MOLECULAR COMPUTING: INTEGRATION OF MOLECULES FOR NANOCOMPUTING - Bio-Inspired and Nanoscale Integrated Computing - page 305

between the dots but are unable to leave the cell [44]. Coulomb repulsion will
force the electrons to occupy dots on opposite corners. The two ground state
polarizations are both energetically equivalent and can be labeled logic ''0'' or
''1.'' Flipping the logic state of one cell, for instance by applying a nega-
tive potential to a lead near the quantum dot occupied by an electron, will
result in the next door cell flipping ground states in order to reduce Coulomb
repulsion. Amlani and co-workers have demonstrated experimental switching
of six-dot QCA cells with Al islands on an oxidized Si surface [45, 46].
They later demonstrated a functioning majority gate with logic AND and OR
operations [47].
While the use of quantum dots in the demonstration of QCA is a good first
step, the ultimate goal is to use individual molecules to hold the electrons and pass
electrostatic potentials down QCA wires. We have synthesized molecules that
have the capability of transferring information from one molecule to another
through electrostatic potentials (Fig. 11.6) [38]. The potentials only employ a
millionth of an electron per bit of information, which is extreme attractive in terms
of energy consumption.
Reshapes of the electron density due to the input signals produce electrostatic
interactions. The electrostatic potential interactions between molecules could
transport the information. They obviate electron currents or electron transfers
as in present devices: A small change in the electrostatic potential of one molecule
could suffice for intermolecular communication, leading to minute charge trans-
fer, far less than one electron. External fields or excitations are able to change the
boundary conditions of the molecule producing a change in the electrostatic
potential generated by these sources. As an example, a charge or field on the left
side of a molecule would reshape the electron density, providing a different
potential at the output side. The change observed in the electrostatic potentials is
in the range of values of nonbonded interactions, such as van der Waals
interactions, which are easily detected by neighboring molecules. In fact, these
are precisely the ranges of signal energies that would be attractive if we will
ultimately utilize large-scale integration in very small areas for power density and
heat dissipation considerations.
Although we synthesized molecules that included three-terminal molecular
junctions, switches, and molecular logic gates to demonstrate the electrostatics
methodology [38], none of the molecules were incorporated into an actual
assembly. All results were based on simulation only because the QCA and
electrostatics approach have major obstacles to overcome before even simple
laboratory tests can be attempted. While relatively large quantum dot arrays can
be fabricated using existing methods, a major problem is that placement of
molecules in precisely aligned arrays at the nanoscale is very difficult to achieve.
Another problem is that degradation of only one molecule in the array can cause
failure of the entire circuit. Even small examples of two-dots have yet to be
demonstrated using molecules because addressing of the molecular-sized inputs
and the recording of a signal based on fractions of an electron make the hurdles
enormous.