The Development of Quantum-Dot Cellular Automata - Field-Coupled Nanocomputing: Paradigms, Progress, and Perspectives - page 7

of stray charge in the vicinity of the QCA cell, which affects the bias point and
polarization degeneracy of the cell. The arrangement of this charge changes with time,
and the gate biases applied to each dot in the cell must be adjusted to keep the cell
operational. The low operating temperature of metal dot QCA cell is due to the size of
the lithographically defined cell.
The metal dot QCA are composed of aluminum islands separated by tunnel
junctions. Fabrication of the cell is done by electron-beam lithography using the
Dolan bridge technique [ 7 ] where the tunnel junctions are formed by evaporation of
aluminum from two angles, with an intervening oxidation step. The resulting tunnel
junctions are composed of two layers of aluminum separated by a thin layer, 1-2 nm,
of aluminum oxide. The area of the overlap between the two layers of aluminum
determines the capacitance of the junction, and since it is typically the dominant
capacitance of the dot, determines the operating temperature of the QCA cell.
Cells and Logic
The first QCA cell was demonstrated in 1997 [ 36 ]. This device had a junction overlap
area of approximately 50 9 50 nm, giving an operating temperature of 70 mK. As it
was the first demonstration, the layout was very conservative and optimized for high
yield, which resulted in a relatively large overlap and low operating temperature. In
this first demonstration the goal was to use gate electrodes to move an electron
between the top and bottom dots on the left side of a cell. The electron in the right half
of the cell would move in the opposite direction to maintain the lowest energy con-
figuration. To measure the polarization of the cell, single-electron transistors, which
are the most sensitive electrometers demonstrated to date [ 37 ], are used to measure the
potential of the dots in the right half of the cell. Measurements of the output of the two
electrometers move in opposite directions, confirming that an electron in the right half
moves in the opposite direction to the electron in the left half, confirming QCA
operation. Full details of the experimental methods are given elsewhere [ 38 - 41 ].
The next step in the development of metal dot QCA was the demonstration of a
logic gate [ 42 ]. The basic logic element in the QCA paradigm is the majority gate,
where three inputs vote on the polarization of a QCA cell. For this experiment we
again used metal dots defined by the Dolan bridge method. For the inputs of the
majority gate we applied voltages to the input electrodes that mimicked the potentials
of three input cells. The applied voltages were varied to step through the logic truth
table. The polarization of the cell was measured by electrometers and the output of the
cell confirmed proper operation of the gate.
These experiments showed the basic functionality of QCA cells. The next
experiment [ 43 ] showed that a QCA line could switch without getting stuck in a
metastable, partially switched, state. In this experiment three 2-dot cells were fabri-
cated in a line, and an input applied to the left side of the line. Electrometers coupled
to the output side of the line confirmed the proper switching.
Power Gain
These initial experiments used unclocked QCA cells, but clocking is an important
element in QCA systems. Clocking of QCA cells is crucial to achieve perhaps the