Carbon nanotubes (CNT) were first systematically described by Sumio Iijima of NEC, who discovered them in 1991 as a by-product of C₆₀ fullerene synthesis [1], and almost simultaneously by a group of researchers led by L.A. Chernozatonsky [2]. The existence of extraordinary forms of carbon with similar morphology had been mentioned before [3,4], but those research efforts remained unnoticed.

A graphene sheet may be wrapped into a regular cylinder along different directions, which gives rise to a broad family of nanotubes. Single walled carbon nanotubes (SWCNTs) are characterised by the chiral vector (n,m), which links pairs of atoms that coincide upon this imaginary process of wrapping, where n and m (n ≥ m) are coordinates of this vector in the basis of lattice vectors of a graphene sheet. Depending on the values of n and m, nanotubes may exhibit totally different properties: nanotubes with n – m divisible by 3 are metallic (or narrow-gap semiconductors) while the rest of the nanotubes are semiconductors, although their band gap approaches zero with the increase of the diameter. The values of n and m uniquely define the diameter and band structure of nanotubes, which is broadly used for their characterisation using electron (absorption and fluorescence) and vibrational Raman spectroscopy.

There are “zigzag” nanotubes, also known as (n, 0) nanotubes and “armchair” (n,n) nanotubes. These two classes of nanotubes are optically inactive while all other nanotubes are chiral. Besides SWCNTs, there are multi-walled carbon nanotubes (MWCNTs) made up of several single-walled nanotubes inserted one into another. Another distinction lies between open and capped nanotubes. In capped nanotubes, the ends are closed with dome-shaped carbon caps that include six pentagonal faces and constitute halves of certain fullerene molecules. With the higher curvature of these caps causing them to be more reactive than the cylindrical walls, capped nanotubes may be transformed into the open by controlled oxidation. The latter approach, combined with ultrasonic treatment, is also an approach to cut long nanotubes into shorter fragments.

There are several techniques of manufacturing nanotubes. Originally, they were produced using the arc discharge technique, similarly to fullerene synthesis, that yielded mixtures of SWCNTs and MWCNTs. Later, a technique based on laser ablation (see pulsed laser deposition) of graphite in the presence of metal particles (cobalt, nickel) acting as catalysts was proposed. This technique made it possible to produce primarily single-walled nanotubes with controllable diameters and good yields.

Lately, techniques based on vapour deposition have been gaining popularity as the most commercially viable methods. These techniques are based on the thermal decomposition of carbon-containing gases (carbon monoxide, lower hydrocarbons and alcohols, or more complex molecules) on catalytic nanoparticles of metals, which results in the growth of nanotubes from their catalyst-bound end.

In the plasma enhanced deposition technique, the direction of nanotubes’ growth can be controlled via manipulating the electric field. Vapour deposition techniques are used to produce dense linear nanotube arrays with thickness (array height) of up to several millimetres and make it possible to control the type of nanotubes formed.

Separation of nanotubes is an important issue since particular applications may require nanotubes of a certain type (e.g., metallic or semiconductor nanotubes) in a non-aggregated state, whereas as-synthesised nanotubes may be quite firmly bonded into strands due to Van der Waals interactions. Existing separation methods employ centrifugation, electrophoresis, chromatography, etc. Single nanotubes can be obtained using different surfactants and even nanotube-DNA systems. Perhaps researchers will be able to address many present challenges in the field when they master more advanced techniques for the directed catalytic synthesis of nanotubes of desired types.

Nanotubes may find application in a wide range of industries due to their unique electrical, magnetic, optical and mechanical properties. For example, CNTs are an order of magnitude stronger than steel; the Young modulus of SWCNT reaches the order of 1–5 TPa. The latter fact has triggered interest in modulating the strength of materials via the addition of nanotubes. Nanotubes can be used in organic diodes and field effect transistors, and current density in metallic nanotubes may be several orders greater than in metals. Molecular electronics can considerably benefit from the use of defective nanotubes where local defects may bind nanotubes of different types and may even create triplex (branched) contacts.

Scientists are studying potential applications of nanotubes in innovative ultra-strong and ultralight composite materials. Nanotubes are used as needles in scanning tunneling and atomic force microscopy, as well as in the development of semiconductor heterostructures. Prototypes of thin flat displays based on CNT matrices have been designed and tested. In this respect, of importance is the essential difference between nanotubes and many conventional materials: the anisotropy of their properties. While nanotubes show extremely high electric and thermal conductivity along the tube axis, in the lateral directions they act as insulators.