different materials in highly oriented thin films.
A key issue in nanoelectronics is the connexion of molecular wires
to create circuits. James Tour and Jeffry Schumm of the University of South
Carolina at Columbia have been able to cross-connect molecular wires made
of short chains of polythiophene rings through a junction box using silicon
tetrachloride.
An other important topic is related to molecular memory networks. Last
year a Japanese group of the University of Tokyo headed by Z. Liu has produced
LB films of a photochromic memory molecule derived from azobenzene wich
changes shapes when it is struck by a photon of ultraviolet light allowing
it to store a binary digit.
Molecular diodes is also a hot subjet since the pioneering work of
A. Aviram of IBM since 1974. Junction between silicon and organic conducting
polymers to produce diodes is an important step on that road. Michael Sailor
and his team from CalTech recently produced diodes usig a low-temperature
chemical procedure in which contact to the semiconductor is made by a layer
of polyacetylene.
Ultimatly it will become possible to manipulate single molecules or
clusters of molecules on any type of supra-molecular assembly through the
use of a scanning and tunelling electron microscope (STEM) as demonstrated
by Phaedon Avouris and In-Whan Lyo from IBM on silicon surfaces.
These are only a few examples of recent developments in the field of
molecular electronics. Further progress relies on the use of integrated
and multidisciplinary strategies. For instance, in order to make ultramicrocircuits,
several laboratories are trying to synthesize proteins which do not exist
in nature. Such syntheses are feasible using present day genetic engineering
techniques, automatic gene and protein synthesis, and computer reproductions
of the bi and tridimensional structures of amino acid sequences. It is
likely that from this research a large variety of molecular electronic
devices and ultramicroeircuits could be conceived and manufactured. They
could be rendered biocompatible, thus allowing the production of implantable
logic circuits, offering the prospect of direct interface between the central
nervous system of animals or human beings and computers. Such biocompatible
circuits are now being implanted and tested in the brains of rats. Other
applications of these ultramicrocircuits could be the production of prostheses
for the blind, different transducers, or solar energy converters on soft
plastic sheets.
The production and assembly of these molecular circuits can be considered
from two different approaches: a "passive" one (successive depositions,
etching, grafting, doping) using technologies close to those presently
used in the manufacturing of microcircuits; or an "active" one
resulting from the spontaneous "growth" of the molecular circuit.
In fact, automatic machines used today for synthesizing and analyzing genes
and proteins, offer new models which may inspire the automation of the
successive operations of molecular circuit production: growth of polymers,
successive washing, reactions with other active groups, blocking, and reactivation
of chemical groups.
However, at the level where the presence or absence of a single chemical
link can affect the performance of an entire circuit, it is virtually impossible
to construct and assemble circuits with traditional macroscopic control
techniques. It becomes necessary to use auto-assembly properties of biological
macromolecules, observed for instance in Langmuir-Blodgett films or during
the auto-organization of viruses or predissociated cellular organelles.
In other words, instead of introducing the information from the outside,
as we do today with most of our machinery (drill press, lathe, or even
car construction robots), we will use information from the biopolymers
themselves. Such information stored in the primary sequences of amino acids,
allows the three dimensional folding of a protein. It is thus possible
to benefit from the properties of biological macromolecules to assemble
three dimensional molecular ultramicrocircuits.
Many questions still remain unanswered : will these ultramicrocircuits
be repairable? Will we be able to selectively break chemical linkages or
rearrange them? Is it still necessary to utilize boolean logic, presently
used in all computers? Will we be inspired by the neuronal networks of
the brain? These circuits presently work, in all probability, in a non-boolean
fashion, using parallel processing. To build the logic of the future, the
convergence between molecular neurobiology and micro-electronics is in
the forefront and it holds great promises. |
|
