Now, transistors made from carbon nanotubes have electronic characteristics that compete with silicon components. The concept of using molecular electronic components began at least in 1974, when Ari Aviram and Mark Ratner were at IBM in New York, and then put forward their theory at New York University, that is, placing molecules between two metal electrodes can act as rectifiers. However, it has taken more than 20 years for a single molecule to be successfully connected to the nano electrode in the laboratory. The difficulty lies in the ability to formulate the processing operation of signal molecules and establish the separation electrode with only a few nanometers in size.

But by the middle of 1990, an exciting new material, carbon nanotubes, had appeared. Carbon nanotubes are just like curled drawing paper with a diameter of several nanometers. Carbon nanotubes with a diameter of only a few nanometers have both metal and semiconductor habits according to their electronic arrangement. With the help of mechanical strength and nanometer length, research institutes around the world quickly linked them into macromolecules.

A common method for obtaining carbon nanotubes includes manufacturing input and output electrodes on a conductive substrate covered with an oxide layer, and then dropping the nanotubes onto the substrate. It is often called the substrate of the gate electrode. The nanotube can act as a two piece capacitor. Therefore, with the use of different voltages on the gate, the amount of charge carried on the nanotube changes.

The measurement of semiconductor nanotubes shows undesirable current characteristics, including increasing the gate voltage, that is, increasing the resistance by several orders of magnitude. The current characteristics of early components are very similar to those of conventional metal oxygen silicon field effect transistors (MOSFETs), and the operation is very difficult.

Since the first carbon nanotube field effect transistor was published by Cees Dekker Research Institute at Netherlands Delft University in 1998, exciting progress has been made in improving the operation of the original, especially in the last two months. These efforts appear in the recent reports of Phaedon Avouris and his collaborators working at IBM. Their improved carbon nanotube FETs can compete with the leading prototype silicon transistors currently used. （S Wind et.2002 .Phys.Appl Lett.80 3817）。    

The nanotube FET developed by Aouris's research team has a new scheme, which is similar to the conventional MOSFET structure and has a conductive channel gate. This arrangement means that the gap between the nanotube and the gate can be made very small, so the resistance is very sensitive to the change of the gate voltage. Indeed, the coupling between the gate and the nanotube is now strong enough to amplify the signal. In contrast, the output voltage change of the first generation nano transistor is too small to control the input of the second transistor, so the nanotube cannot be integrated into the circuit.

Semiconductor nanotubes typically operate like P-type semiconductors, so they conduct holes rather than electrons. This habit is caused by the molecules from the atmosphere are adsorbed on the nanotubes and the charge conversion of doped electrodes. It is a problem because adsorbed molecules affect the regeneration ability of the original. However, IBM has also made progress on this issue. They control molecular adsorption by embedding nanotubes in films and annealing in vacuum (V Derycke et al. 2002 Appl. Phys. Lett. 80 2773). Avouris and his collaborators found that the same process can be used to make n-type components, which conduct electrons, such as the alternating current method of doping nanotubes with alkali metals.

Perhaps the most important goal in the past few years has been to make components that conduct electrons better. High current means faster transistors that can produce high-power integrated circuits. The current of early nanotube components is limited by the internal resistance of nanotubes and the contact resistance on the electrodes. Recently, however, advanced new processes such as chemical vapor precipitation have produced high-quality materials and improved operating methods. For example, Michael Fuhrer and his collaborators recently reported on the production of carbon nanotubes that are higher than silicon MOSFETs and have a mobility of up to 20000 cm2v-1s-1 at the University of Maryland.

Significant progress has also been made in reducing resistance at the nanotube electrode interface. Paul McEuen of Cornell University and Hongjie Dai of Stanford University have produced large-diameter nanotubes, which have smaller band gap and thinner "Schottky barrier", making it easier for charges to pass through the barrier between nanotubes and metal electrodes. Another line proposed by the IBM research team is the internal energy heat related to the titanium electrode. It is a device with a new gate scheme. It can handle the boasting current of this method to 3uA and compete with the best primary silicon transistor.

All these efforts have been able to assemble different nanotubes into basic logic circuits, which is an important step towards nano electronics. These circuits manufactured by Delft, IBM and Chongwu Zhou's team at the University of Southern California in Los Angeles include: converters, logic NORs, electrostatic random access memory cells and ring oscillators.

Although the successful manufacturing of nanotube circuits has left great challenges, this direction is advanced, let alone exciting. But it is true that the current process is far from meeting the requirements of large-scale production. For example, in order to conduct the same amount of current as micro wide silicon transistors, we need to synthesize 2D chips with parallel tubes of considerable width. At the same time, it is still impossible to control the current properties of synthetic nanotubes, which results in a mixture of metal and semiconductor tubes. Perhaps the most important problem is the lack of control over how to place tubes in preset positions in equipment manufacturing.

A great deal of effort is being made towards this goal. In order to improve the current and even future performance of carbon nanotube FETs, many universities and industrial enterprises are working hard. We have seen amazing progress in the past four years, and we will hear more in the future.