Page 78 - Nanotechnology an introduction

Figure 7.5 Input–output relationships approaching the ideal of a step (Heaviside) function. The input might be a voltage (e.g., applied to the coil of a relay, Figure 7.2, or the gate of a transistor, Figure 7.6) and the output might be the
current flowing through the rest of the circuit. Note that in this example the response characterized by a thick solid line will give an output of one for any input (voltage) exceeding about 0.6. For the response characterized by the dashed
line, the input would have to exceed about 0.8; i.e., it is less tolerant to deviations from the ideal (the input of one yielding an output of one). The dotted line marks a possible hysteretic response when decreasing the input from 1 to 0 and
beyond to negative values, the existence of which opens up the possibility of using the device as a memory element.

Figure 7.6 A field-effect transistor (FET). Regions marked “n” and “p” are n-type and p-type semiconductors (e.g., appropriately doped silicon). The hatched region is an insulator (e.g., silica). Conductors S, G and D are, respectively, the
source, gate and drain. Application of a voltage to G (which plays the role of the coil in the relay) increases the concentration of conduction electrons in the p region and allows current to flow from S to D.
A related device is an information store, or memory. A relay or transistor having the property of bistability could function as an information store
(memory), with the disadvantage that it would need to be constantly supplied with electrical power. A more elaborate relay, with two separate coils
for switching the current “on” and “off”, would be better in this regard, since once its position had been flipped, power could be cut off (see Figure
7.5). Read-only memories do not even require the flipping to be reversible: an early type of read-only memory was paper tape in which holes were
punched. Reading was carried out by passing the tape between a pair of electrically conducting rollers. In the absence of a hole, there would be no
electrical contact between the rollers. A later development was the use of ferromagnets, which could be poled “up” or “down”. Since
ferromagnetism cannot exist below a certain volume (cf. Section 2.6), this technology is not suitable for ultimate nanoscale miniaturization, but this
limit is still far from being reached—the current limitation is the sensitivity of the magnetic field detector (reading head). Memories based on
electrical resistance can be fabricated from materials (e.g., NiO) that can be switched from a conducting to an insulating state by applying a voltage
pulse. Other materials can have their phase changed from amorphous to crystalline by exposure to light or by passing an electric current, with
corresponding changes in reflectance and resistance, but these materials are not especially “nano”.
Apart from the logic gates acting as the components of information processors, the other main types of device to be considered are sensors and
actuators. A sensor has a clear transduction function. Examples are a magnetic sensor that registers whether the spin in a memory cell is “up” or
“down”; a light sensor such as a photodiode that converts light into electricity; a chemical sensor that converts the presence of a certain chemical
compound into electricity or light. The main issue in miniaturization is whether the signal exceeds the noise level. An example of an actuator is the
coil in the relay (Figure 7.2).
7.3. Quantum Computing
Extrapolation of Moore's law to about the year 2020 indicates that component's sizes will be sufficiently small for the behavior of electrons within
them to be perturbed by quantum effects. This implies a profound perturbation of the proper functioning of the technology, and no solution to this
problem within the current framework is in view. Quantum computing can be thought of as “making a virtue out of necessity”, creating computational
devices based on the principles of quantum logic. The essence of quantum computing is to use not binary logic bits as the representation of the
elementary unit of information, but qubits, which can exist in an infinite number of superpositions of the binary units zero and one (Figure 7.7).

Figure 7.7 Qubits. Top: qubits can be represented by spheres (here simplified as disks), a point on the circumference (indicated by an arrow from the center) indicating the net result of the superpositions of all its states. A measurement
forces the state into either 1 (with a probability of 0.7 in this example) or 0 (with the complementary probability of 0.3). Bottom: the quantum logic gate. To create an inverter (the NOT operation, cf. Figure 7.3) the arrow must be rotated by